With the release of Fedora Core 5 Test 2, the Fedora Core Steering Committee would like to announce the transfer of Fedora Core 3 to the Fedora Legacy Project.
For more information on the Fedora Legacy Project, or if you wish to join the team, please see http://fedoralegacy.org/.
Bill Nottingham wrote:
With the release of Fedora Core 5 Test 2, the Fedora Core Steering Committee would like to announce the transfer of Fedora Core 3 to the Fedora Legacy Project.
I would like the Fedora Core Steering Committee to change its mind.
Suely, the time to retire FC3 is after one can sanely upgrade to FC5. Say, a month after releasing FC5, not a month before.
John Summerfied wrote:
Bill Nottingham wrote:
With the release of Fedora Core 5 Test 2, the Fedora Core Steering Committee would like to announce the transfer of Fedora Core 3 to the Fedora Legacy Project.
I would like the Fedora Core Steering Committee to change its mind.
Suely, the time to retire FC3 is after one can sanely upgrade to FC5. Say, a month after releasing FC5, not a month before.
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
http://fedoraproject.org/wiki/Legacy
From: "Rahul Sundaram" sundaram@redhat.com
John Summerfied wrote:
Bill Nottingham wrote:
With the release of Fedora Core 5 Test 2, the Fedora Core Steering Committee would like to announce the transfer of Fedora Core 3 to the Fedora Legacy Project.
I would like the Fedora Core Steering Committee to change its mind.
Suely, the time to retire FC3 is after one can sanely upgrade to FC5. Say, a month after releasing FC5, not a month before.
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
It seems John and I understand it. We still think it was an ill considered decision.
{^_^}
Hi
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
It seems John and I understand it. We still think it was an ill considered decision.
Why?. If this is a community project, then the community has some right and responsibility to contribute. If you care about older releases why not step up and do the required work in getting Fedora Legacy going better?. I am sure everyone in the team would welcome your contributions.
--- Rahul
Fedora Bug Triaging - http://fedoraproject.org/wiki/BugZappers
Rahul Sundaram wrote:
Hi
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
It seems John and I understand it. We still think it was an ill considered decision.
Why?. If this is a community project, then the community has some right
I should think that the answer was contained in the prior message. AFTER it is reasonable to install FC5 rather than before. Anyway, policy is policy, and is what it is.
and responsibility to contribute. If you care about older releases why not step up and do the required work in getting Fedora Legacy going better?. I am sure everyone in the team would welcome your contributions.
I think you missed the point somehow. Some people think that FC5 isn't ready yet.
Mike
Hi
and responsibility to contribute. If you care about older releases why not step up and do the required work in getting Fedora Legacy going better?. I am sure everyone in the team would welcome your contributions.
I think you missed the point somehow. Some people think that FC5 isn't ready yet.
It doesn't have to be. Fedora Core 4 still gets updates from Fedora Core developers. Fedora Legacy is ready. You will continue to receive updates for Red Hat Linux 7.2, 8, Fedora Core 1,2,3.
Rahul Sundaram wrote:
Hi
and responsibility to contribute. If you care about older releases why not step up and do the required work in getting Fedora Legacy going better?. I am sure everyone in the team would welcome your contributions.
I think you missed the point somehow. Some people think that FC5 isn't ready yet.
It doesn't have to be. Fedora Core 4 still gets updates from Fedora Core developers. Fedora Legacy is ready. You will continue to receive updates for Red Hat Linux 7.2, 8, Fedora Core 1,2,3.
I thought 7.2 was dropped. Nine too.
If we wanted FC4, we'd have installed it ages ago. In fact, had _I_ installed any FC machines in the past six months ago, I would have installed FC4.
At this time in its life, I don't care for FC4; if I must choose between upgrading my FC3 boxes and switching to Legacy, the choice I want is to upgrade to FC5.
Hi
If we wanted FC4, we'd have installed it ages ago. In fact, had _I_ installed any FC machines in the past six months ago, I would have installed FC4.
At this time in its life, I don't care for FC4; if I must choose between upgrading my FC3 boxes and switching to Legacy, the choice I want is to upgrade to FC5.
Thats your choice. Wait till FC5 is released and upgrade. There is no major security bugs left unfixed in FC3. All of those was taken care of before handing it over to the legacy team. So you can just enable legacy and get updates if you want to for a period of merely two months.
On Tue, 2006-17-01 at 14:35 +0800, John Summerfied wrote:
Rahul Sundaram wrote:
Hi
and responsibility to contribute. If you care about older releases why not step up and do the required work in getting Fedora Legacy going better?. I am sure everyone in the team would welcome your contributions.
I think you missed the point somehow. Some people think that FC5 isn't ready yet.
It doesn't have to be. Fedora Core 4 still gets updates from Fedora Core developers. Fedora Legacy is ready. You will continue to receive updates for Red Hat Linux 7.2, 8, Fedora Core 1,2,3.
I thought 7.2 was dropped. Nine too.
If we wanted FC4, we'd have installed it ages ago. In fact, had _I_ installed any FC machines in the past six months ago, I would have installed FC4.
At this time in its life, I don't care for FC4; if I must choose between upgrading my FC3 boxes and switching to Legacy, the choice I want is to upgrade to FC5.
I will add to this, saying FC4 had a very rocky start, and I don't recall any announcement that the ISO images were fixed and re-released. With the nightmare I went through getting FC3 installed and the reported problems installing FC4 seeming trivial in comparison, I have not wanted to chance it.
I have tried FC5t2 and it is a little fussy but I was able to get it to install.
I also intend on moving straight from FC3 to FC5.
I have great hopes for FC5 and appreciate the hard work done by the developers.
Hi
I will add to this, saying FC4 had a very rocky start, and I don't recall any announcement that the ISO images were fixed and re-released.
An few informal updates were made available at http://fedora.isphuset.no/
On Fri, 2006-20-01 at 05:31 +0530, Rahul Sundaram wrote:
Hi
I will add to this, saying FC4 had a very rocky start, and I don't recall any announcement that the ISO images were fixed and re-released.
An few informal updates were made available at http://fedora.isphuset.no/
Why informal and on a site I would never think to go to?
Guy Fraser wrote:
On Fri, 2006-20-01 at 05:31 +0530, Rahul Sundaram wrote:
Hi
I will add to this, saying FC4 had a very rocky start, and I don't recall any announcement that the ISO images were fixed and re-released.
An few informal updates were made available at http://fedora.isphuset.no/
Why informal and on a site I would never think to go to?
If you read my earlier mails we are working on guidelines and collaborating with many such efforts to help them be a formal project if they are interested in this. For this particular project, we have recently started discussing it. Lets see.
On Fri, 2006-27-01 at 04:33 +0530, Rahul Sundaram wrote:
Guy Fraser wrote:
On Fri, 2006-20-01 at 05:31 +0530, Rahul Sundaram wrote: ...
Hi ...
I will add to this, saying FC4 had a very rocky start, and I don't recall any announcement that the ISO images were fixed and re-released. ...
An few informal updates were made available at http://fedora.isphuset.no/ ...
Why informal and on a site I would never think to go to? ...
If you read my earlier mails we are working on guidelines and collaborating with many such efforts to help them be a formal project if they are interested in this. For this particular project, we have recently started discussing it. Lets see.
... Sorry Rahul,
My eye were getting blurry, I had to skip a couple hundred messages.
Rahul Sundaram wrote:
Hi
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
It seems John and I understand it. We still think it was an ill considered decision.
Why?. If this is a community project, then the community has some right and responsibility to contribute. If you care about older releases why not step up and do the required work in getting Fedora Legacy going better?. I am sure everyone in the team would welcome your contributions.
--- Rahul Fedora Bug Triaging - http://fedoraproject.org/wiki/BugZappers
I I don't want to join it for the sake of a month; I wish to go from three to five. I prefer not to rejig my yum configurations and/or up2date configurations, and I don't wish to go to four at this late state.
"responsibility to contribute."
Jo and I don't contribute? You gotta be joking. We've both been helping out on these lists for years (with a minor break or two when RH left us).
Hi
I I don't want to join it for the sake of a month; I wish to go from three to five. I prefer not to rejig my yum configurations and/or up2date configurations, and I don't wish to go to four at this late state.
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
"responsibility to contribute."
Jo and I don't contribute? You gotta be joking. We've both been helping out on these lists for years (with a minor break or two when RH left us).
If you would like to get updates from Fedora Legacy project for a longer duration you will have to contribute to that project specifically which is what I meant earlier in context.
Rahul Sundaram wrote:
Hi
I I don't want to join it for the sake of a month; I wish to go from three to five. I prefer not to rejig my yum configurations and/or up2date configurations, and I don't wish to go to four at this late state.
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
Take your fingers out of you ears; I alread said that's not what I want to do.
"responsibility to contribute."
Jo and I don't contribute? You gotta be joking. We've both been helping out on these lists for years (with a minor break or two when RH left us).
If you would like to get updates from Fedora Legacy project for a longer duration you will have to contribute to that project specifically which is what I meant earlier in context.
We don't want to get updates from FL. We'd like support for the outgoing release (FC3 now, FC4 next time) to remain as it was until the second successor is out and (somewhat) stabilised, so there are always two releases supported by FP.
Jo and I (and a lot of others) switched from RH back when RH announced its changed arrangements, and the commencement of the Fedora project because neither RH not FC provided what we wanted.
We've dabbled in others including Debian, Ubuntu, SUSE, Mandrake/Mandribble and some others.
I currently run Debian/Woody, Debian/Sarge, Ubuntu, FC3, SUSE 10 and WBEL, so I know what the major alternatives are,
We're asking just for a teensy extension to the support offered by FP. Instead of 13 months for FC3, maybe 14 (FC5+1 say, instead of FC5-1).
Hi
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
Take your fingers out of you ears; I alread said that's not what I want to do.
Thats your personal choice. If you choose to refuse updates from FL, that a project issue.
Jo and I (and a lot of others) switched from RH back when RH announced its changed arrangements, and the commencement of the Fedora project because neither RH not FC provided what we wanted.
Fedora Core 3 was supported by 1 year and 2 months. Thats far more than previous release Red Hat Linux not even counting the updates from FL.
We're asking just for a teensy extension to the support offered by FP. Instead of 13 months for FC3, maybe 14 (FC5+1 say, instead of FC5-1).
Present you detailed rationale to fedora-devel list then.
Rahul Sundaram wrote:
Hi
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
Take your fingers out of you ears; I alread said that's not what I want to do.
Ok, but isn't that a little bit of a harsh way to say it? I think that Rahul has used a very even tone so far. If I may ask, why do you have an aversion to FL? I have been using FL with my machine for FC2 since it was declared "stale" and have been very satisfied. It's not hard to set up for updates from there (I use "yum"), and the people are very nice and helpful.
Thats your personal choice. If you choose to refuse updates from FL, that a project issue.
Jo and I (and a lot of others) switched from RH back when RH announced its changed arrangements, and the commencement of the Fedora project because neither RH not FC provided what we wanted.
Fedora Core 3 was supported by 1 year and 2 months. Thats far more than previous release Red Hat Linux not even counting the updates from FL.
Well, this is another topic, I think. In some parts, Red Hat is considered notorious for churn. Trying to justify passing FC3 off by referring to previous track record is *ahem* somewhat weak, IMO.
We're asking just for a teensy extension to the support offered by FP. Instead of 13 months for FC3, maybe 14 (FC5+1 say, instead of FC5-1).
Present you detailed rationale to fedora-devel list then.
Seems like a reasonable course of action.
Mike
Hi
Well, this is another topic, I think. In some parts, Red Hat is considered notorious for churn. Trying to justify passing FC3 off by referring to previous track record is *ahem* somewhat weak, IMO
Might be a famous change rather than a notorious change. Release early, Release often ;-)
On Tue, 2006-17-01 at 12:50 +0530, Rahul Sundaram wrote:
Hi
Well, this is another topic, I think. In some parts, Red Hat is considered notorious for churn. Trying to justify passing FC3 off by referring to previous track record is *ahem* somewhat weak, IMO
Might be a famous change rather than a notorious change. Release early, Release often ;-)
Especially if you mean release as in "catch and release" rather than "new product release".
Guy Fraser wrote:
On Tue, 2006-17-01 at 12:50 +0530, Rahul Sundaram wrote:
Hi
Well, this is another topic, I think. In some parts, Red Hat is considered notorious for churn. Trying to justify passing FC3 off by referring to previous track record is *ahem* somewhat weak, IMO
Might be a famous change rather than a notorious change. Release early, Release often ;-)
Especially if you mean release as in "catch and release" rather than "new product release".
Kindly help in testing if you want a better release.
Mike McCarty wrote:
Rahul Sundaram wrote:
Hi
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
Take your fingers out of you ears; I alread said that's not what I want to do.
Ok, but isn't that a little bit of a harsh way to say it? I think that Rahul has used a very even tone so far. If I may ask, why do you have an aversion to FL? I have been using FL with my machine for FC2 since it was declared "stale" and have been very satisfied. It's not hard to set up for updates from there (I use "yum"), and the people are very nice and helpful.
It seemed to me that Rahul was being particularly obtuse, and does not recognise that the Fedora Project's service forces people into undesireable choices. It makes me wonder whether he properly understands "support."
It doesn't provide as good a service as, with just a little more effort, it could.
Thats your personal choice. If you choose to refuse updates from FL, that a project issue.
The issue is that the project doesn't provide quite the service some of its users think it ought.
I have better thinks to do than make pointless changes to my computer systems.
Jo and I (and a lot of others) switched from RH back when RH announced its changed arrangements, and the commencement of the Fedora project because neither RH not FC provided what we wanted.
Fedora Core 3 was supported by 1 year and 2 months. Thats far more than previous release Red Hat Linux not even counting the updates from FL.
Well, this is another topic, I think. In some parts, Red Hat is considered notorious for churn. Trying to justify passing FC3 off by referring to previous track record is *ahem* somewhat weak, IMO.
We're asking just for a teensy extension to the support offered by FP. Instead of 13 months for FC3, maybe 14 (FC5+1 say, instead of FC5-1).
Present you detailed rationale to fedora-devel list then.
Seems like a reasonable course of action.
OTOH I am a user (and a supporter), but not a developer, and the level of support is a user concern.
Hi
It seemed to me that Rahul was being particularly obtuse, and does not recognise that the Fedora Project's service forces people into undesireable choices. It makes me wonder whether he properly understands "support."
Setting aside the fact, that Red Hat nor me has any obligation to provide support services for Fedora, Fedora Project provides does provide both updates through legacy as a service. Unsure why you would consider my responses obtuse. You havent expanded on why you wouldnt choose to use Fedora Legacy updates or what concerns you have in waiting till the GA release of Fedora Core 5 considering the fact that there isnt any major security issues with FC3 that needs to be addressed before FC5.
I have better thinks to do than make pointless changes to my computer systems.
What is pointless specifically? Would Fedora legacy repository available by default resolve this issue?
Seems like a reasonable course of action.
OTOH I am a user (and a supporter), but not a developer, and the level of support is a user concern.
It has to be addressed by the developers so if you want any action to be done, you have to post to the developer's list. Users cannot address any development related concerns.
Rahul Sundaram wrote:
Hi
It seemed to me that Rahul was being particularly obtuse, and does not recognise that the Fedora Project's service forces people into undesireable choices. It makes me wonder whether he properly understands "support."
Setting aside the fact, that Red Hat nor me has any obligation to provide support services for Fedora, Fedora Project provides does provide both updates through legacy as a service. Unsure why you would consider my responses obtuse. You havent expanded on why you wouldnt choose to use Fedora Legacy updates or what concerns you have in waiting
Fedora Project isn't providing as good a service as it could, with just a teensy bit more effort. Retire the oldest release _after_ the second later release instead of before it.
If all the security problems really are fixed, it won't even cost signficant effort.
I thought I did offer some reasons I don't wish to use FL; even if you don't understand my reluctance to use it, you don't need to, any more than Ford needs understand why some people prefer white cars.
Let me try again. Today, IMV upgrading to FC4 is not sensible. It offers neither the latest techology, nor any reasonable lifetime, nor the stability (I'm guessing, but them my opinion's the one that colours my decisions) of FC3.
I am prepared to update FC3 to a new release when a sensible choice is available. FC5 will, I expect offer new technlogy and a life of about a year, two advantages over FC4. I'll take a chance on stability[as in not crashing] (but not where stability is of primary concern).
Now, there is a window where Fedora Project will not support the release I am using at present, and so forcing me to a decision I should not need to make.
I don't mind that FL exists, and I am in fact using it to support a RHL 7.3 box that's doing all it needs to do and has for years.
However, I resent that FP presumes to force me to use FL, all for the lack of a lick of paint.
till the GA release of Fedora Core 5 considering the fact that there isnt any major security issues with FC3 that needs to be addressed before FC5.
The fact you don't know about any security issues doesn't mean one won't surface in the next five minutes.
I have better thinks to do than make pointless changes to my computer systems.
What is pointless specifically? Would Fedora legacy repository available by default resolve this issue?
Seems like a reasonable course of action.
OTOH I am a user (and a supporter), but not a developer, and the level of support is a user concern.
It has to be addressed by the developers so if you want any action to be done, you have to post to the developer's list. Users cannot address any development related concerns.
Again, I don't think it a reasonable expectation that all of us users who want to discuss this matter should be expected to join the developers' list and the FP list and the ...
There _are_ developers on this list, and this list is archived so the developers can catch up on the discussion. As can the FL folk.
Hi
Fedora Project isn't providing as good a service as it could, with just a teensy bit more effort. Retire the oldest release _after_ the second later release instead of before it.
If all the security problems really are fixed, it won't even cost signficant effort.
I thought I did offer some reasons I don't wish to use FL; even if you don't understand my reluctance to use it, you don't need to, any more than Ford needs understand why some people prefer white cars.
Let me try again. Today, IMV upgrading to FC4 is not sensible. It offers neither the latest techology, nor any reasonable lifetime, nor the stability (I'm guessing, but them my opinion's the one that colours my decisions) of FC3.
Your explanations here wont have the desired effect here. It might in fedora-devel list. Thats what I have been trying to communicate to you.
The fact you don't know about any security issues doesn't mean one won't surface in the next five minutes.
You can help in Fedora Legacy to fix them.
There _are_ developers on this list, and this list is archived so the developers can catch up on the discussion. As can the FL folk.
You can be expecting all the relevant developers involved in the decision making processes of release cycle and release lifetime to sign up to this list and track all the discussions here.
Hi
There _are_ developers on this list, and this list is archived so the developers can catch up on the discussion. As can the FL folk.
You can be expecting all the relevant developers involved in the decision making processes of release cycle and release lifetime to sign up to this list and track all the discussions here.
Err. I meant you *cant be.
Rahul Sundaram wrote:
There _are_ developers on this list, and this list is archived so the developers can catch up on the discussion. As can the FL folk.
You can't be expecting all the relevant developers involved in the decision making processes of release cycle and release lifetime to sign up to this list and track all the discussions here.
It's no less sensible than your expectation all us users join the developer list:-)
However, that's not my suggestion. Read again.
Hi
You can't be expecting all the relevant developers involved in the decision making processes of release cycle and release lifetime to sign up to this list and track all the discussions here.
It's no less sensible than your expectation all us users join the developer list:-)
However, that's not my suggestion. Read again.
You can subscribe and post to the list if you are interested. Not asking all of the users to do it. Only those who want to talk about development related changes.
From: "Rahul Sundaram" sundaram@redhat.com
Hi
Fedora Project isn't providing as good a service as it could, with just a teensy bit more effort. Retire the oldest release _after_ the second later release instead of before it.
If all the security problems really are fixed, it won't even cost signficant effort.
I thought I did offer some reasons I don't wish to use FL; even if you don't understand my reluctance to use it, you don't need to, any more than Ford needs understand why some people prefer white cars.
Let me try again. Today, IMV upgrading to FC4 is not sensible. It offers neither the latest techology, nor any reasonable lifetime, nor the stability (I'm guessing, but them my opinion's the one that colours my decisions) of FC3.
Your explanations here wont have the desired effect here. It might in fedora-devel list. Thats what I have been trying to communicate to you.
John, I suspect beleagured Rahul has a rather significant point here. You can visit the devel list, make sure you're understood there even if shouted down, then resign the list. A few weeks should suffice. It is a pain; but, it does help to go where the intended audience is.
The fact you don't know about any security issues doesn't mean one won't surface in the next five minutes.
You can help in Fedora Legacy to fix them.
There _are_ developers on this list, and this list is archived so the developers can catch up on the discussion. As can the FL folk.
You can be expecting all the relevant developers involved in the decision making processes of release cycle and release lifetime to sign up to this list and track all the discussions here.
Rahul, I suspect it would be a good idea if they visited this list from time to time to get a sense of the "buzz". That is what I do with the LKML list. (Which is what convinces me that 2.6 is a development kernel rather than a stable kernel, even number notwithstanding. 2.6.15 hardly resembles 2.6.12 in much of any regard. And 16's going to be farther afield. It seems like Linus has lost control. But that's another rant, or should be. {^_-} I simply sort new messages by subject, scan the subjects to see what is happening, and read maybe two or three messages a day that look like they affect items in which I am interested. Then I scan some of the messages in long running argument threads to see what is going on. I don't do that too often. I get disillusioned easily.)
{^_^}
On Tue, 2006-01-17 at 17:53, jdow wrote:
Rahul, I suspect it would be a good idea if they visited this list from time to time to get a sense of the "buzz". That is what I do with the LKML list. (Which is what convinces me that 2.6 is a development kernel rather than a stable kernel, even number notwithstanding. 2.6.15 hardly resembles 2.6.12 in much of any regard. And 16's going to be farther afield. It seems like Linus has lost control. But that's another rant, or should be. {^_-}
You make it sound like that's a surprise - or a change. Serious problems were still being fixed up though about X.X.20 of the previous kernel "stable" releases.
From: "Les Mikesell" lesmikesell@gmail.com
On Tue, 2006-01-17 at 17:53, jdow wrote:
Rahul, I suspect it would be a good idea if they visited this list from time to time to get a sense of the "buzz". That is what I do with the LKML list. (Which is what convinces me that 2.6 is a development kernel rather than a stable kernel, even number notwithstanding. 2.6.15 hardly resembles 2.6.12 in much of any regard. And 16's going to be farther afield. It seems like Linus has lost control. But that's another rant, or should be. {^_-}
You make it sound like that's a surprise - or a change. Serious problems were still being fixed up though about X.X.20 of the previous kernel "stable" releases.
Oddly, I see a basic difference between making extreme changes in the modules in the Kernel and fixing bugs. The changes are big enough that they introduce their own serious bugs from time to time. (Latest is NFS crawling at about 1/4 speed. One thing I looked at closely was the changes to NTP operation. I have tools for setting up a GPS reference here. But the kernel's been changing that area faster than I can keep up. So I've not been motivated to get that GPS antenna outside yet.)
{^_^} Joanne
On Tue, 2006-01-17 at 20:11, jdow wrote:
You make it sound like that's a surprise - or a change. Serious problems were still being fixed up though about X.X.20 of the previous kernel "stable" releases.
Oddly, I see a basic difference between making extreme changes in the modules in the Kernel and fixing bugs.
How is it different if the unfixed bugs affect what you are trying to do? Did you, for example, try to use NFS in a heterogeneous network when it was moved into the 2.2 kernel?
Hi
Rahul, I suspect it would be a good idea if they visited this list from time to time to get a sense of the "buzz". That is what I do with the LKML list.
Some of them are active in the lists or read it atleast regularly but knowing the schedule of the developers involved I am very reluctant to ask them do more work in keep tracking of the list. They rather be concentrating on their expertise of churning out code that will have a better impact. I have an effort on progress to help everyone including the developers and users of Fedora understand and keep track of important changes and discussions within the project by providing summarized weekly reports. I am hoping that the relevant developers would benefit from this too.
http://fedoraproject.org/wiki/Projects
If you anyone here who read/participate in this list would like to summarize it on a weekly basis in a short way you can contribute - http://fedoraproject.org/wiki/Projects/WeeklyReports/Contribute. Contact me off list if you need any help with that.
From: "Rahul Sundaram" sundaram@redhat.com
Hi
Rahul, I suspect it would be a good idea if they visited this list from time to time to get a sense of the "buzz". That is what I do with the LKML list.
Some of them are active in the lists or read it atleast regularly but knowing the schedule of the developers involved I am very reluctant to ask them do more work in keep tracking of the list. They rather be concentrating on their expertise of churning out code that will have a better impact. I have an effort on progress to help everyone including the developers and users of Fedora understand and keep track of important changes and discussions within the project by providing summarized weekly reports. I am hoping that the relevant developers would benefit from this too.
http://fedoraproject.org/wiki/Projects
If you anyone here who read/participate in this list would like to summarize it on a weekly basis in a short way you can contribute - http://fedoraproject.org/wiki/Projects/WeeklyReports/Contribute. Contact me off list if you need any help with that.
That's a good idea. It'd be good if people here on this list volunteered some time summarizing what goes on here that might be of special developer interest.
{^_^}
On Tue, 2006-01-17 at 16:23, John Summerfied wrote:
I thought I did offer some reasons I don't wish to use FL; even if you don't understand my reluctance to use it, you don't need to, any more than Ford needs understand why some people prefer white cars.
Do you expect Ford to always supply the color of your choice regardless of what that might be? If you do proceed to make this argument to the developers or legacy list, please present it in terms of more than your own best interest. There are reasons to make the transition transparent, but your color preference probably isn't the most compelling one.
However, I resent that FP presumes to force me to use FL, all for the lack of a lick of paint.
It shouldn't come as a surprise.
till the GA release of Fedora Core 5 considering the fact that there isnt any major security issues with FC3 that needs to be addressed before FC5.
The fact you don't know about any security issues doesn't mean one won't surface in the next five minutes.
And if it does, it would be reasonable to assume that the legacy support group will make the effort to back-port the fix.
On Wed, Jan 18, 2006 at 06:23:48AM +0800, John Summerfied wrote:
Fedora Project isn't providing as good a service as it could, with just a teensy bit more effort.
I think you underestimate the effort involved with maintaining another branch of packages. Bumping up to the newest version of a package isn't always feasible, and it's a ton of work backporting fixes to older releases. Tracking three streams of development is a nightmare (And this doesn't even count RHEL streams that your average Red Hat developer works on in parallel).
Handing off releases to legacy has been a great way of freeing up developer time so that the stability of the current release increases and the next release gets into a state ready for shipping.
Just the amount of time saved doing bugzilla triage on FC3 is going to pay off huge now that it's reached EOL. I got the kernel bugcount pretty low at EOL, and now I can go about doing the same thing for FC4, which has been suffering at the expense of FC3 for a while.
Dave
Dave Jones wrote:
On Wed, Jan 18, 2006 at 06:23:48AM +0800, John Summerfied wrote:
Fedora Project isn't providing as good a service as it could, with just a teensy bit more effort.
I think you underestimate the effort involved with maintaining another branch of packages. Bumping up to the newest version of a package isn't always feasible, and it's a ton of work backporting fixes to older releases. Tracking three streams of development is a nightmare (And this doesn't even count RHEL streams that your average Red Hat developer works on in parallel).
Handing off releases to legacy has been a great way of freeing up developer time so that the stability of the current release increases and the next release gets into a state ready for shipping.
Just the amount of time saved doing bugzilla triage on FC3 is going to pay off huge now that it's reached EOL. I got the kernel bugcount pretty low at EOL, and now I can go about doing the same thing for FC4, which has been suffering at the expense of FC3 for a while.
Dave
Dave, we're only talking about a short period of support. If Rahul is right (and I don't believe he is), there won't be any fixes.
If he's wrong, I don't imagine that there will be many fixes, so we're talking about only a little bit more support - a short period of time, maybe a few important fixes.
And those fixes are being generated anyway: all that's needed is better timing and better coordination. I really don't see much change to the workload at all, and I can imagine it might even be better: do the handover when the panic of pushing the new release out the door is over and things have calmed down a bit, instead of when the excitement's building and everyone's running round trying to get last-minute changes in, trying to justify rebuilding everything so we can get the latest prerelease gcc or glibc or whatever in.
Hi
Dave, we're only talking about a short period of support. If Rahul is right (and I don't believe he is), there won't be any fixes.
Ya. Continue to believe I am lying all the way to you to make you convince something that isnt really true. Nice way to get anything done. I said there wasnt any major security fixes involved at this point and that is the information that I gather from the #fedora-legacy discussions. If it isn't true, then it isnt. We still have a lot more work on maintaining a release than merely security fixes. Anyway fedora-devel list has a ongoing discussion. Feel free to share your thoughts there.
Rahul Sundaram wrote:
Hi
Dave, we're only talking about a short period of support. If Rahul is right (and I don't believe he is), there won't be any fixes.
Ya. Continue to believe I am lying all the way to you to make you convince something that isnt really true. Nice way to get anything done. I said there wasnt any major security fixes involved at this point and that is the information that I gather from the #fedora-legacy discussions. If it isn't true, then it isnt. We still have a lot more work on maintaining a release than merely security fixes. Anyway fedora-devel list has a ongoing discussion. Feel free to share your thoughts there.
This is getting out of hand. I haven't seen anyone claim you are lying, Rahul.
Could we all please calm down?
Why didn't the "demise" of FC2 cause all this furor? I meekly went along to Legacy, and still run it. Maybe I should have raised a ruckus.
Mike
Mike McCarty wrote:
Rahul Sundaram wrote:
Hi
Dave, we're only talking about a short period of support. If Rahul is right (and I don't believe he is), there won't be any fixes.
Ya. Continue to believe I am lying all the way to you to make you convince something that isnt really true. Nice way to get anything done. I said there wasnt any major security fixes involved at this point and that is the information that I gather from the #fedora-legacy discussions. If it isn't true, then it isnt. We still have a lot more work on maintaining a release than merely security fixes. Anyway fedora-devel list has a ongoing discussion. Feel free to share your thoughts there.
This is getting out of hand. I haven't seen anyone claim you are lying, Rahul.
what is it all about "I dont believe it is" every single mail then?. Explain it to me if you can.
Rahul Sundaram wrote:
Mike McCarty wrote:
Rahul Sundaram wrote:
Hi
Dave, we're only talking about a short period of support. If Rahul is right (and I don't believe he is), there won't be any fixes.
Ya. Continue to believe I am lying all the way to you to make you convince something that isnt really true. Nice way to get anything done. I said there wasnt any major security fixes involved at this point and that is the information that I gather from the #fedora-legacy discussions. If it isn't true, then it isnt. We still have a lot more work on maintaining a release than merely security fixes. Anyway fedora-devel list has a ongoing discussion. Feel free to share your thoughts there.
This is getting out of hand. I haven't seen anyone claim you are lying, Rahul.
what is it all about "I dont believe it is" every single mail then?. Explain it to me if you can.
Well, I don't know what your native language is. You seem pretty competent in English. However that may be, the subject of the sentence "You are lying" is "you", and the subject of "I don't believe that" is "I". One makes a statement about you, and the other makes a statement about the speaker.
So much for grammar.
"I don't believe this is true." makes the claim that Dave does not believe what Rahul has said.
"You are lying." makes the claim that Rahul does not believe what Rahul has said.
That's the semantics of it.
It's a matter of imputing motive. The first statement, while it allows that the person may be lying, does not imply that he is. The second statement imputes bad motives, and deceptiveness.
Mike
On Tue, 2006-01-17 at 00:56, John Summerfied wrote:
Jo and I (and a lot of others) switched from RH back when RH announced its changed arrangements, and the commencement of the Fedora project because neither RH not FC provided what we wanted.
We've dabbled in others including Debian, Ubuntu, SUSE, Mandrake/Mandribble and some others.
I currently run Debian/Woody, Debian/Sarge, Ubuntu, FC3, SUSE 10 and WBEL, so I know what the major alternatives are,
If you want a free long-supported, very-much-FC3-like distribution, look at Centos4.x. The only substantial difference is having mysql 4.x and dropping some kernel modules (which an optional kernel rebuild adds back).
Les Mikesell wrote:
On Tue, 2006-01-17 at 00:56, John Summerfied wrote:
Jo and I (and a lot of others) switched from RH back when RH announced its changed arrangements, and the commencement of the Fedora project because neither RH not FC provided what we wanted.
We've dabbled in others including Debian, Ubuntu, SUSE, Mandrake/Mandribble and some others.
I currently run Debian/Woody, Debian/Sarge, Ubuntu, FC3, SUSE 10 and WBEL, so I know what the major alternatives are,
If you want a free long-supported, very-much-FC3-like distribution, look at Centos4.x. The only substantial difference is having mysql 4.x and dropping some kernel modules (which an optional kernel rebuild adds back).
I am using WBEL on my server, where longevity is important.
I'd not use FC (haven't you seen me describe it as a rolling beta?) on anything _that_ important.
I don't have a particular attachment to FC3, it just happens to be the one that's been retired right now. I didn't use prior FCs and so wasn't affected when they retired.
Retirement plans for other software are based on some period of time _after_ the release of a subsequent (eg second, third) release, not _before_.
The current policy of retiring an FC release immediately _before_ the release of the second subsequent release seems completely bizarre to me.
Hi
Retirement plans for other software are based on some period of time _after_ the release of a subsequent (eg second, third) release, not _before_.
The current policy of retiring an FC release immediately _before_ the release of the second subsequent release seems completely bizarre to me.
You can post to fedora-devel list and ask the developers the reason.
On Tue, 2006-01-17 at 15:34, John Summerfied wrote:
The current policy of retiring an FC release immediately _before_ the release of the second subsequent release seems completely bizarre to me.
It is really in no one's best interest to have developers wasting their time backporting fixes and updates into an old distribution. If you want updates, everyone is better off if you move on to the next version, and a policy that encourages this makes a lot of sense. In theory, FC4 should be very well tested and reliable by now and FC5 should be usable if you don't mind a surprise or two. It is also next to no trouble to update your yum configuration to pull from the legacy repository (much less than complaining about it...). My only concern in this issue is for the large number of machines whose administrators don't watch the mail list and won't be aware of the need to change. When the next security problem and fix come around they won't get the fix even if they continue to run their updates.
Les Mikesell wrote:
On Tue, 2006-01-17 at 15:34, John Summerfied wrote:
The current policy of retiring an FC release immediately _before_ the release of the second subsequent release seems completely bizarre to me.
It is really in no one's best interest to have developers wasting their time backporting fixes and updates into an old distribution.
If Rahul is right there aren't any, so the workload won't be great. If he's wrong (as I expect) there won't be many, and still thr workload's not great.
If you want updates, everyone is better off if you move on to the next version, and a policy that encourages this makes a lot of sense. In theory, FC4 should be very well tested and reliable by now and FC5 should be usable if you don't mind a surprise or two.
I'm content to move to FC5, but it's not even released yet, and that is the point of the argument.
It is also next to no trouble to update your yum configuration to pull from the legacy repository (much less than complaining about it...). My only concern in this issue is for the large
If FP listens, then it can solve the problem for a lot more than one person for a lot more (we hope) than one release.
Any time anyone makes any changes to any file, there is some likelihood of getting it wrong, and if a hundred people make changes to their own copies, then its almost certain that someone will do exactly that.
number of machines whose administrators don't watch the mail list and won't be aware of the need to change. When the next security problem and fix come around they won't get the fix even if they continue to run their updates.
There is that too.
Hi
If Rahul is right there aren't any, so the workload won't be great. If he's wrong (as I expect) there won't be many, and still thr workload's not great.
Please dont misquote me. I never said there is no workload involved in extending a release within Fedora Core by Red Hat. I said there arent major security issues at this point which are two differnt things.
If FP listens, then it can solve the problem for a lot more than one person for a lot more (we hope) than one release.
You are talking in a users list. It will just get lost in the list noise.
Rahul Sundaram wrote:
Hi
If Rahul is right there aren't any, so the workload won't be great. If he's wrong (as I expect) there won't be many, and still thr workload's not great.
Please dont misquote me. I never said there is no workload involved in
You said no security fixes. I did not say no workload.
extending a release within Fedora Core by Red Hat. I said there arent major security issues at this point which are two differnt things.
If FP listens, then it can solve the problem for a lot more than one person for a lot more (we hope) than one release.
You are talking in a users list. It will just get lost in the list noise.
Two things Red Hat and Fedora Project should be doing 1. Surveying the oppsition, just as Microsoft does. MS is running RHEL and some other distros.
2. Listening to users. This list is for users to talk stuff over, so this is a good place to listen to users. I'd think this thread is big enough to be noticed, and the subject should also highlight its importance to RH & FP people.
On Wed, 2006-01-18 at 04:29 +0530, Rahul Sundaram wrote:
If FP listens, then it can solve the problem for a lot more than one person for a lot more (we hope) than one release.
You are talking in a users list.
You are receiving user feedback - One of the most valuable things an enterprise can receive!
It will just get lost in the list noise.
Well, RH could have somebody monitoring this list on a professional basis and communicate the issues reported here to their management.
Ralf
Hi
It will just get lost in the list noise.
Well, RH could have somebody monitoring this list on a professional basis and communicate the issues reported here to their management.
This list is mostly noise at the management level. Here again is a much better idea. One of the guys summaries important happenings here in http://fedoraproject.org/wiki/Projects
On Wed, 2006-01-18 at 09:34 +0530, Rahul Sundaram wrote:
Hi
It will just get lost in the list noise.
Well, RH could have somebody monitoring this list on a professional basis and communicate the issues reported here to their management.
This list is mostly noise at the management level.
Didn't you read what I wrote?
I said: Pay somebody for reading this list and filter the noise into reports to the management.
Such are report would then contain: "There is a strong demand to extend the life time of Fedora releases".
BTW: I do agree on this topic - FC(current+1)t2 is too early to discontinue FC(current-1). FC(current+1) + 1 month would be better.
Ralf
Hi
This list is mostly noise at the management level.
Didn't you read what I wrote?
I said: Pay somebody for reading this list and filter the noise into reports to the management.
That doesnt sound like a good idea to me really. It would be grossly inefficient for anyone to track all the mailing lists users post into and then try filter out anything important at the management level. The signal to noise level ratio is just too much. Users can very well bring in discussions to fedora-devel list if its about development issues. I dont understand the reluctance to do something as simple as signing up another mailing list and suggesting that someone get paid to manage this as a workaround.
On Wed, 2006-01-18 at 09:50 +0530, Rahul Sundaram wrote:
Hi
This list is mostly noise at the management level.
Didn't you read what I wrote?
I said: Pay somebody for reading this list and filter the noise into reports to the management.
That doesnt sound like a good idea to me really. It would be grossly inefficient for anyone to track all the mailing lists users post into and then try filter out anything important at the management level.
Ask yourself: You apparently are a RH employee and already are reading many of the lists. How much effort would it take to you to communicate the essentials to the appropriate channels inside of RH, be it as formal reports, by notes/memos, or by simply forwarding some mails?
It's probably not your formal task nor duty - But it could be and probably would not make much difference to you.
The signal to noise level ratio is just too much. Users can very well bring in discussions to fedora-devel list if its about development issues. I dont understand the reluctance to do something as simple as signing up another mailing list and suggesting that someone get paid to manage this as a workaround.
I can tell you how I see it:
* People expect RH to have established such communication channels internally. * Non-developers hesitate to post to devel lists because they repeatedly have been instructed not to post end-user feedback to devel lists. * RH lists require subscription. This is not feasible for "one time postings". * It's not the developers on devel lists talking the decisions on release cycles. It's RH's management.
Ralf
Hi
Ask yourself: You apparently are a RH employee and already are reading many of the lists. How much effort would it take to you to communicate the essentials to the appropriate channels inside of RH, be it as formal reports, by notes/memos, or by simply forwarding some mails?
A huge amount of effort really. I dont read all the mails by far.
It's probably not your formal task nor duty - But it could be and probably would not make much difference to you.
My job has obsoletely nothing to do with Fedora. It takes hours of my own volunteer time to respond here in a limited way and I rather spend it on doing things that improve Fedora Project rather than merely spending time on discussions that dont change a single thing.
I can tell you how I see it:
- People expect RH to have established such communication channels
internally.
Right.
- Non-developers hesitate to post to devel lists because they repeatedly
have been instructed not to post end-user feedback to devel lists.
When they posted end user queries or send in rants that arent Fedora development specific.
- RH lists require subscription. This is not feasible for "one time
postings".
It is way less than the time people spend arguing about it here where it makes no difference.
- It's not the developers on devel lists talking the decisions on
release cycles. It's RH's management.
Development people can talk to management about it if they are convinced of the strategy. That is the way things work here.
On Tue, 2006-01-17 at 22:58, Rahul Sundaram wrote:
- Non-developers hesitate to post to devel lists because they repeatedly
have been instructed not to post end-user feedback to devel lists.
When they posted end user queries or send in rants that arent Fedora development specific.
Heh... How is a user supposed to know ahead of time which problem description is going to be called a rant by the developers and which is going to result in some development to fix it?
Hi
Heh... How is a user supposed to know ahead of time which problem description is going to be called a rant by the developers and which is going to result in some development to fix it?
No idea. I try. Sometimes I am told to screw myself off. Sometimes I do get a response, positive or otherwise. Sometimes it is this golden silence. I guess the same rules apply to you.
From: "Les Mikesell" lesmikesell@gmail.com
On Tue, 2006-01-17 at 22:58, Rahul Sundaram wrote:
- Non-developers hesitate to post to devel lists because they repeatedly
have been instructed not to post end-user feedback to devel lists.
When they posted end user queries or send in rants that arent Fedora development specific.
Heh... How is a user supposed to know ahead of time which problem description is going to be called a rant by the developers and which is going to result in some development to fix it?
Elliot sent out that monetary support question recently. You are describing another way that the Fedora projects can be materially assisted by individuals. While you cannot get a tax write off for donated hours (you didn't get paid so no income exists to be deducted against) figuring out how to digest the lists into something really simple for the developers to follow might be a good idea. And if you can manage to thread the messages over a period of time you are most of the way to developing the digests. One or two lines of digest text per message thread that looks like something other than idle bantering should be enough to keep the developers aware of the buzz on the list. And if you include the message IDs for the messages, in a clickable format, the developers can read the more interesting and well put messages (if you score them in any way) in this thread or one like it.
It does seem that for most on this list (mea culpa) the proposition is all take and little or no give. That does seem unbalanced, bad karmic practice, and unfair when you stand back far enough and look at the whole picture.
{^_^}
From: "Rahul Sundaram" sundaram@redhat.com
Hi
This list is mostly noise at the management level.
Didn't you read what I wrote?
I said: Pay somebody for reading this list and filter the noise into reports to the management.
That doesnt sound like a good idea to me really. It would be grossly inefficient for anyone to track all the mailing lists users post into and then try filter out anything important at the management level. The signal to noise level ratio is just too much. Users can very well bring
Too poor you mean, I bet. It is impossible to get a signal to noise ratio too good - in most cases.
{^_-}
From: "Ralf Corsepius" rc040203@freenet.de
On Wed, 2006-01-18 at 09:34 +0530, Rahul Sundaram wrote:
Hi
It will just get lost in the list noise.
Well, RH could have somebody monitoring this list on a professional basis and communicate the issues reported here to their management.
This list is mostly noise at the management level.
Didn't you read what I wrote?
I said: Pay somebody for reading this list and filter the noise into reports to the management.
Would you rather waste money on that critter, who'd go insane within two months, or on another bug fighting developer? Seems to me you may have just volunteered yourself to right up weekly summary reports to feed to the developers? {^_-}
Such are report would then contain: "There is a strong demand to extend the life time of Fedora releases".
So appear on the devel list and report that point.
BTW: I do agree on this topic - FC(current+1)t2 is too early to discontinue FC(current-1). FC(current+1) + 1 month would be better.
I can see Rahul's standpoint on the problem. The weeks before release are hectic. However, there is something to be said for making the change after release. Test 2 should be good enough that the amount of changes for release would be small. So the work load may not be so high that this time could be used to bring the Legacy crew onboard for the maintenance effort. That would allow a smooth transition to legacy. This would create smaller "bumps" for people who have critical systems that are working and they do not want to mess with except for extraordinary purposes - heavy security fixes.
Of course, John is worried about the transition from 3 to 4 being a problem followed by the transition from 4 to 5 being a problem. I wonder if the transition from 3 to 5 might not be an even bigger problem than the two stage transition. The level of change would be smaller so he could settle in his adaptations to the 3 to 4 changes over time while 5 matures a little. Then he can switch to five as six emerges. That might be better than waiting until he can make a double jump about the time support ends for 3 on his new proposed schedule. Then if he has transition problems he has to beat them to death with no means of not going forward, if he is unwilling to use legacy support. It almost looks like John has built himself a nicely bricked up chimney with himself inside rather than outside. He has no good way out of his setup given the number of things he's not willing to do.
{^_^}
On Tue, 2006-01-17 at 22:58 -0800, jdow wrote:
From: "Ralf Corsepius" rc040203@freenet.de
On Wed, 2006-01-18 at 09:34 +0530, Rahul Sundaram wrote:
Hi
It will just get lost in the list noise.
Well, RH could have somebody monitoring this list on a professional basis and communicate the issues reported here to their management.
This list is mostly noise at the management level.
Didn't you read what I wrote?
I said: Pay somebody for reading this list and filter the noise into reports to the management.
Would you rather waste money on that critter, who'd go insane within two months, or on another bug fighting developer? Seems to me you may have just volunteered yourself to right up weekly summary reports to feed to the developers? {^_-}
I am not working for RH, and am not on "RH's hidden channels to the decision makers being relevant here" ... but others on this list, probably comprising son having been involved into this threat, probably are. ;)
Such are report would then contain: "There is a strong demand to extend the life time of Fedora releases".
So appear on the devel list and report that point.
You are barking up the wrong tree. I am around there for a long time, but actually I don't think it will make a difference.
All I can do is "to shout into the woods, loudly" and hope somebody will listen. It's essentially the same as you do here and will have the same impact - Except that the bias will be different: User tend to be "demanding", developers tend to be wanting to "concentrate on the interesting things", management tends to "find things too expensive" - And all of them prefer not to change anything ;)
BTW: I do agree on this topic - FC(current+1)t2 is too early to discontinue FC(current-1). FC(current+1) + 1 month would be better.
I can see Rahul's standpoint on the problem. The weeks before release are hectic.
In practice, it doesn't make much of a difference to most maintainers[1], if a distro is formally discontinued or not.
Most maintainers will have to go after bugs/PRs in any case, no matter how old the release/distro is, somebody complained about, and maintainers will have to decide case-by-case if and how to apply changes to a package in any case.
One of such decision can be not to apply/skip a change to packages in older distros or not to update a package there-in. Similar effects can also be caused by a maintainer's time constraints/resources - Most fixes/updates happen when "he has gotten around to it"[2].
Whether a distro is formally discontinued or not, is secondary.
This even more applies to FE than it does to FC: There, maintainers decide themselves on whether to update a package or not. As you might have noticed, many packages in FC3 already don't receive regular updates, "just because the maintainer didn't", for whatever reasons ;)
I.e. unless RH is going to turn of the FC3 buildsystem and stops hosting FC3 extras on their hosts, the impact of discontinuing FC3 on maintaining FE packages is close to zero.
What will have a significant impact, is RH moving devel to FC5 and adding devel for FC4, inside of the build system.
Ralf
[1] Exception: Fundamental packages in Core, such as kernel, glibc, etc. [2] Exception: Security fixes.
On Wed, Jan 18, 2006 at 06:55:08AM +0800, John Summerfied wrote:
Les Mikesell wrote:
On Tue, 2006-01-17 at 15:34, John Summerfied wrote:
The current policy of retiring an FC release immediately _before_ the release of the second subsequent release seems completely bizarre to me.
It is really in no one's best interest to have developers wasting their time backporting fixes and updates into an old distribution.
If Rahul is right there aren't any, so the workload won't be great. If he's wrong (as I expect) there won't be many, and still thr workload's not great.
Even if you're not pushing out updates, the fact that it's still supported means you still need to do bugzilla triage, which takes lots of time.
Reading bugzillamail takes up the bulk of my time, as opposed to actually writing/fixing any code.
Dave
Dave Jones wrote:
On Wed, Jan 18, 2006 at 06:55:08AM +0800, John Summerfied wrote:
Les Mikesell wrote:
On Tue, 2006-01-17 at 15:34, John Summerfied wrote:
The current policy of retiring an FC release immediately _before_ the release of the second subsequent release seems completely bizarre to me.
It is really in no one's best interest to have developers wasting their time backporting fixes and updates into an old distribution.
If Rahul is right there aren't any, so the workload won't be great. If he's wrong (as I expect) there won't be many, and still thr workload's not great.
Even if you're not pushing out updates, the fact that it's still supported means you still need to do bugzilla triage, which takes lots of time.
Someone's still going to be doing that. Can the transition to FL be smoothed over?
My concern is that FC3 (maybe we're actually too late for FC3, but it's a convenient label) be supported until FC5 is released and supported so that users who want to upgrade have a smooth upgrade path directly to FC5.
I actually do not care _how_ its done, except that from the users' perspective it should "just happen."
I'm not entirely convinced that the suggestions to update the yum package are entirely flawless[1], but so far as I can see it will work for most people. The prospects of someone at FP getting the change right are enormously better than all of us users doing so.
Reading bugzillamail takes up the bulk of my time, as opposed to actually writing/fixing any code.
1. Some of us unofficially mirror the updates, but I'm happy to live with the consequences of my actions should I do that:-)
Dave Jones wrote:
On Wed, Jan 18, 2006 at 06:55:08AM +0800, John Summerfied wrote:
Les Mikesell wrote:
On Tue, 2006-01-17 at 15:34, John Summerfied wrote:
The current policy of retiring an FC release immediately _before_ the release of the second subsequent release seems completely bizarre to me.
It is really in no one's best interest to have developers wasting their time backporting fixes and updates into an old distribution.
If Rahul is right there aren't any, so the workload won't be great. If he's wrong (as I expect) there won't be many, and still thr workload's not great.
Even if you're not pushing out updates, the fact that it's still supported means you still need to do bugzilla triage, which takes lots of time.
Reading bugzillamail takes up the bulk of my time, as opposed to actually writing/fixing any code.
Dave
One behavior of Bugzilla that makes bug tracking hard is that bugzilla entries no longer appear on the filer's homepage and are easily lost until the expiring release bug reports were closed or questioned for feedback. I forgot all about the Synaptic related bug until questioned and it did not appear as a category on the filer homepage. Another category related to these types of reports might give developers a less cluttered with "lost and found" bug reports.
The issue that I reported was long ago resolved. I had no idea it was still open. If the report showed on the page, I would have closed the report as resoved.
Jim
Hi
One behavior of Bugzilla that makes bug tracking hard is that bugzilla entries no longer appear on the filer's homepage and are easily lost until the expiring release bug reports were closed or questioned for feedback. I forgot all about the Synaptic related bug until questioned and it did not appear as a category on the filer homepage. Another category related to these types of reports might give developers a less cluttered with "lost and found" bug reports.
The issue that I reported was long ago resolved. I had no idea it was still open. If the report showed on the page, I would have closed the report as resoved.
We have been having some discussions on improving bugzilla. I have a handily list now. I will this feature request to that list. Continue using bugzilla though. It has its quirks but its a mature and reliable workhorse.
Rahul Sundaram wrote:
Hi
One behavior of Bugzilla that makes bug tracking hard is that bugzilla entries no longer appear on the filer's homepage and are easily lost until the expiring release bug reports were closed or questioned for feedback. I forgot all about the Synaptic related bug until questioned and it did not appear as a category on the filer homepage. Another category related to these types of reports might give developers a less cluttered with "lost and found" bug reports.
The issue that I reported was long ago resolved. I had no idea it was still open. If the report showed on the page, I would have closed the report as resoved.
We have been having some discussions on improving bugzilla. I have a handily list now. I will this feature request to that list. Continue using bugzilla though. It has its quirks but its a mature and reliable workhorse.
My current listing of bugs reported are under 10 items. i am sure that there are many floating around that have ong been resolved but I lost track of where they were filed. some of these improvments will clear room for the developers to do acual bug resolving by reducing the "never mind, it is fixed now" bugs thst are part of the clutter. Better tracking of bugs individuals file will help things along a lot.
Bugzilla in general has improves a lot from where it was a short time back. I am sure the trend will continue.
For this thread in general, FC3 is now legacy and cleaning out the bugzilla outstanding bug load is one of the better aspects with it transferring hands for responsibility for keeping it a secure release. Dedication for keeping FC3 updated pretty much in sync with FC4 updates helped to make its transition to legacy a well maintained version.
Jim
On Tue, 2006-01-17 at 16:55, John Summerfied wrote:
I'm content to move to FC5, but it's not even released yet, and that is the point of the argument.
The point is that you are supposed to be testing the test releases, otherwise they don't accomplish anything.
Les Mikesell wrote:
On Tue, 2006-01-17 at 16:55, John Summerfied wrote:
I'm content to move to FC5, but it's not even released yet, and that is the point of the argument.
The point is that you are supposed to be testing the test releases, otherwise they don't accomplish anything.
Oh, I did. Stupid installer installed the wrong kernel and the system wouldn't boot. I mentioned it at the time on fc-test.
ALso, I'm on dialup. I cannot possibly cope with the flood of updates.
Since I actually want to do work on that system, I installed SUSE 10 instead.
John Summerfied wrote:
Les Mikesell wrote:
On Tue, 2006-01-17 at 16:55, John Summerfied wrote:
I'm content to move to FC5, but it's not even released yet, and that is the point of the argument.
The point is that you are supposed to be testing the test releases, otherwise they don't accomplish anything.
Oh, I did. Stupid installer installed the wrong kernel and the system wouldn't boot. I mentioned it at the time on fc-test.
Fixed on rawhide within days.
On 1/17/06, Les Mikesell lesmikesell@gmail.com wrote:
On Tue, 2006-01-17 at 16:55, John Summerfied wrote:
I'm content to move to FC5, but it's not even released yet, and that is the point of the argument.
The point is that you are supposed to be testing the test releases, otherwise they don't accomplish anything.
This is completely disingenuous. If the only "point" is to do nothing other than test "testing" versions, then there is no "point" to having a "release" version. The test releases *do* accomplish something, they are tested by those people interested in *testing* them. That's not everyone (in fact, it's a very small minority). The full release versions are for "everyone else" (that's FC1, FC2, FC3, and FC3, note the lack of -test suffix). If it weren't for the users of release versions, all your development efforts and testing are nothing more than mental masturbation. The "point" is to get a released product into the hands of end users wanting to get work done with it.
Personally, I don't have any issues with using Fedora Legacy, and I frankly don't see why anyone else does, either. The people putting updates together for Fedora Legacy aren't somehow "substandard" compared to those putting updates into "supported" releases of Feodra Core. However, I do see it as entirely appropriate to keep 2 trees supported.
This overlap, in fact, occurs right up until -test2 of a new release is sent out the door. What conditions change at the release of -test2 that makes continuing that support through the release of the final version untenable? I don't think there *are* any conditions that make it untenable other than a drive to move people forward. Well, that drive is statisfied by moving FC3 into Legacy the day FC5 is released, rather than the day FC5-test2 is released.
People are playing theorycraft here with regards to why the Project has this policy. I'd like some actual word from the people that made the decision.
"We do it this way because . . ."
-- Chris
"I trust the Democrats to take away my money, which I can afford. I trust the Republicans to take away my freedom, which I cannot."
On Thu, 2006-01-19 at 03:32, Christofer C. Bell wrote:
I'm content to move to FC5, but it's not even released yet, and that is the point of the argument.
The point is that you are supposed to be testing the test releases, otherwise they don't accomplish anything.
This is completely disingenuous. If the only "point" is to do nothing other than test "testing" versions, then there is no "point" to having a "release" version.
If no one tests the testing version, nothing will change between it and the release version.
The test releases *do* accomplish something, they are tested by those people interested in *testing* them.
Ummm, yeah - I'll bet there are just a whole bunch of people out there whose favorite thing is to try to run software that probably won't work just so they can report bugs.
That's not everyone (in fact, it's a very small minority).
And that's a problem, because bugs don't get found until lots of people run the software under exactly the same conditions as they will in production and on all of the same hardware.
The full release versions are for "everyone else" (that's FC1, FC2, FC3, and FC3, note the lack of -test suffix). If it weren't for the users of release versions, all your development efforts and testing are nothing more than mental masturbation. The "point" is to get a released product into the hands of end users wanting to get work done with it.
That's not the point of the fedora project, although it has to be a side effect that keeps enough people using it. When all the problems that can be fixed in a version have been fixed, the project isn't accomplishing anything until the next release.
On Tue, 2006-01-17 at 16:39 -0600, Les Mikesell wrote:
On Tue, 2006-01-17 at 15:34, John Summerfied wrote:
The current policy of retiring an FC release immediately _before_ the release of the second subsequent release seems completely bizarre to me.
It is really in no one's best interest to have developers wasting their time backporting fixes and updates into an old distribution. If you want updates, everyone is better off if you move on to the next version, and a policy that encourages this makes a lot of sense. In theory, FC4 should be very well tested and reliable by now and FC5 should be usable if you don't mind a surprise or two. It is also next to no trouble to update your yum configuration to pull from the legacy repository (much less than complaining about it...). My only concern in this issue is for the large number of machines whose administrators don't watch the mail list and won't be aware of the need to change. When the next security problem and fix come around they won't get the fix even if they continue to run their updates.
A yum update to yum would solve that. Change the .repo files to point to the legacy project and then any new updates would be available totally transparent to the user.
And, leave that update on the fedora site so anyone, at any time, who tries to update from fedora would get the change.
This won't, of course, take care of those who fail to keep up to date systems. (But then they weren't keeping the updates installed anyway.) It will however, give them the new location the first time they do an actual update attempt.
-- Les Mikesell lesmikesell@gmail.com
On Tue, 2006-17-01 at 14:56 +0800, John Summerfied wrote:
Rahul Sundaram wrote:
Hi
I I don't want to join it for the sake of a month; I wish to go from three to five. I prefer not to rejig my yum configurations and/or up2date configurations, and I don't wish to go to four at this late state.
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
Take your fingers out of you ears; I alread said that's not what I want to do.
"responsibility to contribute."
Jo and I don't contribute? You gotta be joking. We've both been helping out on these lists for years (with a minor break or two when RH left us).
If you would like to get updates from Fedora Legacy project for a longer duration you will have to contribute to that project specifically which is what I meant earlier in context.
We don't want to get updates from FL. We'd like support for the outgoing release (FC3 now, FC4 next time) to remain as it was until the second successor is out and (somewhat) stabilised, so there are always two releases supported by FP.
Jo and I (and a lot of others) switched from RH back when RH announced its changed arrangements, and the commencement of the Fedora project because neither RH not FC provided what we wanted.
We've dabbled in others including Debian, Ubuntu, SUSE, Mandrake/Mandribble and some others.
I currently run Debian/Woody, Debian/Sarge, Ubuntu, FC3, SUSE 10 and WBEL, so I know what the major alternatives are,
We're asking just for a teensy extension to the support offered by FP. Instead of 13 months for FC3, maybe 14 (FC5+1 say, instead of FC5-1).
Hey John, there is no use beating a dead horse. Once a decision is made there is never any chance it will be reversed. I have finally accepted this fact. And if I continue to use FC, I have resigned my self to live with what I am given and just be happy if it works for me, when it doesn't, I figure out how to live with what I can make work myself. It has been made very clear that this is what is expected of FC users. When something breaks, send a bugzilla if there aren't already a bunch for the same problem, then figure out how to live with problem until the package is no longer supported or possibly gets fixed. Remember we are the extended testing ground for the Enterprise products and should be resigned to be treated as such.
By the way FC5t2 seems OK so far, but I will get Ubuntu just in case the final release is as broken as FC4 was, so I have something stable.
Guy Fraser wrote:
On Tue, 2006-17-01 at 14:56 +0800, John Summerfied wrote:
Rahul Sundaram wrote:
Hi
I I don't want to join it for the sake of a month; I wish to go from three to five. I prefer not to rejig my yum configurations and/or up2date configurations, and I don't wish to go to four at this late state.
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
Take your fingers out of you ears; I alread said that's not what I want to do.
"responsibility to contribute."
Jo and I don't contribute? You gotta be joking. We've both been helping out on these lists for years (with a minor break or two when RH left us).
If you would like to get updates from Fedora Legacy project for a longer duration you will have to contribute to that project specifically which is what I meant earlier in context.
We don't want to get updates from FL. We'd like support for the outgoing release (FC3 now, FC4 next time) to remain as it was until the second successor is out and (somewhat) stabilised, so there are always two releases supported by FP.
Jo and I (and a lot of others) switched from RH back when RH announced its changed arrangements, and the commencement of the Fedora project because neither RH not FC provided what we wanted.
We've dabbled in others including Debian, Ubuntu, SUSE, Mandrake/Mandribble and some others.
I currently run Debian/Woody, Debian/Sarge, Ubuntu, FC3, SUSE 10 and WBEL, so I know what the major alternatives are,
We're asking just for a teensy extension to the support offered by FP. Instead of 13 months for FC3, maybe 14 (FC5+1 say, instead of FC5-1).
Hey John, there is no use beating a dead horse. Once a decision is made there is never any chance it will be reversed. I have finally accepted this fact. And if I continue to use FC, I have resigned my self to live with what I am given and just be happy if it works for me, when it doesn't, I figure out how to live with what I can make work myself. It has been made very clear that this is what is expected of FC users. When something breaks, send a bugzilla if there aren't already a bunch for the same problem, then figure out how to live with problem until the package is no longer supported or possibly gets fixed. Remember we are the extended testing ground for the Enterprise products and should be resigned to be treated as such.
By the way FC5t2 seems OK so far, but I will get Ubuntu just in case the final release is as broken as FC4 was, so I have something stable.
You guys sound like the people that use Mandriva. ;-)
Hi
. Remember we are the extended testing ground for the Enterprise products and should be resigned to be treated as such.
Fedora Project includes various sub projects such as core, extras, documentation and legacy among others supported by the Fedora Foundation and sponsored by Red Hat. Fedora Core as one of the sub projects provides a base operating system and packages which shares a similar code base to Red Hat Enterprise Linux. Others such as Fedora Extras are entirely community driven and has no relationship with RHEl. Not all of the packages and technologies included in Fedora Core go into RHEL. Some of the packages are in RHEL but has been removed from Fedora Core - exim for example. Many of them like GFS, SystemTap and Frysk were included in RHEL first before reaching Fedora Core. Also the RHEL updates lifecyle is entirely different from Fedora Core. So Fedora Core cannot act as a testing ground through it does provide a valuable means to gather community feedback.
On Fri, Jan 20, 2006 at 06:06:23AM +0530, Rahul Sundaram wrote:
Hi
. Remember we are the extended testing ground for the Enterprise products and should be resigned to be treated as such.
Fedora Project includes various sub projects such as core, extras, documentation and legacy among others supported by the Fedora Foundation and sponsored by Red Hat. Fedora Core as one of the sub projects provides a base operating system and packages which shares a similar code base to Red Hat Enterprise Linux. Others such as Fedora Extras are entirely community driven and has no relationship with RHEl. Not all of the packages and technologies included in Fedora Core go into RHEL. Some of the packages are in RHEL but has been removed from Fedora Core - exim for example. Many of them like GFS, SystemTap and Frysk were included in RHEL first before reaching Fedora Core. Also the RHEL updates lifecyle is entirely different from Fedora Core. So Fedora Core cannot act as a testing ground through it does provide a valuable means to gather community feedback.
This is total double talk. If Fedora Core is not a testing ground for RHEL features why exactly does it exist? What you are saying makes no sense at all. ------------------------------------------- Aaron Konstam Computer Science Trinity University telephone: (210)-999-7484
HI
This is total double talk. If Fedora Core is not a testing ground for RHEL features why exactly does it exist? What you are saying makes no sense at all.
Mission to provide a community oriented completely Free and open source operating system.
On Sat, 2006-01-21 at 12:18 +0530, Rahul Sundaram wrote:
HI
This is total double talk. If Fedora Core is not a testing ground for RHEL features why exactly does it exist? What you are saying makes no sense at all.
Mission to provide a community oriented completely Free and open source operating system.
OK, you seem to be provoking it:
Please elaborate: * What in Fedora is community oriented? * What would be different in FC if it was "not community oriented"?
The only differences I can sense, are * we would not have FE, but would be using 3rd party repositories instead. * we probably would not be using FC but would be using a "rebuilt RHEL" as substitute for RHL.
Ralf
Ralf Corsepius wrote:
On Sat, 2006-01-21 at 12:18 +0530, Rahul Sundaram wrote:
HI
This is total double talk. If Fedora Core is not a testing ground for RHEL features why exactly does it exist? What you are saying makes no sense at all.
Mission to provide a community oriented completely Free and open source operating system.
OK, you seem to be provoking it:
You get provoked unnecessarily I must say.
Please elaborate:
- What in Fedora is community oriented?
What is not?
- What would be different in FC if it was "not community oriented"?
Red Hat Linux.
The only differences I can sense, are
- we would not have FE, but would be using 3rd party repositories
instead.
You forget the presence of various other projects such as Legacy, Documentation, Websites, Infrastructure etc...
- we probably would not be using FC but would be using a "rebuilt RHEL"
as substitute for RHL.
Very unlikely. RHEL is too boring for a community to participate. It is almost stagnant which is great for enterprises but not for the community. If RHEL would have been suitable for everyone Red Hat wouldnt started Fedora or vice versa. Fedora's rapid release cycle with tons of feature updates, many of them from Red Hat is a key process of enabling the community. The development methodology of "Release early, Release often" combined with donations from Red Hat combined with the many community sub projects within the Fedora Foundation is what provides value for Fedora.
On Sat, 2006-01-21 at 12:51 +0530, Rahul Sundaram wrote:
Ralf Corsepius wrote:
On Sat, 2006-01-21 at 12:18 +0530, Rahul Sundaram wrote:
HI
This is total double talk. If Fedora Core is not a testing ground for RHEL features why exactly does it exist? What you are saying makes no sense at all.
Mission to provide a community oriented completely Free and open source operating system.
OK, you seem to be provoking it:
You get provoked unnecessarily I must say.
Please elaborate:
- What in Fedora is community oriented?
What is not?
I haven't seen much "community oriented by RH" in FC yet.
I sense, RH not wanting to listen to the community and pushing "their interests".
- What would be different in FC if it was "not community oriented"?
Red Hat Linux.
Uuh? RHL would have continued to exist? I.e. RH would be doing the work volunteers are doing in FE now? Great to here that, let's discontinue FE, then - We contributors must be dumber than we think we are ;)
The only differences I can sense, are
- we would not have FE, but would be using 3rd party repositories
instead.
You forget the presence of various other projects such as Legacy, Documentation, Websites, Infrastructure etc...
Are they provided by RH? RH provides some of the servers, yes. But the contents? Most probably written and provided by volunteers from the community to the community. I.e. direction: Community => community.
If RHL had continued to exist, RH would have had to provide some of this contents themselves, some would be found on 3rd party servers, some would not be necessary at all, and some would not exist.
It would not make an essential difference.
- we probably would not be using FC but would be using a "rebuilt RHEL"
as substitute for RHL.
Very unlikely. RHEL is too boring for a community to participate.
Right, if RHL would not exist anymore, people having switched to other distros is more likely to have happened.
It is almost stagnant which is great for enterprises but not for the community. If RHEL would have been suitable for everyone Red Hat wouldnt started Fedora or vice versa. Fedora's rapid release cycle with tons of feature updates, many of them from Red Hat is a key process of enabling the community.
I don't buy that. It's a key process for RH to be able to provide sufficient stability for RHEL.
I am still waiting for RH to act community-oriented, i.e. to let the community participate actively in their decision processes, to let the community actively work on packages in FC, etc., etc.
The development methodology of "Release early, Release often" combined with donations from Red Hat combined with the many community sub projects within the Fedora Foundation is what provides value for Fedora.
The community did not implement the Fedora Foundation. It's RH who are about to implement them for reason, RH still didn't communicate to the community. Reducing costs/spare taxes usually is the reason for founding foundations (I've been working for one for >10 years).
Ralf
Hi
Please elaborate:
- What in Fedora is community oriented?
What is not?
I haven't seen much "community oriented by RH" in FC yet.
I sense, RH not wanting to listen to the community and pushing "their interests".
What specific things didnt Red Hat listen to the community. Not just your personal opinions.
- What would be different in FC if it was "not community oriented"?
Red Hat Linux.
Uuh? RHL would have continued to exist? I.e. RH would be doing the work volunteers are doing in FE now? Great to here that, let's discontinue FE, then - We contributors must be dumber than we think we are ;)
No. If it was not Fedora Project it would have been Red Hat Linux plus third party repositories which dont integrate well. Now its Fedora foundation managing several sub projects, one of them happens to be maintained by Red Hat.
You forget the presence of various other projects such as Legacy, Documentation, Websites, Infrastructure etc...
Are they provided by RH? RH provides some of the servers, yes. But the contents? Most probably written and provided by volunteers from the community to the community. I.e. direction: Community => community.
If community can set direction then that makes it it a community project. Dont you think? If Red Hat provided direction then you will turn around and call it control like you do for Fedora Core project. Dont argue both ways.
It is almost stagnant which is great for enterprises but not for the community. If RHEL would have been suitable for everyone Red Hat wouldnt started Fedora or vice versa. Fedora's rapid release cycle with tons of feature updates, many of them from Red Hat is a key process of enabling the community.
I don't buy that. It's a key process for RH to be able to provide sufficient stability for RHEL.
Of course it is but its also a Free and open source community operating system. Win-win.
I am still waiting for RH to act community-oriented, i.e. to let the community participate actively in their decision processes, to let the community actively work on packages in FC, etc., etc.
You can actively work on packages by providing patches. Which problem werent you able to solve because Red Hat is working on core packages?. Core is getting reduced and might even not exist in future releases anyway.
The development methodology of "Release early, Release often" combined with donations from Red Hat combined with the many community sub projects within the Fedora Foundation is what provides value for Fedora.
The community did not implement the Fedora Foundation.
Community cannot implement a foundation. It requires funding to the tune that the community cannot provide. How many millions do you want to fund?. It requires a legal team and management direction. Foundation is meant for the community though.
It's RH who are about to implement them for reason, RH still didn't communicate to the community.
You didnt see my earlier mail yet I guess. See https://www.redhat.com/archives/fedora-list/2006-January/msg03413.html. What questions are left unanswered there?
Reducing costs/spare taxes usually is the reason for founding foundations (I've been working for one for >10 years).
You might not understand the law in the US to determine tax savings. Wait for the public announcement for details. Let me quote out, the goal of the foundation is to enable the following.
*
By providing a non-profit entity to organize and manage volunteers.
*
By ensuring that the work of these volunteers remains forever free.
*
By providing a fundraising arm for the development and protection of Fedora and related open source projects.
*
By providing an entity for copyright assignment, so that what is free is also defensible in a court of law.
*
By funding patent filings for inventors in the open source community, so that dedicated individuals can help to build a protective patent shield around open source code.
On Sat, 2006-01-21 at 15:25 +0530, Rahul Sundaram wrote:
I don't buy that. It's a key process for RH to be able to provide sufficient stability for RHEL.
Of course it is but its also a Free and open source community operating system. Win-win.
I am still waiting for RH to act community-oriented, i.e. to let the community participate actively in their decision processes, to let the community actively work on packages in FC, etc., etc.
You can actively work on packages by providing patches.
I could do (and did) that upstream, I could do so (and did) with RHL, before Fedora, I could do so with RHEL.
In comparison to RHL, nothing has changed for me, except that I am building and shipping packages through the FE build system, instead of building them myself. Both situations have/had pros and cons, which approximately balance each other.
So, in the end, nothing much has changed "by Fedora having been introduced".
The development methodology of "Release early, Release often" combined with donations from Red Hat combined with the many community sub projects within the Fedora Foundation is what provides value for Fedora.
The community did not implement the Fedora Foundation.
Community cannot implement a foundation.
Well, the question is: Does it need a foundation?
Reducing costs/spare taxes usually is the reason for founding foundations (I've been working for one for >10 years).
You might not understand the law in the US to determine tax savings.
Right, I am not familiar with US tax law.
In Germany, donations to a "non-profit foundation" are tax deducable up to a certain amount, and are commonly used by enterprises to reduce their taxable income.
Working principle: Enterprises found a foundation, thereby reduce their taxable income and delegate employees/resources to the foundation. Somewhat oversimplified, the only thing a foundation must prove is it not to work "ROI-oriented". Typically this is basic research and long term R'n'D. Fedora probably would comply to the tax laws' criterions.
Ralf
Hi
Working principle: Enterprises found a foundation, thereby reduce their taxable income and delegate employees/resources to the foundation. Somewhat oversimplified, the only thing a foundation must prove is it not to work "ROI-oriented". Typically this is basic research and long term R'n'D. Fedora probably would comply to the tax laws' criterions.
You really need to understand US laws to go through the Foundation procedures. I am sure you can have some of the US buddies help you with that.
Hi
I could do (and did) that upstream, I could do so (and did) with RHL, before Fedora, I could do so with RHEL.
In comparison to RHL, nothing has changed for me, except that I am building and shipping packages through the FE build system, instead of building them myself. Both situations have/had pros and cons, which approximately balance each other.
So, in the end, nothing much has changed "by Fedora having been introduced".
Fedora Extras enabled by default makes a huge difference. Anaconda will support it in the subsequent releases which again make it very relevant.
Well, the question is: Does it need a foundation?
It is being created because the community wanted assurance that their work will remain open source, defensible and to serve other stated goals.
http://fedoraproject.org/wiki/Foundation
On Sun, 2006-01-22 at 17:17 +0530, Rahul Sundaram wrote:
Hi
I could do (and did) that upstream, I could do so (and did) with RHL, before Fedora, I could do so with RHEL.
In comparison to RHL, nothing has changed for me, except that I am building and shipping packages through the FE build system, instead of building them myself. Both situations have/had pros and cons, which approximately balance each other.
So, in the end, nothing much has changed "by Fedora having been introduced".
Fedora Extras enabled by default makes a huge difference.
As I've said many times before, if using pure rpm based installation/updates it doesn't make any difference.
Well, the question is: Does it need a foundation?
It is being created
... by RH, not by community ...
because the community wanted assurance that their work will remain open source, defensible and to serve other stated goals.
It's the work-around RH has decided by themselves, because RH found this would be an applicable work-around. The community, so far, has not, or at least, only be negligibly been involved, and not even been informed until last week.
Ralf
Hi
because the community wanted assurance that their work will remain open source, defensible and to serve other stated goals.
It's the work-around RH has decided by themselves, because RH found this would be an applicable work-around.
Foundation work is done for the community with community input. We have a entire list of very prominent community people involved in the effort. It was basically done to assign all copyrights etc to the foundation and make sure what is community contributions stays open source and free forever. We are also helping to fund so many other things. Calling it a "workaround" is a insult to the people involved with the work and you are being rude as always for no good reason.
The community, so far, has not, or at least, only be negligibly been involved, and not even been informed until last week.
Wrong again as usual. How about you stop being misleading? Let me point you out how many announcements and messages related to the foundation has been made so far over a period of more than 6 months.
http://www.redhat.com/archives/fedora-announce-list/2005-June/msg00008.html http://www.redhat.com/magazine/008jun05/departments/fedora_status/#fedora-fo... http://www.redhat.com/magazine/010aug05/features/fedora_foundation/ http://www.redhat.com/en_us/USA/home/company/news/prarchive/2005/press_fedor... http://www.redhat.com/magazine/012oct05/departments/fedora_status/#fedora-fo...
Hi
You forget the presence of various other projects such as Legacy, Documentation, Websites, Infrastructure etc...
Are they provided by RH? RH provides some of the servers, yes. But the contents? Most probably written and provided by volunteers from the community to the community. I.e. direction: Community => community.
Oh btw I have reviewed documentation, contribution to release notes, contribute to legacy (legacy is chaired by Fedora Core release maintainer) , contributed to websites and infrastructure team both of which has other Red Hat people working on it. Fedora Documentation Steering Committee is chaired by Red Hat. Red Hat has been team members of all these projects. So try and understand the processes and people involved better. Its all documented in the wiki @ http://fedoraproject.org/wiki/Projects
Rahul Sundaram wrote:
Hi
You forget the presence of various other projects such as Legacy, Documentation, Websites, Infrastructure etc...
Are they provided by RH? RH provides some of the servers, yes. But the contents? Most probably written and provided by volunteers from the community to the community. I.e. direction: Community => community.
Oh btw I have reviewed documentation, contribution to release notes, contribute to legacy (legacy is chaired by Fedora Core release maintainer) , contributed to websites and infrastructure team both of which has other Red Hat people working on it. Fedora Documentation Steering Committee is chaired by Red Hat. Red Hat has been team members of all these projects. So try and understand the processes and people involved better. Its all documented in the wiki @ http://fedoraproject.org/wiki/Projects
And Red Hat has contributed tons to Fedora Ambassdors project which I am others from Red Hat are members. Again similar for Bug triaging and artwork... Want more?
On Sat, 2006-01-21 at 15:41 +0530, Rahul Sundaram wrote:
Rahul Sundaram wrote:
And Red Hat has contributed tons to Fedora Ambassdors project which I am others from Red Hat are members. Again similar for Bug triaging and artwork... Want more?
I give in. You are trying to sell us marketing campaigns (RH user clubs) as new achievements.
Ralf
Ralf Corsepius wrote:
On Sat, 2006-01-21 at 15:41 +0530, Rahul Sundaram wrote:
Rahul Sundaram wrote:
And Red Hat has contributed tons to Fedora Ambassdors project which I am others from Red Hat are members. Again similar for Bug triaging and artwork... Want more?
I give in. You are trying to sell us marketing campaigns (RH user clubs) as new achievements.
Ralf
You are calling all the steering committee's with a partnership between community members and Red Hat people as Red Hat user clubs?. Thats really underestimating the work involved here.
On Sun, 2006-01-22 at 17:20 +0530, Rahul Sundaram wrote:
Ralf Corsepius wrote:
On Sat, 2006-01-21 at 15:41 +0530, Rahul Sundaram wrote:
Rahul Sundaram wrote:
And Red Hat has contributed tons to Fedora Ambassdors project which I am others from Red Hat are members. Again similar for Bug triaging and artwork... Want more?
I give in. You are trying to sell us marketing campaigns (RH user clubs) as new achievements.
Ralf
You are calling all the steering committee's with a partnership between community members and Red Hat people as Red Hat user clubs?.
Yes, because that's what they more or less are: Badges to attach to your chest - I fail to see the difference between an "Jack Daniels' apprentice tour guide certificate", "Disney club membership" and a "Fedora Ambassador" and the like.
Also, people can launch as many sites and write as many web-pages, as you want. It's something you can do with or without RH's or the FF - It had been done before Fedora and is still being done.
In fact, RH, FESCO and the FF are dictating the "rules of the game as they want them to be" and it's up to users and contributors to decide how far to respect these rules and to apply/use the resources these organizations provide.
Ralf
Hi
In fact, RH, FESCO and the FF are dictating the "rules of the game as they want them to be" and it's up to users and contributors to decide how far to respect these rules and to apply/use the resources these organizations provide.
Ralf
So you choose to not trust the foundation, all the community members at FESCO or Red Hat itself. I would suggest contributing to a project that you trust instead. Good luck.
Hi ,
I have problem in installing FC 4 , Basically my system was windows XP ,
I added a new hard disk , but unfortunately newly added disk is not shown in while installing linux .
But how ever , system bios is able to detect the newly added hard disk .
Can any one help me , how to proceed.
Regards Gowri
On Mon, 2006-01-23 at 19:15 +0900, Gowri wrote:
Hi ,
I have problem in installing FC 4 , Basically my system was windows XP ,
I added a new hard disk , but unfortunately newly added disk is not shown in while installing linux .
But how ever , system bios is able to detect the newly added hard disk .
Can any one help me , how to proceed.
What type of hard drive (SATA2) did you add and what type chip set (nForce4?) do you have?
I have a new Biostar GEForce 6100 which has a very new chip set that the installer doesn't recognize and had to install to an IDE drive instead of SATA2 but after the updates it saw everything. FC5 T1 and T2 do recognize all hardware for me though.
On Mon, 2006-01-23 at 14:47 +0530, Rahul Sundaram wrote:
Hi
In fact, RH, FESCO and the FF are dictating the "rules of the game as they want them to be" and it's up to users and contributors to decide how far to respect these rules and to apply/use the resources these organizations provide.
So you choose to not trust the foundation, all the community members at FESCO or Red Hat itself. I would suggest contributing to a project that you trust instead. Good luck.
Seems to me as if you haven't understand a glimpse of what I am saying.
Therefore let me try to elaborate: Like all other OSS projects, Fedora is based on users', developers' and sponsoring/financing organs' opinions and objectives are sufficiently intersecting. These parties own genuine interests are irrelevant to the other parties.
E.g. to users and developers it's irrelevant what RH, the FF or FESCO wants. The only point that matters to them is that the project's products (eg. distribution, buildsys, etc.) sufficiently caters/serves their needs. I.e. whatever RH, FF or FESCO do, and which persons or organizations are involved, is their business. May they establish as many marketing events as they like, may they be a foundation or an enterprize, or ...
Neither do they control users/developers nor does the user/developer community control them, nor do they control my mind - It's users and developers who are using YOU, and it's up to YOU to make Fedora attractive to them.
Steps into this directions would be to improve your communication, to listen to the community and ... WHAT you actually do is less important, the questions are "WHY" and "HOW".
Ralf
Ralf Corsepius wrote:
On Mon, 2006-01-23 at 14:47 +0530, Rahul Sundaram wrote:
Hi
In fact, RH, FESCO and the FF are dictating the "rules of the game as they want them to be" and it's up to users and contributors to decide how far to respect these rules and to apply/use the resources these organizations provide.
So you choose to not trust the foundation, all the community members at FESCO or Red Hat itself. I would suggest contributing to a project that you trust instead. Good luck.
Seems to me as if you haven't understand a glimpse of what I am saying.
I understand it fine and dandy but unless you are really happy to contribute and make some differences instead of just about being rude and abusing everyone who is trying to work on the team - both volunteers and full time staff then find a different project that makes sense for you to contribute instead of continuously harping about one issue after another without any constructive criticisms or ideas. I requested repeatedly in our earlier conversations that you post a good list of QA issues that you alleged are present in the Fedora Extras process. Instead of doing that you indulged in name calling everyone doing the work in Fedora Extras. This is not the means to work on anything at all in a positive manner.
Therefore let me try to elaborate: Like all other OSS projects, Fedora is based on users', developers' and sponsoring/financing organs' opinions and objectives are sufficiently intersecting. These parties own genuine interests are irrelevant to the other parties.
E.g. to users and developers it's irrelevant what RH, the FF or FESCO wants.
Completely wrong. The FESCO or FF and RH represent interests with the community in mind. If you dont believe it why are you working on the project at all?
Neither do they control users/developers nor does the user/developer community control them, nor do they control my mind - It's users and developers who are using YOU, and it's up to YOU to make Fedora attractive to them.
Let me make this clear again. If you dont want feel like all my explanations before doesnt meet your requirements find a different project.
Steps into this directions would be to improve your communication, to listen to the community and ... WHAT you actually do is less important, the questions are "WHY" and "HOW".
I am doing the best I can. Stop using rhetoric and rudeness. It wont help at all.
On Mon, 2006-01-23 at 20:45 +0530, Rahul Sundaram wrote:
Ralf Corsepius wrote:
On Mon, 2006-01-23 at 14:47 +0530, Rahul Sundaram wrote:
Seems to me as if you haven't understand a glimpse of what I am saying.
I understand it fine and dandy but unless you are really happy to contribute and make some differences instead of just about being rude and abusing everyone who is trying to work on the team - both volunteers and full time staff then find a different project that makes sense for you to contribute instead of continuously harping about one issue after another without any constructive criticisms or ideas. I requested repeatedly in our earlier conversations that you post a good list of QA issues that you alleged are present in the Fedora Extras process.
Great to experience somebody at RH exposing PM on the list ;)
Instead of doing that you indulged in name calling everyone doing the work in Fedora Extras. This is not the means to work on anything at all in a positive manner.
Therefore let me try to elaborate: Like all other OSS projects, Fedora is based on users', developers' and sponsoring/financing organs' opinions and objectives are sufficiently intersecting. These parties own genuine interests are irrelevant to the other parties.
E.g. to users and developers it's irrelevant what RH, the FF or FESCO wants.
Completely wrong. The FESCO or FF and RH represent interests with the community in mind. If you dont believe it why are you working on the project at all?
That's the core of us talking pass each other:
I am participating in FE, because MY personal interests intersect sufficiently with those of the Fedora Project and expect mutual benefits for all parties being involved. I don't really care about RH, the FF or FESCO.
The problem is RH, the FF and FESCO are increasingly interfering with my interests, and don't really help. Instead of concentrating on technical issues, I see more and more marketing projects being launched, FESCO drawing more and more questionable decisions and RH becoming more and more non-communicational.
I acknowledge RH providing servers and the buildsystem, but apart that, I don't see any substantial contribution of RH to Fedora, which is substantially different from what the situation used to be before Fedora.
Ralf
Hi
I understand it fine and dandy but unless you are really happy to contribute and make some differences instead of just about being rude and abusing everyone who is trying to work on the team - both volunteers and full time staff then find a different project that makes sense for you to contribute instead of continuously harping about one issue after another without any constructive criticisms or ideas. I requested repeatedly in our earlier conversations that you post a good list of QA issues that you alleged are present in the Fedora Extras process.
Great to experience somebody at RH exposing PM on the list ;)
I started the discussed in public about the QA issues and I can very well mention my idea and what I said again in public. Unless I exposure your private conversations in your name to a public list you have no rooms for complaints about privacy.
The problem is RH, the FF and FESCO are increasingly interfering with my interests, and don't really help. Instead of concentrating on technical issues, I see more and more marketing projects being launched, FESCO drawing more and more questionable decisions and RH becoming more and more non-communicational.
Be specific and post to fedora-devel list or fedora-extras list. This is NOT the place where it would make ANY difference at all.
I acknowledge RH providing servers and the buildsystem, but apart that, I don't see any substantial contribution of RH to Fedora, which is substantially different from what the situation used to be before Fedora.
You have repeatedly and steadily ignored all the contributions then. Fedora Core as a sub project within the Fedora Foundation has tons of engineering contributions from RH. FESCO (extras committee) and FDSCO(docs) , FAMSO(ambassadors) have RH members working along with community members. FESCO is now chaired by a non-RH member to enable the community to steer better the direction of extras repository . Fedora Legacy chair is the Fedora Core release engineer working for RH. RH has members working on websites, infrastructure, RH has funded and continues to donate very heavily in money and in kind resources to the project. Let me just take a look at my work as an example of this . I work on core, docs, i18n, legacy, ambassadors, websites. my regional efforts and so on under the Fedora Project banner. Unless you want to discuss any alleged issues in specific details publicly within the appropriate mailing lists there is zero difference this discussion will make. Continuing to post where it makes zero difference is waste of your effort and mine - both volunteering. Do understand that.
On Mon, 2006-01-23 at 21:56 +0530, Rahul Sundaram wrote:
Hi
You have repeatedly and steadily ignored all the contributions then. Fedora Core as a sub project within the Fedora Foundation has tons of engineering contributions from RH. FESCO (extras committee) and FDSCO(docs) , FAMSO(ambassadors) have RH members working along with community members.
You are missing the fact, that I am a technical person and am only interested in technical contributions, technical cooperations and technical coordination.
All the other things you mention might have some benefits to share for some users and contributors - To me they are highly non-interesting and irrelevant [1].
Ralf
[1] I have a "Jack Daniels apprentice tour guide certificate" somewhere. This fact probably is as interesting to you as Fedora Ambassadors are to me.
Ralf Corsepius wrote:
On Mon, 2006-01-23 at 21:56 +0530, Rahul Sundaram wrote:
Hi
You have repeatedly and steadily ignored all the contributions then. Fedora Core as a sub project within the Fedora Foundation has tons of engineering contributions from RH. FESCO (extras committee) and FDSCO(docs) , FAMSO(ambassadors) have RH members working along with community members.
You are missing the fact, that I am a technical person and am only interested in technical contributions, technical cooperations and technical coordination.
Do it in fedora-extras list.
All the other things you mention might have some benefits to share for some users and contributors - To me they are highly non-interesting and irrelevant
Your opinion.
On Mon, 2006-01-23 at 12:01 +0100, Ralf Corsepius wrote:
On Mon, 2006-01-23 at 14:47 +0530, Rahul Sundaram wrote:
Hi
In fact, RH, FESCO and the FF are dictating the "rules of the game as they want them to be" and it's up to users and contributors to decide how far to respect these rules and to apply/use the resources these organizations provide.
So you choose to not trust the foundation, all the community members at FESCO or Red Hat itself. I would suggest contributing to a project that you trust instead. Good luck.
Seems to me as if you haven't understand a glimpse of what I am saying.
Therefore let me try to elaborate: Like all other OSS projects, Fedora is based on users', developers' and sponsoring/financing organs' opinions and objectives are sufficiently intersecting. These parties own genuine interests are irrelevant to the other parties.
Your point is exactly what Rahul has been saying. If you want to make a difference, step up to the plate and assist the project.
If you are not willing to do that, then you are relegating yourself to the field of those users who are only interested in what is already available in a format someone else has created.
Jeff Vian wrote:
Your point is exactly what Rahul has been saying. If you want to make a difference, step up to the plate and assist the project.
If you are not willing to do that, then you are relegating yourself to the field of those users who are only interested in what is already available in a format someone else has created.
"Relegate" is in this context a pejorative term.
Mike
On Thu, 2006-01-26 at 16:59 -0600, Jeff Vian wrote:
On Mon, 2006-01-23 at 12:01 +0100, Ralf Corsepius wrote:
On Mon, 2006-01-23 at 14:47 +0530, Rahul Sundaram wrote:
Hi
In fact, RH, FESCO and the FF are dictating the "rules of the game as they want them to be" and it's up to users and contributors to decide how far to respect these rules and to apply/use the resources these organizations provide.
So you choose to not trust the foundation, all the community members at FESCO or Red Hat itself. I would suggest contributing to a project that you trust instead. Good luck.
Seems to me as if you haven't understand a glimpse of what I am saying.
Therefore let me try to elaborate: Like all other OSS projects, Fedora is based on users', developers' and sponsoring/financing organs' opinions and objectives are sufficiently intersecting. These parties own genuine interests are irrelevant to the other parties.
Your point is exactly what Rahul has been saying.
Not quite: RH is dictating and seem to expect all others to swallow.
What I say: OSS projects are give and take. If what others give (here: RH) is non-interesting from an individual's perspective (here: me, rsp. those who want FC3 to be continued), the whole project becomes non-interesting to an individual.
I don't deny Rahul is contributing something.
The question to me are: How much about RH's work is "community oriented" and how much about RH's work wouldn't have be done if Fedora wasn't "community oriented", and how much about Fedora is a marketing hype?
If you want to make a difference, step up to the plate and assist the project.
In case you don't know: I maintain ca. 50 packages in FE and reviewed many others.
If you are not willing to do that, then you are relegating yourself to the field of those users who are only interested in what is already available in a format someone else has created.
C.f. above - I am "demanding", because "I gave".
Ralf
Ralf Corsepius wrote:
Not quite: RH is dictating and seem to expect all others to swallow.
There are some differences from this sentence which can fairly be applied to MSFT and applying it to RHAT though. RHAT gave out all the pieces and sources needed to pick up package maintainence.
What I say: OSS projects are give and take. If what others give (here: RH) is non-interesting from an individual's perspective (here: me, rsp. those who want FC3 to be continued), the whole project becomes non-interesting to an individual.
Is that actually true that Fedora's value to you is contingent on RHAT specifically continuing to provide security coverage? I guess it comes down to security coverage FC4 is the 'later' version of FC3 apps already (KDE 3.5, etc). For servers that make me feel that bad about dist-updating them I use CentOS.
The question to me are: How much about RH's work is "community oriented" and how much about RH's work wouldn't have be done if Fedora wasn't "community oriented", and how much about Fedora is a marketing hype?
In case you don't know: I maintain ca. 50 packages in FE and reviewed many others.
C.f. above - I am "demanding", because "I gave".
Hum, what is your concept for RHAT's relationship to FC3 then, in terms of what should be? Just asking to understand.
RHAT seem to be saying that Legacy can come up to the same standard as internal security coverage, and if it's true then is there still a problem?
-Andy
On Sat, 2006-01-21 at 15:35 +0530, Rahul Sundaram wrote:
Hi
You forget the presence of various other projects such as Legacy, Documentation, Websites, Infrastructure etc...
Are they provided by RH? RH provides some of the servers, yes. But the contents? Most probably written and provided by volunteers from the community to the community. I.e. direction: Community => community.
Oh btw I have reviewed documentation, contribution to release notes, contribute to legacy (legacy is chaired by Fedora Core release maintainer) , contributed to websites and infrastructure team both of which has other Red Hat people working on it. Fedora Documentation Steering Committee is chaired by Red Hat. Red Hat has been team members of all these projects.
Would have RH not have written these docs if Fedora didn't exist?
To me this is all "relabeling" work you could not have avoided to write otherwise.
Ralf
Hi
Would have RH not have written these docs if Fedora didn't exist?
To me this is all "relabeling" work you could not have avoided to write otherwise.
Community has done pretty good work. Only when the community does the work the complain is that Red Hat didnt. If Red Hat did the work the complaint is that the community was not allowed to participate. Again dont argue both ways
Hi
Uuh? RHL would have continued to exist? I.e. RH would be doing the work volunteers are doing in FE now? Great to here that, let's discontinue FE, then - We contributors must be dumber than we think we are ;)
So you do agree that Fedora Extras as part of the Fedora project is community oriented. Good
On Sat, 2006-01-21 at 01:21, Rahul Sundaram wrote:
Very unlikely. RHEL is too boring for a community to participate. It is almost stagnant which is great for enterprises but not for the community. If RHEL would have been suitable for everyone Red Hat wouldnt started Fedora or vice versa.
Your argument would be more convincing if you could point out some directions that FC has taken that are more in the community interest than in building the next RHEL. Can you?
RHEL4 looks very much like a direct cutover from what was built in FC3 with Mysql4.x dropped in, and so far everything in subsequent FC releases looks like more of the same. If this is really supposed to be a community project, shouldn't it take some directions that aren't aimed at building the next RHEL version?
-- Les Mikesell lesmikesell@gmail.com
On Saturday 21 January 2006 12:04, Les Mikesell wrote:
On Sat, 2006-01-21 at 01:21, Rahul Sundaram wrote:
Very unlikely. RHEL is too boring for a community to participate. It is almost stagnant which is great for enterprises but not for the community. If RHEL would have been suitable for everyone Red Hat wouldnt started Fedora or vice versa.
Your argument would be more convincing if you could point out some directions that FC has taken that are more in the community interest than in building the next RHEL. Can you?
RHEL4 looks very much like a direct cutover from what was built in FC3 with Mysql4.x dropped in, and so far everything in subsequent FC releases looks like more of the same. If this is really supposed to be a community project, shouldn't it take some directions that aren't aimed at building the next RHEL version?
This is a sentiment I have to echo also. The kde/gnome wars being a case in point. kde efforts are only on a time permitted basis it still seems, often done by soemone on his or her own time even after a sufficiently large numnber of users choose to run it. That doesn't equate to being a 'community project' IMNSHO.
IMO the project should demo that its responsive to the users wishes in matters such as this. To show that, we'd need to see support for the eye candy of our choice by rpms instead of being forced into a konstruct build, and a release install that actually worked, unlike the "official" FC4 debacle. I'd imagine you would have many more running FC4 if the users could bring FC4.2 install glitches here, but we've been told pointedly we're on our own, so those cd's are still in my carry case, unused. That wasn't at all conducive to good community relationships. It made it very plain that we were, at the end of the day, still subject to the whims of TPTB at Red Hat.
If and when that time arrives, then this old fart on SS would consider an annual donation to a 'foundation'. A pittance in the grand scheme, but it would show my thanks nonetheless. In the meantime, I serve mainly as a canary in the coal mine, squawking about things that make me itch. And considering a jump to debian if FC5 final won't install, and this is after 8 years of running nothing but RH. And I don't have a 'sacrificial box' other than the one that runs my small milling machine useing emc2. As I'm building emc2 from cvs about weekly, a re-install and a quick cvs grab and I'm pretty much back to a good footing on square one.
OTOH, as long as we are the guinea pigs for the next release of RHEL, then IMO this paper only for tax purposes 'foundation' is a wasted effort.
On Sat, 2006-01-21 at 12:15, Gene Heskett wrote:
IMO the project should demo that its responsive to the users wishes in matters such as this. To show that, we'd need to see support for the eye candy of our choice by rpms instead of being forced into a konstruct build, and a release install that actually worked, unlike the "official" FC4 debacle. I'd imagine you would have many more running FC4 if the users could bring FC4.2 install glitches here, but we've been told pointedly we're on our own, so those cd's are still in my carry case, unused.
Very good example. What kind of community is it that can't be bothered to respin the official isos once in a while so the known-broken install on certain hardware gets fixed and the user that manages to install isn't immediately faced with a gig of update downloads just to get started?
In the meantime, I serve mainly as a canary in the coal mine, squawking about things that make me itch. And considering a jump to debian if FC5 final won't install, and this is after 8 years of running nothing but RH. And I don't have a 'sacrificial box' other than the one that runs my small milling machine useing emc2.
The free vmplayer program from www.vmware.com makes a great test platform since it is non-destructive to the host install. It would really help people evaluate FC5 if someone maintained an installed image with current updates that could be downloaded as you can with ubuntu, suse and many others at: http://www.vmware.com/vmtn/vm/community.html
Les Mikesell wrote:
On Sat, 2006-01-21 at 12:15, Gene Heskett wrote:
IMO the project should demo that its responsive to the users wishes in matters such as this. To show that, we'd need to see support for the eye candy of our choice by rpms instead of being forced into a konstruct build, and a release install that actually worked, unlike the "official" FC4 debacle. I'd imagine you would have many more running FC4 if the users could bring FC4.2 install glitches here, but we've been told pointedly we're on our own, so those cd's are still in my carry case, unused.
Very good example. What kind of community is it that can't be bothered to respin the official isos once in a while so the known-broken install on certain hardware gets fixed and the user that manages to install isn't immediately faced with a gig of update downloads just to get started?
Nobody stopped anyone from doing the work involved. In fact it was already done before. http://fedora.isphuset.no/
On 1/22/06, Rahul Sundaram sundaram@redhat.com wrote:
Nobody stopped anyone from doing the work involved. In fact it was already done before. http://fedora.isphuset.no/
I think the underlying frustration involving the re-spins is one of access and branding. You want it to be accessible for people having problems with the regular installer, but you don't want everyone assuming its an 'official' iso built by the normal release process.
There is a lot more discussion that needs to happen to see how respins, which the core maintainers aren't responsible for, should be branded and should be talked about in easily accessible documentation. Discussion I was hoping the Foundation could take up at some point. Can a respin project get their own bugzilla component in the same bugzilla?
I keep forgetting to ping the http://fedora.isphuset.no/ guys to make sure they have written greg concerning trademark permission for the respins. It's an important usage case, which the trademark policy should try to explicitly address.
-jef
Jeff Spaleta wrote:
On 1/22/06, Rahul Sundaram sundaram@redhat.com wrote:
Nobody stopped anyone from doing the work involved. In fact it was already done before. http://fedora.isphuset.no/
I think the underlying frustration involving the re-spins is one of access and branding. You want it to be accessible for people having problems with the regular installer, but you don't want everyone assuming its an 'official' iso built by the normal release process.
There is a lot more discussion that needs to happen to see how respins, which the core maintainers aren't responsible for, should be branded and should be talked about in easily accessible documentation. Discussion I was hoping the Foundation could take up at some point. Can a respin project get their own bugzilla component in the same bugzilla?
We need to discuss that if only the people who do the work would be responsive to requests repeatedly.
I keep forgetting to ping the http://fedora.isphuset.no/ guys to make sure they have written greg concerning trademark permission for the respins. It's an important usage case, which the trademark policy should try to explicitly address.
-jef
I wrote to them on several occasions with no response whatsoever. Its really disappointing.
On Sunday 22 January 2006 06:58, Rahul Sundaram wrote:
Les Mikesell wrote:
On Sat, 2006-01-21 at 12:15, Gene Heskett wrote:
IMO the project should demo that its responsive to the users wishes in matters such as this. To show that, we'd need to see support for the eye candy of our choice by rpms instead of being forced into a konstruct build, and a release install that actually worked, unlike the "official" FC4 debacle. I'd imagine you would have many more running FC4 if the users could bring FC4.2 install glitches here, but we've been told pointedly we're on our own, so those cd's are still in my carry case, unused.
Very good example. What kind of community is it that can't be bothered to respin the official isos once in a while so the known-broken install on certain hardware gets fixed and the user that manages to install isn't immediately faced with a gig of update downloads just to get started?
Nobody stopped anyone from doing the work involved. In fact it was already done before. http://fedora.isphuset.no/
We ARE well aware of that. As stated above, I have the cd's already burnt. But when asking a question and disclosing that the question involves the unofficial 4.2 release, then the question was summarily ignored, as being off topic I guess. If you are not willing to even discuss something someone has done in the interests of widening the usage, then the whole thing is moot, and we are wasting our collective time with this endeavor called a foundation.
You should have welcomed that effort with open arms, and actively pointed folks having trouble with _your_ release to that one as a possible solution to the install debacle 4 was. Instead, this message mentioning it and supplying the link, and the announcement of its availability, are the only 2 messages I recall going by on this list since 4.2 was released. There may have been a few others, but they didn't make a lasting impression if there were others. A quick search shows 93 hits, out of 23201 messages in this folder. Thats pretty slim pickins considering the message volume.
I'm of the opinion that fedora folks themselves should have done the respin as you call it, within 3-4 days time when it became obvious that 4 was a wholesale x crasher that trashed the disks as it went away. The take it or leave it attitude has the distinct odor of hydrogen sulfide about it.
To me thats ignoreing any 'community effort', and quite frankly I'm less than impressed with this newfound selling of the word community, when its been rather pointedly ignored until now.
If you want it to be a community effort, then its time a hell of a lot more credit was given to those that have contributed.
And that doesn't by any means include me. I'm not looking for any flowers because I haven't contributed squat in terms of code. It (the hardware AND the software) has gotten far more complex than than it was back in 1978 when I built the hardware and wrote an app for the rca 1802 cpu that was still running at that tv station over a decade later. Or a 2nd app running on a trs-80 coco2, started in 1989, that was in daily use till 2002. They could be ported of course, but the hardware no longer exists in the broadcasting business so its moot. That was then, this is now and I haven't done anything but test new Linus kernels for the last 5 years. Anybody can do that.
If there is to be a foundation that we need to support else fedora is going away, then lets see a by-laws draft that will give _us_, the users, a voice.
Hi
Nobody stopped anyone from doing the work involved. In fact it was already done before. http://fedora.isphuset.no/
We ARE well aware of that. As stated above, I have the cd's already burnt. But when asking a question and disclosing that the question involves the unofficial 4.2 release, then the question was summarily ignored, as being off topic I guess. If you are not willing to even discuss something someone has done in the interests of widening the usage, then the whole thing is moot, and we are wasting our collective time with this endeavor called a foundation.
You should have welcomed that effort with open arms, and actively pointed folks having trouble with _your_ release to that one as a possible solution to the install debacle 4 was.
I repeatedly wrote the the people who did it with no responses. We needed to work on several details before adopting such efforts in a formal way.
I'm of the opinion that fedora folks themselves should have done the respin as you call it, within 3-4 days time when it became obvious that 4 was a wholesale x crasher that trashed the disks as it went away. The take it or leave it attitude has the distinct odor of hydrogen sulfide about it.
If you mean Red Hat needs to do all the work then that does not make it a community effort. Other folks contributing does. Luckily someone just did that and if see the fedora-devel list discussion you could have seen me, Warren etc welcoming that effort wholeheartedly. Unfortunately subsequent communication has been stalled due to non responses. We cannot collaborate more since this require changes in the trademark guidelines and other technical details to be worked out. We are looking for more community input on different use cases that help us modify the guidelines to support such needs. If you provide that, do email gdk AT fedoraproject.org
To me thats ignoreing any 'community effort', and quite frankly I'm less than impressed with this newfound selling of the word community, when its been rather pointedly ignored until now.
New found? I remember the original announcements mentioning this idea.
If you want it to be a community effort, then its time a hell of a lot more credit was given to those that have contributed.
But we did. We welcomed it several times
And that doesn't by any means include me. I'm not looking for any flowers because I haven't contributed squat in terms of code.
My code contributions are not that large either but there are several other means to contribute. See http://fedoraproject.org/wiki/HelpWanted
If there is to be a foundation that we need to support else fedora is going away, then lets see a by-laws draft that will give _us_, the users, a voice.
Yes community voice is the aim. If you have experience working on by-laws and can provide input, do email sopwith AT redhat and send a copy to gdk AT fedoraproject.org, directors@fedorafoundation.org. Thank you in advance for your input.
On Sun, 2006-01-22 at 22:34 +0530, Rahul Sundaram wrote:
Hi
Nobody stopped anyone from doing the work involved. In fact it was already done before. http://fedora.isphuset.no/
We ARE well aware of that. As stated above, I have the cd's already burnt. But when asking a question and disclosing that the question involves the unofficial 4.2 release, then the question was summarily ignored, as being off topic I guess. If you are not willing to even discuss something someone has done in the interests of widening the usage, then the whole thing is moot, and we are wasting our collective time with this endeavor called a foundation.
You should have welcomed that effort with open arms, and actively pointed folks having trouble with _your_ release to that one as a possible solution to the install debacle 4 was.
I repeatedly wrote the the people who did it with no responses. We needed to work on several details before adopting such efforts in a formal way.
See my previous private response on this one.
As mentioned I only recieved one mail from you, and it was rather unclear about your intentions. Seeing as it looked a lot like the other "Thanks" and "you should do this" mail I dismissed it as that. If there were more I cannot see that I have received any.
I'm of the opinion that fedora folks themselves should have done the respin as you call it, within 3-4 days time when it became obvious that 4 was a wholesale x crasher that trashed the disks as it went away. The take it or leave it attitude has the distinct odor of hydrogen sulfide about it.
I myself tried to contact redhat people on a few occations, but never got any responses either. I guessed redhat is a big organization and had no time for little me.
What really irritated me was that the installer problems never saw any updates in FC4. What I mean is that anaconda never got fixed. I got several mail telling me to "atleast fix the anaconda problems" but due to legalese I did not dare to deviate from the standard FC4/FC4-updates packages.
The two most prominent problems that also exist in FC4.2: 1. Swap label garbage 2. Installer cd won't boot unless you feed it garbage.
Inquires about whether the kernel sitting in updates-testing for weeks will get released within the next few week or so went unanswered. Resulting in a brand new kernel just a few days after FC4.2 release. (2.6.13* -> 2.6.14* jump)
If you mean Red Hat needs to do all the work then that does not make it a community effort. Other folks contributing does. Luckily someone just did that and if see the fedora-devel list discussion you could have seen me, Warren etc welcoming that effort wholeheartedly. Unfortunately subsequent communication has been stalled due to non responses. We cannot collaborate more since this require changes in the trademark guidelines and other technical details to be worked out. We are looking for more community input on different use cases that help us modify the guidelines to support such needs. If you provide that, do email gdk AT fedoraproject.org
Fixing anaconda sounds to me like a community benefit. I assume it was not fixed because RH and FC4 itself didn't need it. Had it been *insert other community distro of choice here* it would have been maintained and updated like any other package.
To me thats ignoreing any 'community effort', and quite frankly I'm less than impressed with this newfound selling of the word community, when its been rather pointedly ignored until now.
New found? I remember the original announcements mentioning this idea.
If you want it to be a community effort, then its time a hell of a lot more credit was given to those that have contributed.
But we did. We welcomed it several times
What *I* read was several thanks from *redhat.com ppls that seemed to be from the persons, not company. IE: I didn't see them as official.
-HK
Hi
See my previous private response on this one.
As mentioned I only recieved one mail from you, and it was rather unclear about your intentions. Seeing as it looked a lot like the other "Thanks" and "you should do this" mail I dismissed it as that. If there were more I cannot see that I have received any.
I did send subsequent mails with even more details. You seem to have missed out on them. Please accept my apologies if my messages came off as something that was not intended. I send hundreds of mails over a day and I might have been curt on occasions.
I myself tried to contact redhat people on a few occations, but never got any responses either. I guessed redhat is a big organization and had no time for little me.
Can I put it right back to you. I send a mail and I didnt a response either. Does that mean that little you ignored big Red Hat ;-) Which people did you contact about this within Red Hat. If you can give names, I will talk to them and see if they really got a mail. Your mails might have been lost in the MX issue we had with spams earlier on @redhat.com ids.
What really irritated me was that the installer problems never saw any updates in FC4. What I mean is that anaconda never got fixed. I got several mail telling me to "atleast fix the anaconda problems" but due to legalese I did not dare to deviate from the standard FC4/FC4-updates packages.
What legalese? As long as you dont call it a formal Fedora release you can deviate as much as you want. If you want to see this effort happen within the Fedora Project. see below
The two most prominent problems that also exist in FC4.2:
- Swap label garbage
- Installer cd won't boot unless you feed it garbage.
Inquires about whether the kernel sitting in updates-testing for weeks will get released within the next few week or so went unanswered.
Whom did you ask? Might have been lost in list noise
Fixing anaconda sounds to me like a community benefit. I assume it was not fixed because RH and FC4 itself didn't need it.
Most likely because RH was focusing its effort on driving ahead within its finite resources. Respin efforts and Fedora Legacy project is how the community can help and has helped here. Fedora Legacy is now a official project within the Fedora Foundation. You can send a proposal to the Fedora Foundation - directors AT fedorafoundation.org and copy me the mail so that I can follow up on this and get this done as formal project too. I see that you have legal concerns and liability issues. We can help sort that out. Do provide us a list of requirements and concerns while sending a proposal if you want to work on this within the community perhaps as a team then and lets see what we can do to address this in the right way.
What *I* read was several thanks from *redhat.com ppls that seemed to be from the persons, not company. IE: I didn't see them as official.
True but we did send messages welcoming you. If you had wanted to get an official response and atleast responded to me, I would have been able to help you. Not blaming anyone. Just clarifying some misconceptions involved.
On Mon, 2006-01-23 at 14:46 +0530, Rahul Sundaram wrote:
Hi
See my previous private response on this one.
As mentioned I only recieved one mail from you, and it was rather unclear about your intentions. Seeing as it looked a lot like the other "Thanks" and "you should do this" mail I dismissed it as that. If there were more I cannot see that I have received any.
I did send subsequent mails with even more details. You seem to have missed out on them. Please accept my apologies if my messages came off as something that was not intended. I send hundreds of mails over a day and I might have been curt on occasions.
Mails disappearing happens all the time. Seems your second one never made it to our any of our MXs, or atleast no logfile has you -> me on that date.
I myself tried to contact redhat people on a few occations, but never got any responses either. I guessed redhat is a big organization and had no time for little me.
Can I put it right back to you. I send a mail and I didnt a response either. Does that mean that little you ignored big Red Hat ;-)
I guess I did ignore your intention with the first mail.
Which people did you contact about this within Red Hat. If you can give names, I will talk to them and see if they really got a mail. Your mails might have been lost in the MX issue we had with spams earlier on @redhat.com ids.
It might have been wrong of me to send the mails directly to redhat ppls. I'd rather not have to blame anyone since that benefits nobody. I'll provide it in a personal if you really need to investigate what happened.
I'm glad we now have an ongoing communication. (Thank you Gene)
What really irritated me was that the installer problems never saw any updates in FC4. What I mean is that anaconda never got fixed. I got several mail telling me to "atleast fix the anaconda problems" but due to legalese I did not dare to deviate from the standard FC4/FC4-updates packages.
What legalese? As long as you dont call it a formal Fedora release you can deviate as much as you want. If you want to see this effort happen within the Fedora Project. see below
Ok, thats not the impression I got when this was discussed before FC4.1 was released. Thus I applied the only-official rpms policy.
In any case, I tried to fix it but anaconda seemed like a big mess to my eyes. Was hard enough to read the code to figure out how one could build the installer properly.
Fixing anaconda sounds to me like a community benefit. I assume it was not fixed because RH and FC4 itself didn't need it.
Most likely because RH was focusing its effort on driving ahead within its finite resources. Respin efforts and Fedora Legacy project is how the community can help and has helped here.
Now, this is another point where Fedora and Fedora Legacy should join efforts. If Fedora doesnt want/have time to maintain a package, why cant FC-L get access to do it?
In any case I assume that both problems were fixed upstream at some point, and a new rpm with the patches could have been released pretty easily.
I am currently pondering releasing a FC4.3 after FC5 is released. And then using the FC5 anaconda version instead. Have no idea wether it'll work though. Is it possible to have a fixed anaconda instead?
Fedora Legacy is now a official project within the Fedora Foundation. You can send a proposal to the Fedora Foundation - directors AT fedorafoundation.org and copy me the mail so that I can follow up on this and get this done as formal project too. I see that you have legal concerns and liability issues. We can help sort that out. Do provide us a list of requirements and concerns while sending a proposal if you want to work on this within the community perhaps as a team then and lets see what we can do to address this in the right way.
Now, this is where I get shaky hands.
I need someone to explain to me what this would need from me. And what benefits would there be to do this instead of continuing like what I have? (I'm sure there are some, and can think of a few obvious but a few examples wouldn't hurt)
You do need to sell me this idea as I don't like jumping into something unknown with both feet.
What *I* read was several thanks from *redhat.com ppls that seemed to be from the persons, not company. IE: I didn't see them as official.
True but we did send messages welcoming you. If you had wanted to get an official response and atleast responded to me, I would have been able to help you. Not blaming anyone. Just clarifying some misconceptions involved.
Glad to have that out of the way.
I hope we can progress from here to make Fedora all that it can be. I guess we can work the details in private from here.
-HK
Hi
Fedora Legacy is now a official project within the Fedora Foundation. You can send a proposal to the Fedora Foundation - directors AT fedorafoundation.org and copy me the mail so that I can follow up on this and get this done as formal project too. I see that you have legal concerns and liability issues. We can help sort that out. Do provide us a list of requirements and concerns while sending a proposal if you want to work on this within the community perhaps as a team then and lets see what we can do to address this in the right way.
Now, this is where I get shaky hands.
I need someone to explain to me what this would need from me. And what benefits would there be to do this instead of continuing like what I have? (I'm sure there are some, and can think of a few obvious but a few examples wouldn't hurt)
You do need to sell me this idea as I don't like jumping into something unknown with both feet.
How about you email gdk AT fedoraproject.org and tell us what your concerns are and what exactly are the requirements that would help you do your job better. He might be able to tell you better than me what the benefits are.
Glad to have that out of the way.
I hope we can progress from here to make Fedora all that it can be. I guess we can work the details in private from here.
Sure. We can. This is the big bad world out there ;-)
On Sun, 2006-01-22 at 05:58, Rahul Sundaram wrote:
Very good example. What kind of community is it that can't be bothered to respin the official isos once in a while so the known-broken install on certain hardware gets fixed and the user that manages to install isn't immediately faced with a gig of update downloads just to get started?
Nobody stopped anyone from doing the work involved. In fact it was already done before. http://fedora.isphuset.no/
I don't see how you reconcile 'not stopping someone else' with being a community effort. How does this help a new user coming to the fedora web site downloading their first copy long after the fix has been done? At best it only helps people who read the mail list before their first download (probably none...).
Whenever the issue has come up, I've always tried to point out the k12ltsp versions that not only included updates in the respins but add a lot of value with extra programs and one-click links to install some that can't be included. Have you been in touch with Eric Harrison who builds this version? He seems to have a very different opinion about the difficulty vs. value of building respins whenever someone reports a problem. You'd be doing everyone a favor if you could arrange to mirror his rebuild on the official fedora site with a big link that says 'Hey - this version is better for a lot of reasons'. This is work that is already done and it is just a matter of embracing it. As it stands, I don't think people 'get it' that the k12ltsp version really is a stock updated fedora plus a few add-ons that don't have to be installed.
Les Mikesell wrote:
On Sun, 2006-01-22 at 05:58, Rahul Sundaram wrote:
Very good example. What kind of community is it that can't be bothered to respin the official isos once in a while so the known-broken install on certain hardware gets fixed and the user that manages to install isn't immediately faced with a gig of update downloads just to get started?
Nobody stopped anyone from doing the work involved. In fact it was already done before. http://fedora.isphuset.no/
I don't see how you reconcile 'not stopping someone else' with being a community effort. How does this help a new user coming to the fedora web site downloading their first copy long after the fix has been done? At best it only helps people who read the mail list before their first download (probably none...).
They could a announcement to fedora announce list. I have send a few mails to this team involved in the respin effort and got no responses really. Before endorsing such efforts we need to do some ground work to prevent liability issues for them or the project and introduce some changes in the project guidelines too. Even otherwise I might be able to get others to provide mirrors and some backing if I get some responses.
We are working on improving the QA process steadily to make the GA releases better. Hiring release managers. QA people etc who work exclusively on Fedora etc. I would like to hear more ideas on room for improvement. You might want to try out the FC5test2 release or the test3 one released shortly and provide some feedback to the fedora-test list too so we do a better GA release of FC5.
Whenever the issue has come up, I've always tried to point out the k12ltsp versions that not only included updates in the respins but add a lot of value with extra programs and one-click links to install some that can't be included. Have you been in touch with Eric Harrison who builds this version? He seems to have a very different opinion about the difficulty vs. value of building respins whenever someone reports a problem.
No I havent talked to the k12ltsp people yet though I have discussed this otherwise with some Fedora people. Open to the idea.
You'd be doing everyone a favor if you could arrange to mirror his rebuild on the official fedora site with a big link that says 'Hey - this version is better for a lot of reasons'. This is work that is already done and it is just a matter of embracing it. As it stands, I don't think people 'get it' that the k12ltsp version really is a stock updated fedora plus a few add-ons that don't have to be installed.
Sure. We could do that. You seem to be in touch with the relevant people. Can you request them to contact me? . I will see what can be done.
Rahul Sundaram wrote:
Les Mikesell wrote:
Whenever the issue has come up, I've always tried to point out the k12ltsp versions that not only included updates in the respins but add a lot of value with extra programs and one-click links to install some that can't be included. Have you been in touch with Eric Harrison who builds this version? He seems to have a very different opinion about the difficulty vs. value of building respins whenever someone reports a problem.
No I havent talked to the k12ltsp people yet though I have discussed this otherwise with some Fedora people. Open to the idea.
In response to Les' statement that I have a very different opinion about respins, that has to be viewed in context. The K12LTSP problem set is different than the Fedora problem set, as such it is a apples-to-oranges comparison.
I'd love to see official re-spun FC isos, but I certainly understand the issues and work involved.
You'd be doing everyone a favor if you could arrange to mirror his rebuild on the official fedora site with a big link that says 'Hey - this version is better for a lot of reasons'. This is work that is already done and it is just a matter of embracing it. As it stands, I don't think people 'get it' that the k12ltsp version really is a stock updated fedora plus a few add-ons that don't have to be installed.
Sure. We could do that. You seem to be in touch with the relevant people. Can you request them to contact me? . I will see what can be done.
As Les notes, K12LTSP is a stock updated fedora plus a few add-ons (almost all Fedora Extras packages these days). Not only does it carry the extra QA baggage of a respin, but it is not a 100% "pure" respin. As such, I'm not so sure it would be a good idea to recognize K12LTSP as a re-spun Fedora Core.
This also runs contrary to my goals, which is to get out of the "distribution" business. I'd prefer to work on eliminating the need for K12LTSP to exist[1] rather than expand it ;-)
-Eric
[1] I've been working with LTSP & Ubuntu on over-coming the packaging issues with LTSP. The first crack at this is now in Edubuntu & Debian unstable. I'm planning on porting these packages to FC5, if all goes well maybe they'll make it into Extras...
On Tue, 2006-01-24 at 18:02, Eric Harrison wrote:
In response to Les' statement that I have a very different opinion about respins, that has to be viewed in context. The K12LTSP problem set is different than the Fedora problem set, as such it is a apples-to-oranges comparison.
I'd love to see official re-spun FC isos, but I certainly understand the issues and work involved.
Thanks for responding, Eric. I didn't mean to minimize the effort you put in, but just point out that it can be done and it is a great service for the users.
As Les notes, K12LTSP is a stock updated fedora plus a few add-ons (almost all Fedora Extras packages these days). Not only does it carry the extra QA baggage of a respin, but it is not a 100% "pure" respin. As such, I'm not so sure it would be a good idea to recognize K12LTSP as a re-spun Fedora Core.
I think a "pure" respin has been arranged elsewhere now. However on a second thought, much of the reason I prefer the k12ltsp version has to do with your choices about extra packages and default settings for a lot of things not even directly relating to LTSP and getting the updates included was just a nice side effect. Perhaps going forward life would be easier if (a) fedora had a "seriously-minimal" install that was just enough to get yum working and (b) things that are almost-rebrands could be easily built with an rpm that adds one or more repositories where high-level choices are done in meta-packages or groups. This would permit installing a minimum of outdated code and let all concerned parties maintain their philosophical purity concerning what they permit in their own repositories without making a complete rebranding operation necessary.
Les Mikesell wrote:
(a) fedora had a "seriously-minimal" install that was just enough to get yum working and (b) things that are almost-rebrands
I did in fact generate such a thing, well, a somewhat mutant groaning script-created monstrosity that was based on such a thing, for FC3 here:
A much smaller packageset than the minimal install does in fact hang together with the stock fedora packages without any other dependencies (see step 7 on the page). At least it did in FC3. This packageset has full rpm and yum but had the minimal footprint of the options from Fedora packages that would hang together.
I did suggest on fedora-devel a year or more ago that "Fedora Core" would perhaps be more flexible if it really did mean 'Core', ie, was a minimal functional packageset, and there were onionshells around it based on more extensive packageselection. But in the end the problem is that with 'fullblown' package configurations that are suitable for desktop use, you can't really reach down into even ten times the memory footprint possible with a Busybox-style implementation, so Fedora will not stretch into machines that would like a "seriously-minimal" implementation.
-Andy
Andy Green andy@warmcat.com wrote: Les Mikesell wrote:
(a) fedora had a "seriously-minimal" install that was just enough to get yum working and (b) things that are almost-rebrands
I did suggest on fedora-devel a year or more ago that "Fedora Core" would perhaps be more flexible if it really did mean 'Core', ie, was a minimal functional packageset, and there were onionshells around it based on more extensive packageselection. But in the end the problem is that with 'fullblown' package configurations that are suitable for desktop use, you can't really reach down into even ten times the memory footprint possible with a Busybox-style implementation, so Fedora will not stretch into machines that would like a "seriously-minimal" implementation.
Sounds like one heckuva idea. Why not just update the core-utils to a minimal run level, and update the rest as needed? How much could get broke? A little? a Lot? It would sure circumvent the concerns of those with a smoothly running Fedora 3, into Fedora 5, maybe with a couple of hiccups.
<waves hands wildly> I vote for this... Red Hat, wouldn't this be better for the large users as well? Why didn't I think of this? <g> I mean you could darn near yum this thing, and use a boot floppy instead of 4 CD's worth.
Sound idea <tip of the baeball cap to Andy's general direction> Raul? Ric
================================================ My father, Victor Moore (Vic) used to say: "There are two Great Sins in the world... ...the Sin of Ignorance, and ...the Sin of Stupidity. Only the former may be overcome." R.I.P. Dad.
Linux user# 44256 Sign up at: http://counter.li.org/ ================================================ --------------------------------- Do you Yahoo!? With a free 1 GB, there's more in store with Yahoo! Mail.
Rickey Moore wrote:
*/Andy Green andy@warmcat.com/* wrote:
Les Mikesell wrote: > (a) fedora had a "seriously-minimal" install that was just > enough to get yum working and (b) things that are almost-rebrands I did suggest on fedora-devel a year or more ago that "Fedora Core" would perhaps be more flexible if it really did mean 'Core', ie, was a minimal functional packageset, and there were onionshells around it based on more extensive packageselection. But in the end the problem is that with 'fullblown' package configurations that are suitable for desktop use, you can't really reach down into even ten times the memory footprint possible with a Busybox-style implementation, so Fedora will not stretch into machines that would like a "seriously-minimal" implementation.
Sounds like one heckuva! idea. Why not just update the core-utils to a minimal run level, and update the rest as needed? How much could get broke? A little? a Lot? It would sure circumvent the concerns of those with a smoothly running Fedora 3, into Fedora 5, maybe with a couple of hiccups.
<waves hands wildly> I vote for this... Red Hat, wouldn't this be better for the large users as well? Why didn't I think of this? <g> I mean you could darn near yum this thing, and use a boot floppy instead of 4 CD's worth.
Sound idea <tip of the baeball cap to Andy's general direction> Raul? Ric
I would suggest bringing this idea with the above link to the fedora-devel list after FC5 is released. The work being done on OLPC project (http://www.redhat.com/magazine/014dec05/features/olpc/) can also be potentially leveraged here.
HI
This is a sentiment I have to echo also. The kde/gnome wars being a case in point. kde efforts are only on a time permitted basis it still seems, often done by soemone on his or her own time even after a sufficiently large numnber of users choose to run it. That doesn't equate to being a 'community project' IMNSHO.
Red Hat understands this very well. See the KDE proposal in http://fedoraproject.org/wiki/KDE
On Sat, Jan 21, 2006 at 11:04:13AM -0600, Les Mikesell wrote:
On Sat, 2006-01-21 at 01:21, Rahul Sundaram wrote:
Very unlikely. RHEL is too boring for a community to participate. It is almost stagnant which is great for enterprises but not for the community. If RHEL would have been suitable for everyone Red Hat wouldnt started Fedora or vice versa.
Your argument would be more convincing if you could point out some directions that FC has taken that are more in the community interest than in building the next RHEL. Can you?
RHEL4 looks very much like a direct cutover from what was built in FC3 with Mysql4.x dropped in, and so far everything in subsequent FC releases looks like more of the same. If this is really supposed to be a community project, shouldn't it take some directions that aren't aimed at building the next RHEL version?
One can argue to what extent the community project Fedora is. But, it is a community project whose purpose is too improve Red Hat's Linux offering. They pay the core people who design and work on it so in some sense they have a right to use it for their purposes. But as I have said many times before I wish Red Hat would be more open to the opinions of the community. ------------------------------------------- Aaron Konstam Computer Science Trinity University telephone: (210)-999-7484
Hi
One can argue to what extent the community project Fedora is. But, it is a community project whose purpose is too improve Red Hat's Linux offering. They pay the core people who design and work on it so in some sense they have a right to use it for their purposes. But as I have said many times before I wish Red Hat would be more open to the opinions of the community.
Within the constraints, can you state your opinions in detail?
On 1/21/06, Les Mikesell lesmikesell@gmail.com wrote:
Your argument would be more convincing if you could point out some directions that FC has taken that are more in the community interest than in building the next RHEL. Can you?
Define "community interest" This is NOT a monolithic community with a single set of opinions and a single non-overlap set of desires. There is no perfect solution that meets all desires. For every opinion I have.. there will undoubtable be someone else in the community who disagrees with my opinion. Does that mean my opinion doesn't count as "community interest" or that the other person's doesn't?
But if you want an example of FC and RHEL differientiating. I will point out that the incremental switch-over to a yum based set of tools and infrastructure in Fedora which is not paralleled with RHEL and its current reliance on rhn to date. The tools aren't finished, it is an incremental improvement process... but I feel its most certaintly in the fedora community's interest to have a set of default tools which don't hook into the rhn-like client-server model. And someone will undoubtable disagree with my opinion.. proving my point about it being very difficult to define a simple definition of "community interest" by looking at vocal opinionated people.
Would you also count the work going on with Kadischi as being an RHEL inspired subproject? http://fedoraproject.org/wiki/Kadischi I wouldn't. From where I'm sitting Kadischi is a very new development, where outside community are taking an active roll in its development and its direction. Has Kadischi born fruit in the form of an official fedora livecd yet.. no. But people in the community are working towards that goal.. a goal that is not clearly tied to the RHEL business model and appears to be generated in no small part by "community interest" and "community effort"
-jef
On 1/21/06, Jeff Spaleta jspaleta@gmail.com wrote:
On 1/21/06, Les Mikesell lesmikesell@gmail.com wrote:
Your argument would be more convincing if you could point out some directions that FC has taken that are more in the community interest than in building the next RHEL. Can you?
Oh and I forgot.
The ppc and x86_64 releases are both the products of community interest and community contributions. Back in earlier fc releases, it was going to be pc 32bit only.. because that is what RedHat was prepared to maintain. But people outside redhat, in the larger community stepped up and did a lot of the hard work getting the x86_64 distribution to the point where Fedora could ship a release officially. And a similar community effort for ppc happened to make ppc a tier one architecture release that is part of official Core releases now.
-jef
On Sat, 2006-01-21 at 16:32, Jeff Spaleta wrote:
Your argument would be more convincing if you could point out some directions that FC has taken that are more in the community interest than in building the next RHEL. Can you?
Define "community interest"
I'd start with making a new user's first time experience as painless as possible.
This is NOT a monolithic community with a single set of opinions and a single non-overlap set of desires.
Can we have a show of hands from the people who think a new user should have to download an iso image full of bugs that have been fixed for most of a year that may prevent the install from working and that will require a gig or so of additional downloading to fix if the install succeeds?
Would you also count the work going on with Kadischi as being an RHEL inspired subproject? http://fedoraproject.org/wiki/Kadischi I wouldn't.
It looks like a badly-need catch up to all the other popular distributions to me.
Hi
Define "community interest"
I'd start with making a new user's first time experience as painless as possible.
What can be done to improve that. Can you provide a specific list. You might want to discuss these in fedora-devel of file bug reports and RFE's to bugzilla if its specific enough.
Would you also count the work going on with Kadischi as being an RHEL inspired subproject? http://fedoraproject.org/wiki/Kadischi I wouldn't.
It looks like a badly-need catch up to all the other popular distributions to me.
Maybe. Maybe they will explore other new options. Kadischi does generate live cd's from a collection of packages which is quite useful. Others doing such things doesnt negate the usefulness of it for the Fedora Project and its pretty good example of some projects done actively from community with help from Red Hat.
On Sun, 2006-01-22 at 13:16, Rahul Sundaram wrote:
I'd start with making a new user's first time experience as painless as possible.
What can be done to improve that. Can you provide a specific list. You might want to discuss these in fedora-devel of file bug reports and RFE's to bugzilla if its specific enough.
The free availability of the VMware vmplayer really changes the picture for test-drive releases. I'd say the single most dramatic thing that could be done now for new-user, new-version rollout would be to maintain up to date virtual machine images for full and minimal installations and promote them on the main download site. These would take none of the changes that a livecd version takes and are even easier to run since you don't have to reboot or shut down your existing system. Images exist for FC4 but I can't tell if they are current with updates or not, and I can't find any yet for the FC5 test versions where they would really make sense. I'd probably be running under one right now if I had found one.
It would also be a nice touch to make a scripted method to migrate a working/updated virtual machine to physical hardware so you don't have have to start from scratch when you get everything the way you like and decide to replace your current system. This can be done by hand by mucking around with modules.conf and rebuilding the initrd image but it isn't pretty. Actually this should be the same thing you need when moving to different physical hardware and not really tied to the vmware install.
Hi
The free availability of the VMware vmplayer really changes the picture for test-drive releases. I'd say the single most dramatic thing that could be done now for new-user, new-version rollout would be to maintain up to date virtual machine images for full and minimal installations and promote them on the main download site.
It could be but it isnt within the project scope to support any proprietary software. Red Hat has a nice team of developers working on Xen instead which could serve similar purposes. FC5 will likely have a pretty good implementation of Xen.
On Sun, 2006-01-22 at 15:28, Rahul Sundaram wrote:
The free availability of the VMware vmplayer really changes the picture for test-drive releases. I'd say the single most dramatic thing that could be done now for new-user, new-version rollout would be to maintain up to date virtual machine images for full and minimal installations and promote them on the main download site.
It could be but it isnt within the project scope to support any proprietary software.
Please explain how making an image available that runs under the free vmplayer 'supports' proprietary software any more than making an iso image available 'supports' CD and media manufacturers. And more to the point, how does this relate to being community oriented, which I thought was the topic at hand? Was it a community choice to not support any proprietary software (in whatever way you see this as support)? I don't see how this choice helps anyone.
Red Hat has a nice team of developers working on Xen instead which could serve similar purposes. FC5 will likely have a pretty good implementation of Xen.
When Xen provides the same functionality, including the ability to test-drive a linux distribution under a running Windows installation, it would be reasonable to discuss whether it is worth maintaining both. Right now, there's only one choice and using it would help get the product you do have some better exposure.
Les Mikesell wrote:
On Sun, 2006-01-22 at 15:28, Rahul Sundaram wrote:
The free availability of the VMware vmplayer really changes the picture for test-drive releases. I'd say the single most dramatic thing that could be done now for new-user, new-version rollout would be to maintain up to date virtual machine images for full and minimal installations and promote them on the main download site.
It could be but it isnt within the project scope to support any proprietary software.
Please explain how making an image available that runs under the free vmplayer 'supports' proprietary software any more than making an iso image available 'supports' CD and media manufacturers.
Pretty simple. Vmware is proprietary software. ISO images and hardware that use them are not. The CD and media manufactures deal with open interfaces to the hardware supported natively in Linux kernel. Fedora does not provide proprietary drives for some hardware either.
And more to the point, how does this relate to being community oriented, which I thought was the topic at hand? Was it a community choice to not support any proprietary software (in whatever way you see this as support)? I don't see how this choice helps anyone.
It is not but the project also serves other goals. No matter how many people ask for it, Fedora project will continue to support only legally unencumbered Free and open source software. This goal cannot be changed by the community just like you cannot ask Linux kernel developers to include non-GPL'ed software. Fortunately the community is generally supportive of this goal.
http://fedoraproject.org/wiki/Marketing/Fedora
Red Hat has a nice team of developers working on Xen instead which could serve similar purposes. FC5 will likely have a pretty good implementation of Xen.
When Xen provides the same functionality, including the ability to test-drive a linux distribution under a running Windows installation, it would be reasonable to discuss whether it is worth maintaining both.
Xen can very well support Windows. In fact Xen did support a modified version of Windows before MS decided to pull off the project. Now it can be supported through hardware virtualisation. Intel and AMD will start supporting it in their next generation processors. Xen will be able to run more operating systems and support more architectures with better security through fine tuned isolation of processes and near native performance for guest operating systems. See http://fedoraproject.org/wiki/Tools/xen and related references for more information.
Right now, there's only one choice and using it would help get the product you do have some better exposure.
Sure. There are already VMWare images for FC5 test versions. You can google them. I merely said that the project by itself wouldnt support it.
On Sun, 2006-01-22 at 17:56, Rahul Sundaram wrote:
Please explain how making an image available that runs under the free vmplayer 'supports' proprietary software any more than making an iso image available 'supports' CD and media manufacturers.
Pretty simple. Vmware is proprietary software.
No argument there. But, you aren't providing or supporting it by providing an image that can run under it. You would be supporting your users that would find that option useful.
ISO images and hardware that use them are not.
What? An image is an image. It's just a fedora distribution in a different format.
And more to the point, how does this relate to being community oriented, which I thought was the topic at hand? Was it a community choice to not support any proprietary software (in whatever way you see this as support)? I don't see how this choice helps anyone.
It is not but the project also serves other goals. No matter how many people ask for it, Fedora project will continue to support only legally unencumbered Free and open source software.
Well, now we are back to square one. You can't say "community oriented" and "no matter how many people ask for it" at the same time and make any sense.
Sure. There are already VMWare images for FC5 test versions. You can google them. I merely said that the project by itself wouldnt support it.
I wasn't able to find any when I tried a couple of days ago. Sure, anyone can do this, but wouldn't it be a good idea for someone who cared about promoting the project to make sure it actually works and is kept up to date. In fact if such a person had done it with FC5 test1 he would have noticed that it didn't work right and might have been able to get it fixed before the iso image release.
Les Mikesell wrote:
On Sun, 2006-01-22 at 17:56, Rahul Sundaram wrote:
Please explain how making an image available that runs under the free vmplayer 'supports' proprietary software any more than making an iso image available 'supports' CD and media manufacturers.
Pretty simple. Vmware is proprietary software.
No argument there. But, you aren't providing or supporting it by providing an image that can run under it. You would be supporting your users that would find that option useful.
The project does not provide support by proprietary software. Period.
Well, now we are back to square one. You can't say "community oriented" and "no matter how many people ask for it" at the same time and make any sense.
Of course it makes perfect sense. Is not Apache a community project because it does not include proprietary software and only includes software under the Apache license?. Is not the Linux Kernel community project because the Kernel developers wouldnt not support proprietary modules?. Is not Debian a open and community project because it doesnt include proprietary beyond a informal non-free repository (Fedora went beyond this level by not even including such repositories) ?. If all of these projects are community oriented, then so is Fedora. Also remember that community has no way to participate in development of proprietary software and including such software would reduce the community ability to participate. The idea of not including proprietary and enabling the community is tied to one another in this way. Also if you look at the survey when we did ask the community, they were supportive of this goal. So you cant even say they want the project to support it.
http://fedoraproject.org/wiki/Marketing/Fedora.
Sure. There are already VMWare images for FC5 test versions. You can google them. I merely said that the project by itself wouldnt support it.
I wasn't able to find any when I tried a couple of days ago. Sure, anyone can do this, but wouldn't it be a good idea for someone who cared about promoting the project to make sure it actually works and is kept up to date. In fact if such a person had done it with FC5 test1 he would have noticed that it didn't work right and might have been able to get it fixed before the iso image release.
So see someone would have to not only include it but also test and support it. This is something that the project does not want to invest resources in because it isnt against the goals of the project and also because we prioritize and spend time on innovative open source development. Xen, Stateless Linux, SELinux MCS, GFS, Fedora directory server etc, many of them developed by Red Hat and driven for integration within Fedora are good examples of where we are spending our time on.
Rahul Sundaram wrote:
Les Mikesell wrote:
On Sun, 2006-01-22 at 17:56, Rahul Sundaram wrote:
Please explain how making an image available that runs under the free vmplayer 'supports' proprietary software any more than making an iso image available 'supports' CD and media manufacturers.
Pretty simple. Vmware is proprietary software.
Just so that you know we are not leaving you in the dust - http://www.fedorareloaded.com/ has links to Vmware player images done by Red Hat people. It is however NOT in anyway supported by Red Hat or the Fedora Project.
---
Rahul
Fedora Bug Triaging - http://fedoraproject.org/wiki/BugZappers
On Sun, 2006-01-22 at 18:36, Rahul Sundaram wrote:
I wasn't able to find any when I tried a couple of days ago. Sure, anyone can do this, but wouldn't it be a good idea for someone who cared about promoting the project to make sure it actually works and is kept up to date. In fact if such a person had done it with FC5 test1 he would have noticed that it didn't work right and might have been able to get it fixed before the iso image release.
So see someone would have to not only include it but also test and support it.
No, as a side effect of this install the Fedora distribution would be tested, and if anyone cared it could be fixed before releasing the broken iso images. If I didn't make that clear, the FC5 test1 isos were broken in the video handling, not vmware player and that's that needs the support. I'd guess that the same breakage applied to real hardware in the same way as the vmware emulation and it has been fixed in updates. I'm not surprised at a test release having such breakage, of course, but I hope we agree that it would have been in everyone's interest if it had been noticed and fixed before anyone downloaded those isos. It's not a matter of having to do extra support since this had to be fixed anyway - it's a matter of how many people see and are inconvenienced by the broken version and form their opinions based on it.
This is something that the project does not want to invest resources in because it isnt against the goals of the project and also because we prioritize and spend time on innovative open source development.
If making a product easy to test and install isn't a priority, I'm afraid the project may become irrelevant.
Les Mikesell wrote:
On Sun, 2006-01-22 at 18:36, Rahul Sundaram wrote:
I wasn't able to find any when I tried a couple of days ago. Sure, anyone can do this, but wouldn't it be a good idea for someone who cared about promoting the project to make sure it actually works and is kept up to date. In fact if such a person had done it with FC5 test1 he would have noticed that it didn't work right and might have been able to get it fixed before the iso image release.
So see someone would have to not only include it but also test and support it.
No, as a side effect of this install the Fedora distribution would be tested, and if anyone cared it could be fixed before releasing the broken iso images.
Guess we have to miss out testing from people who refuse to use the vmware images regardless of the fact that it has already been provided in http://fedorareloaded.com just like we miss out supporting all those whizbang proprietary kernel drivers by design and if you are testing FC5test2, I would appreciate feedback to fedora-test list. Thanks.
If making a product easy to test and install isn't a priority, I'm afraid the project may become irrelevant.
It is not acceptable to deviate from stated goals. Thats what will destroy what the project is all about.
On Sun, 2006-01-22 at 20:01, Rahul Sundaram wrote:
No, as a side effect of this install the Fedora distribution would be tested, and if anyone cared it could be fixed before releasing the broken iso images.
Guess we have to miss out testing from people who refuse to use the vmware images regardless of the fact that it has already been provided in http://fedorareloaded.com just like we miss out supporting all those whizbang proprietary kernel drivers by design and if you are testing FC5test2, I would appreciate feedback to fedora-test list. Thanks.
Thanks - I guess the Jan 22. date explains why I couldn't find it on Friday. Note that I plan to run this on a machine where I'd be very uncomfortable about letting new and untested drivers touch real hardware, no matter how whizbang they claim to be. I expect it to greatly decrease the time until I am comfortable with using it on a real system, though.
Les Mikesell wrote:
On Sun, 2006-01-22 at 20:01, Rahul Sundaram wrote:
No, as a side effect of this install the Fedora distribution would be tested, and if anyone cared it could be fixed before releasing the broken iso images.
Guess we have to miss out testing from people who refuse to use the vmware images regardless of the fact that it has already been provided in http://fedorareloaded.com just like we miss out supporting all those whizbang proprietary kernel drivers by design and if you are testing FC5test2, I would appreciate feedback to fedora-test list. Thanks.
Thanks - I guess the Jan 22. date explains why I couldn't find it on Friday. Note that I plan to run this on a machine where I'd be very uncomfortable about letting new and untested drivers touch real hardware, no matter how whizbang they claim to be.
They cant possibly damage modern hardware.
I expect it to greatly decrease the time until I am comfortable with using it on a real system, though.
Real systems are better than vmware or xen instances for good testing for several reasons. So even if you can grab such images one way or the other, I would recomming testing in on the hardware directly rather than using virtualisation layers which do no exposure all the hardware for bare testing. Looking forward to feedback on the fedora-test list. Thanks.
On Sun, 2006-01-22 at 20:52, Rahul Sundaram wrote:
Guess we have to miss out testing from people who refuse to use the vmware images regardless of the fact that it has already been provided in http://fedorareloaded.com just like we miss out supporting all those whizbang proprietary kernel drivers by design and if you are testing FC5test2, I would appreciate feedback to fedora-test list. Thanks.
Back to the quality control issue: Clicking the download link at the above site gives a 'Not Found' error:
The requested URL /fedorareloaded.com/podcasts/FedoraCore5-Test2.zip.torrent was not found on this server.
I've had better first impressions...
Note that I plan to run this on a machine
where I'd be very uncomfortable about letting new and untested drivers touch real hardware, no matter how whizbang they claim to be.
They cant possibly damage modern hardware.
I recall seeing discussions of lost data on this mailing list around the introduction of FC2. I think "can't" is a little strong although I didn't expect hardware damage. I just don't want much downtime or chances of mistakes when selecting partitions for the new install.
On Sun, 2006-01-22 at 23:59 -0600, Les Mikesell wrote:
On Sun, 2006-01-22 at 20:52, Rahul Sundaram wrote:
Guess we have to miss out testing from people who refuse to use the vmware images regardless of the fact that it has already been provided in http://fedorareloaded.com just like we miss out supporting all those whizbang proprietary kernel drivers by design and if you are testing FC5test2, I would appreciate feedback to fedora-test list. Thanks.
Back to the quality control issue: Clicking the download link at the above site gives a 'Not Found' error:
The requested URL /fedorareloaded.com/podcasts/FedoraCore5-Test2.zip.torrent was not found on this server.
Didn't Rahul already tell you that is not a fedora project site? Why whine about the broken link here?
I've had better first impressions...
Note that I plan to run this on a machine
where I'd be very uncomfortable about letting new and untested drivers touch real hardware, no matter how whizbang they claim to be.
They cant possibly damage modern hardware.
I recall seeing discussions of lost data on this mailing list around the introduction of FC2. I think "can't" is a little strong although I didn't expect hardware damage. I just don't want much downtime or chances of mistakes when selecting partitions for the new install.
So do like most system admins do. Make sure you have a recovery plan in place in case the install does not go the way it should.
I do backups of my machine before I do anything as destructive as a new install --- Just In Case. Another option would be to install on a new hard drive so the one you are using now can be put back if needed.
That is a common sense approach used by all good admins I know, even on enterprise class servers and with known good proprietary operating systems.
This whole thread started because you are unhappy about the transfer of FC3 to the legacy project. Please get over your attitude of "I want it my way" and realize that you are the only one I recall having seen carrying on like this for the past 2 days.
If a test release has a problem, well --- after all it is a test release and that is why it is being released -- for purposes of identifying show-stopper bugs.
On Mon, 2006-01-23 at 00:27, Jeff Vian wrote:
Back to the quality control issue: Clicking the download link at the above site gives a 'Not Found' error:
The requested URL /fedorareloaded.com/podcasts/FedoraCore5-Test2.zip.torrent was not found on this server.
Didn't Rahul already tell you that is not a fedora project site? Why whine about the broken link here?
Because my entire point about this has been that someone who cares about fedora should be making sure this stuff works. Otherwise people will get the impression that it is not a usable distribution. I've yet to see a response that indicates that there is someone who cares.
This whole thread started because you are unhappy about the transfer of FC3 to the legacy project.
No, I'm pretty neutral on that issue except that I don't want to see the internet full of hacked FC3 boxes where the administrator is still doing updates against a no longer active repository. If the rollover is transparent I don't care when it happens.
Please get over your attitude of "I want it my way" and realize that you are the only one I recall having seen carrying on like this for the past 2 days.
Most of the postings have been to correct other people's knee-jerk misinterpretations of what I said. I can take care of my own boxes - I'd just like fedora to be something I could recommend to someone else.
Les Mikesell wrote:
On Mon, 2006-01-23 at 00:27, Jeff Vian wrote:
Back to the quality control issue: Clicking the download link at the above site gives a 'Not Found' error:
The requested URL /fedorareloaded.com/podcasts/FedoraCore5-Test2.zip.torrent was not found on this server.
Didn't Rahul already tell you that is not a fedora project site? Why whine about the broken link here?
Because my entire point about this has been that someone who cares about fedora should be making sure this stuff works. Otherwise people will get the impression that it is not a usable distribution. I've yet to see a response that indicates that there is someone who cares.
Dont you care?. If you do then you go ahead and create a vmplayer image and host it or atleast complain to the people who created the image and send the feedback to them. Merely complaining in the wrong forum and arguing that a VMware image that doesnt work with anything other than a VMWare proprietary player is not proprietary. #whois fedoraloaded.com for contact details . Dont even think anybody at all cares for Fedora? . Stuff doesnt get happened automatically buddy. You are harping on a non-existent issue.
On Mon, 2006-01-23 at 02:32, Rahul Sundaram wrote:
Because my entire point about this has been that someone who cares about fedora should be making sure this stuff works. Otherwise people will get the impression that it is not a usable distribution. I've yet to see a response that indicates that there is someone who cares.
Dont you care?. If you do then you go ahead and create a vmplayer image and host it or atleast complain to the people who created the image and send the feedback to them.
I care enough to point out the problem but have no real interest in which distribution is most popular. I don't consider myself an expert on fedora or the apps that should be included in someone's first install. I still run FC1 on my own desktop but I'd imagine that you'd like to showcase subsequent improvements that I know nothing about.
Dont even think anybody at all cares for Fedora? . Stuff doesnt get happened automatically buddy. You are harping on a non-existent issue.
I'll believe it when I see improvements in the user install process that match or surpass other distributions.
On 1/23/06, Les Mikesell lesmikesell@gmail.com wrote:
I'll believe it when I see improvements in the user install process that match or surpass other distributions.
fedora doesn't match or surpass the install process of slackware?
-jef
On Mon, 2006-01-23 at 07:59, Jeff Spaleta wrote:
On 1/23/06, Les Mikesell lesmikesell@gmail.com wrote:
I'll believe it when I see improvements in the user install process that match or surpass other distributions.
fedora doesn't match or surpass the install process of slackware?
Interesting comparison. I stopped using slackware as soon as another distro became easier to use. I think it was RedHat 4.x that was the first thing that had a fair chance of installing automatically on a variety of hardware from the CDs. I don't see a lot of reason for brand loyalty in the world of free software.
On 1/23/06, Les Mikesell lesmikesell@gmail.com wrote:
Interesting comparison. I stopped using slackware as soon as another distro became easier to use.
It is an interesting comparison.. considering that slackware topped linuxquestions.org members poll for 2004, which was held in feb of 2005. Perhaps the lesson here is that no-one, not even you, have a good idea as to what the community as a whole is interest in seeing improvement in. If you want to talk from personal experience, fine, be specific as to which improvements in the installer you would like to see, but don't try wrap yourself in the language of the average user perspective nor try to suggest that fedora is out of step with other open source projects simply because you have have decided that certain projects are irrelevant to the common linux experience. You nor I have any way to evaluate whether our opinions are representative of anything more than a vocal minority.
Let me be a tad more blunt. Stop making general comparisons about "other distros." Stop using retorical constructions which attempt to magnify your opinion that fedora is out of step with the larger distribution space. Some distros do things better than others and the evaluation of that depends on the person who is looking over the distros. If you want to talk about the pros/cons of specific functionality be as specific in your comparison drawing as you can possibly be instead of casting wide statements about "other distributions". What you are doing is nothing but grand standing for effect. I demand specifics because I care about this project's success.
What i suggest you do, is to take a moment make a list of specific examples as to where the implementation in fedora installation process could be modified. Start a new thread for each idea, and support each idea with specific examples of similar features implemented in specifically named distributions. If after discussion on each of those threads, you feel that your idea is still worth presenting to the Core developers and Core release team, write an RFE in bugzilla and reference the discussion thread for them to review.
I will speak from personal experience and say, that the more topics this one thread covers, the less likely this discussion can be followed later as a reference. If you are sincerely trying to invoke change in the minds of the people who make decisions, you do yourself a favor by constraining discussion in a particular thread as much as possible to a specific implementable action instead of lumping all your specific ideas together. And it will also be important to follow up with a feature request into bugzilla which references the discussion in the web archive of this mailinglist.
As much fun as it is to argue with you over the definition of things like the line between proprietary hardware and proprietary software... the fact of the matter is.. the final decision on any idea presented here will most likely be made by people who are not active in this discussion. Keep your true audience in mind and make sure you do what you can to keep discussion in any particular thread narrow.
-jef
On Mon, 2006-01-23 at 08:42, Jeff Spaleta wrote:
On 1/23/06, Les Mikesell lesmikesell@gmail.com wrote:
Interesting comparison. I stopped using slackware as soon as another distro became easier to use.
It is an interesting comparison.. considering that slackware topped linuxquestions.org members poll for 2004, which was held in feb of 2005. Perhaps the lesson here is that no-one, not even you, have a good idea as to what the community as a whole is interest in seeing improvement in.
I'd interpret that same data in a somewhat different way. Most likely that fedora users (myself included) have never heard of linuxquestions.org... Or perhaps that it really demonstrates the point that people are not able to make informed decisions easily with test drive versions.
Let me be a tad more blunt. Stop making general comparisons about "other distros."
I don't get it. How else can you evaluate a distribution other than comparing it to existing alternatives?
What you are doing is nothing but grand standing for effect.
Yes, but the only way I can make a point is to mention the obvious counterexample, like the k12ltsp distribution being able to do a respin when needed - even though it is one guy and not his real job.
I demand specifics because I care about this project's success.
It's the process that matters, not any specific thing. When something is broken, don't make users wait for the next release for the fix, only to find that a different set of things are broken at that point. The current process lets that happen. The problem is not any specific thing being broken, it is the difficulty of obtaining the fix as shown by the gig or more of updates you currently would have to download after an FC4 install - assuming your hardware works with the FC4 installer.
What i suggest you do, is to take a moment make a list of specific examples as to where the implementation in fedora installation process could be modified.
The issues have already been covered. Respin the isos after any installer issue is fixed and whenever some reasonable threshold of updates have accumulated - like 500M. Make a test drive version, whether live CD or vmware or both and keep it up to date so any potential new user or someone who has held off updating from earlier releases can easily see if the current version with updates is suitable for any particular use. When rolling support to legacy, update the yum config so security updates continue for administrators who are not tracking the change. The common thread here: make it easier and more attractive for new people to participate.
Hi
I don't get it. How else can you evaluate a distribution other than comparing it to existing alternatives?
Might happen on individual project needs and contributions. If its a commercial product it can be based on revenue and customer demands. It can also be done by establishing new markets through research and so on.
Yes, but the only way I can make a point is to mention the obvious counterexample, like the k12ltsp distribution being able to do a respin when needed - even though it is one guy and not his real job.
I requested a facilitation since you cite this so often. Would you do that?
It's the process that matters, not any specific thing. When something is broken, don't make users wait for the next release for the fix, only to find that a different set of things are broken at that point. The current process lets that happen. The problem is not any specific thing being broken, it is the difficulty of obtaining the fix as shown by the gig or more of updates you currently would have to download after an FC4 install - assuming your hardware works with the FC4 installer.
We are already started to work with the people who did the respins. So if you have anything specific to add on that, mail me off list.
The issues have already been covered. Respin the isos after any installer issue is fixed and whenever some reasonable threshold of updates have accumulated - like 500M.
Anybody could do that. The project would do whatever it can to enable this to happen.
Make a test drive version, whether live CD
Live CD effort is in progress
or vmware
VMware will not be supported or used in Fedora.
When rolling support to legacy, update the yum config so security updates continue for administrators who are not tracking the change.
Effort in progress again.
The common thread here: make it easier and more attractive for new people to participate.
We need specific responses to the development list instead of random user discussions if you want to help make an impact. Thanks
On Mon, 2006-01-23 at 12:12, Rahul Sundaram wrote:
Yes, but the only way I can make a point is to mention the obvious counterexample, like the k12ltsp distribution being able to do a respin when needed - even though it is one guy and not his real job.
I requested a facilitation since you cite this so often. Would you do that?
I really have no connection other than participating on their mailing list, but I'll give it a try. However, the way the rest of this conversion has gone, I foresee it turning out not to be acceptable due to his inclusion of one-click links to install flash, acrobat reader, and realplayer.
Les Mikesell wrote:
On Mon, 2006-01-23 at 08:42, Jeff Spaleta wrote:
[snip]
Let me be a tad more blunt. Stop making general comparisons about "other distros."
I don't get it. How else can you evaluate a distribution other than comparing it to existing alternatives?
[snip]
I think he meant "stop speaking in generalities".
Mike
On Monday 23 January 2006 09:06, Les Mikesell wrote:
On Mon, 2006-01-23 at 07:59, Jeff Spaleta wrote:
On 1/23/06, Les Mikesell lesmikesell@gmail.com wrote:
I'll believe it when I see improvements in the user install process that match or surpass other distributions.
fedora doesn't match or surpass the install process of slackware?
Interesting comparison. I stopped using slackware as soon as another distro became easier to use. I think it was RedHat 4.x that was the first thing that had a fair chance of installing automatically on a variety of hardware from the CDs. I don't see a lot of reason for brand loyalty in the world of free software.
And thats where I have to pipe up Les. Each of all these competing distros would like to be able to carve out enough of a niche in this thing that they can make a living selling the services that inevitably open up as a market when someone who isn't intimately familiar with both the hardware, and the nuances of a missing } in some C code installs this thing called linux because he's heard its better than Winderz from his friends.
RedHat has done the best job of that so far, but SuSe has made some truely great strides in the last 3 years too from what I read on the net. But don't denigrate the efforts of Pat at Slackware, or the other one? man show just 50 miles north of me in Morgantown WV, Mepis. Pat has continued on at a pace thats been hard on his health, simply because he is a believer.
One of the things that sells a distro to Joe Sixpack, is the enthusiasm of a satisfied user he knows and drinks beer with. In fact, at the onsies and twosies level, its probably the single most important factor to the Joes of this world.
You are not showing any brand loyalty when you make a statement like that, and are simply saying up front that you will take the best deal "as long as its free". Thats not a particularly pleasant image when others view it.
Its certainly not an image I want to project in front of myself. I can irritate folks quite well without that, as I've proved here a couple of times, but I'm still here, cause like Calahans Bar, its the best place to be and I know it.
-- Les Mikesell lesmikesell@gmail.com
On Mon, 2006-01-23 at 13:16, Gene Heskett wrote:
You are not showing any brand loyalty when you make a statement like that, and are simply saying up front that you will take the best deal "as long as its free". Thats not a particularly pleasant image when others view it.
Well, I'm honest whether it's pretty or not. I like apache, perl, java, sendmail, X, firefox, evolution, etc. but the package they come from doesn't matter that much beyond the difficulty of the install and update process. The reason I like unix-like operating systems is for the extent that they are the same, not the quirks of any particular flavor.
Its certainly not an image I want to project in front of myself. I can irritate folks quite well without that, as I've proved here a couple of times, but I'm still here, cause like Calahans Bar, its the best place to be and I know it.
RedHat and derivatives deserve respect, but the reason they became popular had very much to do with the ease of installation of their package compared to others at the time. I'm just suggesting that fedora should continue that tradition.
Les Mikesell wrote:
RedHat and derivatives deserve respect, but the reason they became popular had very much to do with the ease of installation of their package compared to others at the time. I'm just suggesting that fedora should continue that tradition.
Well looking at the respin guy, he just upped and did it and now "Fedora" in the form of people, foundations, users and so on deal with the reality of its existence.
That's a lot more powerful in terms of getting where you say we should be than the best-worded explanation of why it SHOULD exist.
-Andy
Jeff Spaleta wrote:
On 1/23/06, Les Mikesell lesmikesell@gmail.com wrote:
I'll believe it when I see improvements in the user install process that match or surpass other distributions.
fedora doesn't match or surpass the install process of slackware?
About Slackware, I dunno, but it does not match or surpass the install process for Debian. I do find that the documentation for Fedora surpasses that for Debian, though.
Speaking as one who has installed and maintains both.
Mike
Hi
I care enough to point out the problem but have no real interest in which distribution is most popular.
Point to the people who did the work or where it would make a impact or do the work yourself. There is nothing stopping you.
Dont even think anybody at all cares for Fedora? . Stuff doesnt get happened automatically buddy. You are harping on a non-existent issue.
I'll believe it when I see improvements in the user install process that match or surpass other distributions.
Be specific. If you have development discussions post to fedora-devel list. For bugs use http://bugzilla.redhat.com
Hi
No, as a side effect of this install the Fedora distribution would be tested, and if anyone cared it could be fixed before releasing the broken iso images. If I didn't make that clear, the FC5 test1 isos were broken in the video handling, not vmware player and that's that needs the support. I'd guess that the same breakage applied to real hardware in the same way as the vmware emulation and it has been fixed in updates. I'm not surprised at a test release having such breakage, of course, but I hope we agree that it would have been in everyone's interest if it had been noticed and fixed before anyone downloaded those isos. It's not a matter of having to do extra support since this had to be fixed anyway - it's a matter of how many people see and are inconvenienced by the broken version and form their opinions based on it.
I am sure people can calliberate their expectations according to the appropriate notes within the announcement, installer, release notes etc about this being a test release. Users are intelligent to understand that test releases might have bugs.
On 1/22/06, Les Mikesell lesmikesell@gmail.com wrote:
On Sun, 2006-01-22 at 17:56, Rahul Sundaram wrote:
Please explain how making an image available that runs under the free vmplayer 'supports' proprietary software any more than making an iso image available 'supports' CD and media manufacturers.
Pretty simple. Vmware is proprietary software.
No argument there. But, you aren't providing or supporting it by providing an image that can run under it. You would be supporting your users that would find that option useful.
It doesn't take a rocket scientist to figure out that if Fedora Core specifically targets a proprietary piece of software that virtualizes a machine and sets out to run Fedora Core under that proprietary software, then Fedora Core is *supporting* that proprietary software.
Mike, you're smarter than this. I know it, I've seen it. Give up this crusade.
-- Chris
"I trust the Democrats to take away my money, which I can afford. I trust the Republicans to take away my freedom, which I cannot."
On Sun, 2006-01-22 at 19:34, Christofer C. Bell wrote:
Pretty simple. Vmware is proprietary software.
No argument there. But, you aren't providing or supporting it by providing an image that can run under it. You would be supporting your users that would find that option useful.
It doesn't take a rocket scientist to figure out that if Fedora Core specifically targets a proprietary piece of software that virtualizes a machine and sets out to run Fedora Core under that proprietary software, then Fedora Core is *supporting* that proprietary software.
I disagree. A virtual machine is no more proprietary than a physical machine. Do we have any complaints about running under any particular physical machine? I'm not asking for any changes to fedora to tie it to the virtual configuration and I'd agree that would be inappropriate. However as long as you can do a backup of the system and restore to a physical machine just as you can between 2 physical machines I don't see any 'specific targeting' involved. I just think that someone who understands fedora and would like to promote its use could do a better job of building and maintaining the image than an outsider - and that such an image would greatly improve a user's first experience. Maybe there isn't any such person, though.
Mike, you're smarter than this. I know it, I've seen it. Give up this crusade.
I'll give up when I'm convinced that I've presented the point in an understandable way. So far no one has responded in a way that makes me think they have any consideration for an end user at all. How about this approach: do you want to control the first impression a lot of people will have or not? VMware images will almost certainly be made. Let's assume they'll be done badly and not maintained. Would, for example, the FC5 test1 install with video that is partly doubled and out-of-sync looking be a good thing for new users to find as their first test drive? If the image building is left to disinterested outsiders or perhaps even people with a vested interest in other distributions, that is very likely to happen.
Hi
I disagree. A virtual machine is no more proprietary than a physical machine. Do we have any complaints about running under any particular physical machine?
A physical machine might not require proprietary drivers or software to support it. A proprietary vm image requires a proprietary software to support it. See the difference now?
Mike, you're smarter than this. I know it, I've seen it. Give up this crusade.
I'll give up when I'm convinced that I've presented the point in an understandable way. So far no one has responded in a way that makes me think they have any consideration for an end user at all. How about this approach: do you want to control the first impression a lot of people will have or not?
Its a test release dude.
VMware images will almost certainly be made. Let's assume they'll be done badly and not maintained.
Why do you assume that. Check http://fedorareloaded.com. Let me know what is done badly?
Would, for example, the FC5 test1 install with video that is partly doubled and out-of-sync looking be a good thing for new users to find as their first test drive?
Considering that its a first test release of a rapidly developed system I consider that acceptable. Can you kindly let me know the bug report number of this video issue so that I can check its not still present in the second test release and several updates after that?
If the image building is left to disinterested outsiders or perhaps even people with a vested interest in other distributions, that is very likely to happen.
Like I said it is being done by Red Hat people with a very close association and interest in the project. The only thing thats being done is that the work is not associated with the project at any formal level. So go take it for the test ride and let us know the results in fedora-test list. Hope that helps.
On Sun, 2006-01-22 at 20:31, Rahul Sundaram wrote:
I disagree. A virtual machine is no more proprietary than a physical machine. Do we have any complaints about running under any particular physical machine?
A physical machine might not require proprietary drivers or software to support it. A proprietary vm image requires a proprietary software to support it. See the difference now?
No, not as long as you can backup/restore a working system between them. Hmmm, maybe the way to resolve the conceptual difference would be for a 'fedora expert' to provide a tar image of what they would like a new user to experience, and let someone else drop that into a runnable vmware image. (See the 'no difference' now?) The expertise of the installer/maintainer is what I'd really like to have made available to the end user. The vmplayer platform is just a handy mechanism that is going to be used whether anyone likes it or not. The only question is how well it will be done.
VMware images will almost certainly be made. Let's assume they'll be done badly and not maintained.
Why do you assume that. Check http://fedorareloaded.com. Let me know what is done badly?
That was just for argument to make the point that someone who has a reason to care about fedora should maintain them. By badly I mostly meant that they may be allowed to get out of date. At this point that image can't be more than a few days out of date. If it stays the same after the next iso release I'd consider that bad.
Would, for example, the FC5 test1 install with video that is partly doubled and out-of-sync looking be a good thing for new users to find as their first test drive?
Considering that its a first test release of a rapidly developed system I consider that acceptable. Can you kindly let me know the bug report number of this video issue so that I can check its not still present in the second test release and several updates after that?
It's fixed already in the updates. I did a text-mode 'yum update xorg*' and that came up working. But then when I could see what I was doing and yum offered to update just about everything else it seemed like it would be easier to start over, and much easier if an image was available instead of starting from new isos.
Les Mikesell wrote:
On Sun, 2006-01-22 at 20:31, Rahul Sundaram wrote:
I disagree. A virtual machine is no more proprietary than a physical machine. Do we have any complaints about running under any particular physical machine?
A physical machine might not require proprietary drivers or software to support it. A proprietary vm image requires a proprietary software to support it. See the difference now?
No, not as long as you can backup/restore a working system between them. Hmmm, maybe the way to resolve the conceptual difference would be for a 'fedora expert' to provide a tar image of what they would like a new user to experience, and let someone else drop that into a runnable vmware image. (See the 'no difference' now?) The expertise of the installer/maintainer is what I'd really like to have made available to the end user. The vmplayer platform is just a handy mechanism that is going to be used whether anyone likes it or not. The only question is how well it will be done.
Can you or can you not use the VMware image without VMWare player? . If it cannot be used with a proprietary application then the vmware image is proprietary. It might be available for gratis or you might able to backup or redistribute it even but as long as the content in the image is unusable without a proprietary application it is definitely a proprietary piece of software. Thats the best I can explain it. I am not sure anybody here is going to support your argument that somehow VMWare images are non proprietary. I will drop this discussion here since I dont have anything further to explain on this particular issue. Thanks
On Sun, 2006-01-22 at 21:35, Rahul Sundaram wrote:
Can you or can you not use the VMware image without VMWare player? . If it cannot be used with a proprietary application then the vmware image is proprietary.
I thought I had seen something about the disk image being the same as those for qemu but was never motivated enough to find out what that means.
It might be available for gratis or you might able to backup or redistribute it even but as long as the content in the image is unusable without a proprietary application it is definitely a proprietary piece of software.
Is a tar image proprietary because you have to load it back on proprietary hardware before it runs again? I'd expect a tar<->vmware file converter to be possible and don't see a lot of difference in principle.
Thats the best I can explain it. I am not sure anybody here is going to support your argument that somehow VMWare images are non proprietary. I will drop this discussion here since I dont have anything further to explain on this particular issue. Thanks
As long as someone else does a good job of maintaining the images it doesn't really matter. If they don't, perhaps we could skirt the notion that copying onto a certain file format somehow makes the content proprietary by having a person with fedora expertise supply a kickstart file and a script of yum commands to build and maintain something that would give a desirable first impression of the current state of the project. Someone who isn't offended by free-as-in-gratis can do the rest without knowing what the ideal fedora setup of the day looks like or how to build it.
On Sun, 2006-01-22 at 20:25 -0600, Les Mikesell wrote:
On Sun, 2006-01-22 at 19:34, Christofer C. Bell wrote:
Pretty simple. Vmware is proprietary software.
No argument there. But, you aren't providing or supporting it by providing an image that can run under it. You would be supporting your users that would find that option useful.
It doesn't take a rocket scientist to figure out that if Fedora Core specifically targets a proprietary piece of software that virtualizes a machine and sets out to run Fedora Core under that proprietary software, then Fedora Core is *supporting* that proprietary software.
I disagree. A virtual machine is no more proprietary than a physical machine. Do we have any complaints about running under any particular physical machine? I'm not asking for any changes to fedora to tie it to the virtual configuration and I'd agree that would be inappropriate. However as long as you can do a backup of the system and restore to a physical machine just as you can between 2 physical machines I don't see any 'specific targeting' involved. I just think that someone who understands fedora and would like to promote its use could do a better job of building and maintaining the image than an outsider - and that such an image would greatly improve a user's first experience. Maybe there isn't any such person, though.
A virtual machine on VMware certainly is proprietary. It provides a virtual interface to the physical hardware and who knows if the interface is perfect (only the vendor).
As has already been said. Testing in a virtual environment is not a true test of how the release will perform with the hardware, since it never touches the hardware.
An OS that works perfectly in a virtual environment may break as soon as it is installed on the real thing. Alternatively, it may have bugs in the virtual environment that are introduced by the virtual machine and that do not exist when testing on the real hardware.
An OS must be targeted for where it will be used -- the real physical hardware.
Mike, you're smarter than this. I know it, I've seen it. Give up this crusade.
I'll give up when I'm convinced that I've presented the point in an understandable way. So far no one has responded in a way that makes me think they have any consideration for an end user at all. How about this approach: do you want to control the first impression a lot of people will have or not? VMware images will almost certainly be made. Let's assume they'll be done badly and not maintained. Would, for example, the FC5 test1 install with video that is partly doubled and out-of-sync looking be a good thing for new users to find as their first test drive? If the image building is left to disinterested outsiders or perhaps even people with a vested interest in other distributions, that is very likely to happen.
I seriously doubt any user that is test driving Linux for the first time is likely to go out and grab a FC5T1 or even FC5T2 release. It is clearly stated what those are and any _new_ user will go with the tried-and-true FC4 release that has been around for some time and has had most of the problems triaged out.
An experienced Linux user who wants to see what is new/different/better in FC5 and how it compares to what he is using now will do the testing. Your argument that this is a problem for new users does not hold water here Mike.
-- Les Mikesell lesmikesell@gmail.com
On Sun, 2006-01-22 at 23:59, Jeff Vian wrote:
I seriously doubt any user that is test driving Linux for the first time is likely to go out and grab a FC5T1 or even FC5T2 release. It is clearly stated what those are and any _new_ user will go with the tried-and-true FC4 release that has been around for some time and has had most of the problems triaged out.
Yes, but we've already covered the problems of the known-broken install on some hardware with FC4 and the gig of update downloads if the install works. That doesn't sound like a lot of fun either for a preview.
An experienced Linux user who wants to see what is new/different/better in FC5 and how it compares to what he is using now will do the testing. Your argument that this is a problem for new users does not hold water here Mike.
Heh - OK, I'll play along. The experienced fedora user *knows* there will be problems to overcome with a new release. That would make them even more likely to want a canned virtual machine. Or at least that's the way I see it. Visit the list archives around prior release times if you've forgotten.
Remember in the distant past this topic started because people running FC3 didn't think installing FC4 made sense at this point and were considering other distributions like ubuntu. Shouldn't they have an equally easy way to preview FC5 to see if it is ready for use or worth waiting for?
On 1/23/06, Les Mikesell lesmikesell@gmail.com wrote:
Heh - OK, I'll play along. The experienced fedora user *knows* there will be problems to overcome with a new release. That would make them even more likely to want a canned virtual machine. Or at least that's the way I see it. Visit the list archives around prior release times if you've forgotten.
Remember in the distant past this topic started because people running FC3 didn't think installing FC4 made sense at this point and were considering other distributions like ubuntu. Shouldn't they have an equally easy way to preview FC5 to see if it is ready for use or worth waiting for?
Yes absolutely! There SHOULD be a way to do virtualization with open source technology. Feel free to dig in and help with the xen effort. In the meantime, there is absolutely nothing stopping you or any other community member from purchasing the proprietary vmware software necessary to create the image and then you and any other community member can make that image available as you see fit.
I find it really really odd that you have problems with the qa of the established release process, but yet you only trust the established fedora project release team to provide the player images for users. I smell a contradiction.
-jef
On Mon, 2006-01-23 at 06:33, Jeff Spaleta wrote:
I find it really really odd that you have problems with the qa of the established release process, but yet you only trust the established fedora project release team to provide the player images for users. I smell a contradiction.
No, from an outside viewpoint it all looks like a symptom of the same thing. If there is anyone within the project that cares about usability they need to do better maintenance. The effects of having a broken installer for the entire life of FC4 are the same as not providing a nice, easy test drive version in terms of user experience and opinions.
An outsider would almost think that fedora management really doesn't want it to be a usable product that might compete with some other distribution.
On Mon, 2006-01-23 at 08:15 -0600, Les Mikesell wrote:
On Mon, 2006-01-23 at 06:33, Jeff Spaleta wrote:
An outsider would almost think that fedora management really doesn't want it to be a usable product that might compete with some other distribution.
I'm a future VMware Certified Professional (should be in about a month) and a lead for my company in the Virtualization space. As someone also fairly active in the Fedora community, I'm inclined to think the *best* mix that satisfies everyone's needs here is to see if VMware would be willing to put together and support a FC5T2 and later VM. Anyone care to submit a sample VM to VMware and see if they bite? See the VMware Virtual Machine Center http://www.vmware.com/vmtn/vm/ for details.
There's a Red Hat VM there now, as well as 2 FC4 VMs (minimal install with and without selinux) and a number of VMs for other distros and OSes available in their "Community" section. They also use Ubuntu as a base for their free VMware Browser appliance. What's interesting is that the *source* for the VM is available (a.k.a. the Ubuntu CD and all of the components). The VM itself is most certainly proprietary though (by Raul's definition) since it can only be played using the VMware Player.
Now this does bring a new issue to mind - because something GPL is being distributed inside a proprietary piece of technology, is the whole now a derivative product, thus making the virtualization technology also subject to the GPL? In other words, does distributing a Linux distro in a VM force you to also distribute the VM as GPL? I don't believe this to be the case - but it could be a grey area...
Cheers,
Chris
On Mon, 2006-01-23 at 07:49 -0700, Christopher A. Williams wrote:
On Mon, 2006-01-23 at 08:15 -0600, Les Mikesell wrote:
On Mon, 2006-01-23 at 06:33, Jeff Spaleta wrote:
An outsider would almost think that fedora management really doesn't want it to be a usable product that might compete with some other distribution.
I'm a future VMware Certified Professional (should be in about a month) and a lead for my company in the Virtualization space. As someone also fairly active in the Fedora community, I'm inclined to think the *best* mix that satisfies everyone's needs here is to see if VMware would be willing to put together and support a FC5T2 and later VM. Anyone care to submit a sample VM to VMware and see if they bite? See the VMware Virtual Machine Center http://www.vmware.com/vmtn/vm/ for details.
There's a Red Hat VM there now, as well as 2 FC4 VMs (minimal install with and without selinux) and a number of VMs for other distros and OSes available in their "Community" section. They also use Ubuntu as a base for their free VMware Browser appliance. What's interesting is that the *source* for the VM is available (a.k.a. the Ubuntu CD and all of the components). The VM itself is most certainly proprietary though (by Raul's definition) since it can only be played using the VMware Player.
Now this does bring a new issue to mind - because something GPL is being distributed inside a proprietary piece of technology, is the whole now a derivative product, thus making the virtualization technology also subject to the GPL? In other words, does distributing a Linux distro in a VM force you to also distribute the VM as GPL? I don't believe this to be the case - but it could be a grey area...
---- isn't the entire discussion of Fedora VM image distribution best had on the fedora-developers list?
Craig
On 1/23/06, Christopher A. Williams chrisw01@comcast.net wrote:
As someone also fairly active in the Fedora community, I'm inclined to think the *best* mix that satisfies everyone's needs here is to see if VMware would be willing to put together and support a FC5T2 and later VM
What an excellent idea. The community section already has an fc5t1 image in it as well. And if you follow the link to the fc5t1 community image you see there is already an fc5t2 image at http://www.thoughtpolice.co.uk/vmware/ which isn't in the vmware community list yet.. but it will be. Seems to me the community is responding organically... this is a good thing{tm}.
But I do believe the hilighted "platform vendor" section are images produced by the vendors and not by Vmware. I still don't think its appropriate as a matter of project goals to produce virtual images that require the use of proprietary tools to produce or consume as part of the Fedora project officially. If vmplayer images was simply a compressed disk image created by something like dd, I'd have no problem with it... but it isn't.
Now this does bring a new issue to mind - because something GPL is being distributed inside a proprietary piece of technology, is the whole now a derivative product, thus making the virtualization technology also subject to the GPL? In other words, does distributing a Linux distro in a VM force you to also distribute the VM as GPL? I don't believe this to be the case - but it could be a grey area...
That is a much deeper question, a question you'll need to ask the FSF about, vmware about or get your own expert legal opinion to comment on. Hopefully Vmware took the time to ask the FSF about this issue before opening up their images catelog to include linux. I suspect there isn't an issue here, but I'd much rather avoid parsing the legality of how the GPL interacts with vmplayer as part of a user oriented mailinglist. I don't think we are going to get an appropriate expert opinion on the matter here.
-jef
On Mon, 2006-01-23 at 10:10, Jeff Spaleta wrote:
Now this does bring a new issue to mind - because something GPL is being distributed inside a proprietary piece of technology, is the whole now a derivative product, thus making the virtualization technology also subject to the GPL? In other words, does distributing a Linux distro in a VM force you to also distribute the VM as GPL? I don't believe this to be the case - but it could be a grey area...
That is a much deeper question, a question you'll need to ask the FSF about, vmware about or get your own expert legal opinion to comment on. Hopefully Vmware took the time to ask the FSF about this issue before opening up their images catelog to include linux. I suspect there isn't an issue here, but I'd much rather avoid parsing the legality of how the GPL interacts with vmplayer as part of a user oriented mailinglist. I don't think we are going to get an appropriate expert opinion on the matter here.
I'm not an expert, but I have to think that the ability to back up a virtual machine and restore onto physical hardware would demonstrate clearly that it is 'mere aggregration' with the other work and no dependency or derivation is involved. And even if there were, the vmplayer code would fit into the special exception under section 3 of the GPL permitting components of the operating system where the program runs to be used without enforcing GPL compliance on them.
On 1/23/06, Les Mikesell lesmikesell@gmail.com wrote:
I'm not an expert,
Thank you for proving my point about this not being the appropriate place to get an expert opinion on that matter. Thank you for dragging this discussion further off the original topic into the bowels of layperson intepretation of copyright law and copyright licenses. This is always a productive discussion to have among end-users. Please, every user who has an opinion about any problematic details about specific GPL inteperatations, please respond on this thread, your opinion matters a great deal when it comes to how the GPL is intepreted under copyright law. If enough end-users intepret the GPL a certain way, that will absolutely positively matter if the issue becomes part of a legal proceeding.
Do you have a blog? I'd love to be able to read your non-expert opinions concern all matter of legal or public policy issues without having to wade through all the other fedora specific discussion in this list.
-jef"my sarcasm tags are broken"spaleta
Les Mikesell wrote:
Remember in the distant past this topic started because people running FC3 didn't think installing FC4 made sense at this point and were considering other distributions like ubuntu. Shouldn't they have an equally easy way to preview FC5 to see if it is ready for use or worth waiting for?
Hum maybe that can be useful for some people, but why should RHAT be the people providing it. It's only quite recently IIRC that vmware made a free (not Free) player. Anyone with the full vmware paid-for can install their own test OS. Hm look Dec 12th 2005
http://www.vmware.com/news/releases/player.html
So the FC3 -> FC4 thing couldn't've been done anyway for folks that didn't have full vmware and they're the folks that can do an install themselves.
Also in vmware the guest OS gets a fake set of peripherals that is always the same, it tells them a limited amount about compatibility with their physcial hardware to run the thing in vmware and physical hardware support is where many problems seem to be coming from.
Vmware question is like flash (Macromedia type of flash) again, not sure why it's Fedora's business to be doing proprietary content production free player or not.
-Andy
On 1/22/06, Les Mikesell lesmikesell@gmail.com wrote:
Was it a community choice to not support any proprietary software (in whatever way you see this as support)? I don't see how this choice helps anyone.
If you don't agree with this project's goals of providing only open source software, then perhaps this isn't the project for you.
Or let me ask it another way, can the fedora project create virtual images for vmplayer without using proprietary vmware to produce the images? Would it be appropriate for fedora to use proprietary compilers and other infrastructure pieces to produce the binary packages that make up the distribution?
If people in the community want to spend money on vmware products to produce vmplayer images for other people in the community who want to spend money on vmware products to play vmplayer images.. they are free to do so. But as a contributing member of this open source project, I feel its highly inappropriate for this project to be producing content that requires proprietary tools to either produce or consume. And i will passionately argue against you and anyone else with words both small and polysyllabic who would desire to see this project formally produce or encourage the production of content that requires proprietary tools. If you'd like me to tell you how strongly i feel about this, I invite you to come into the #fedora channel on irc so you can I can have a nice little private chat without disturbing anyone else who might be offended by the words i choose to describe your opinions on the matter.
-jef
On Sun, 2006-01-22 at 18:14, Jeff Spaleta wrote:
Was it a community choice to not support any proprietary software (in whatever way you see this as support)? I don't see how this choice helps anyone.
If you don't agree with this project's goals of providing only open source software, then perhaps this isn't the project for you.
I certainly have no 'brand loyalty' to any single distribution. I mostly like fedora because all the quirky non-standard admin stuff I had to learn for RH 4 through 7 still make it look like I know what I'm doing.
Or let me ask it another way, can the fedora project create virtual images for vmplayer without using proprietary vmware to produce the images?
I don't see how this is different than proprietary versions of real hardware. Or the software in the bios included in them. You have to run something, virtual or real - and all real machines I've seen are proprietary.
Would it be appropriate for fedora to use proprietary compilers and other infrastructure pieces to produce the binary packages that make up the distribution?
I'm neutral on that issue too as long as the result is portable. If using proprietary tools internally would make the developers more productive, then it helps everyone, just as using a proprietary revision control system made Linus more productive when his associates were willing to observe the license.
If people in the community want to spend money on vmware products to produce vmplayer images for other people in the community who want to spend money on vmware products to play vmplayer images.. they are free to do so.
There is no need to spend money on the player. A free version has been released - which is the point of all this. People will be using it to preview all sorts of distributions whether you approve or not. You only need the full version once to create an image which then can be distributed freely. Subsequent updates and maintenance can be done with just the player.
But as a contributing member of this open source project, I feel its highly inappropriate for this project to be producing content that requires proprietary tools to either produce or consume.
It obviously doesn't 'require' vmware to run fedora. And it still won't 'require' it just because someone makes the effort to make it convenient to run there.
And i will passionately argue against you and anyone else with words both small and polysyllabic who would desire to see this project formally produce or encourage the production of content that requires proprietary tools.
No one has even hinted about making it 'require' anything. The issue is just about whether someone close to the project should have a 'quality control' hand in the images that are made available. That is, does anyone care what a potential new user sees on his first test drive?
If you'd like me to tell you how strongly i feel about this, I invite you to come into the #fedora channel on irc so you can I can have a nice little private chat without disturbing anyone else who might be offended by the words i choose to describe your opinions on the matter.
I have no problem making making my opinions public and not much interest in ones that aren't.
On Sun, 2006-01-22 at 17:39 -0600, Les Mikesell wrote:
On Sun, 2006-01-22 at 15:28, Rahul Sundaram wrote:
The free availability of the VMware vmplayer really changes the picture for test-drive releases. I'd say the single most dramatic thing that could be done now for new-user, new-version rollout would be to maintain up to date virtual machine images for full and minimal installations and promote them on the main download site.
It could be but it isnt within the project scope to support any proprietary software.
Please explain how making an image available that runs under the free vmplayer 'supports' proprietary software any more than making an iso image available 'supports' CD and media manufacturers. And more to the point, how does this relate to being community oriented, which I thought was the topic at hand? Was it a community choice to not support any proprietary software (in whatever way you see this as support)? I don't see how this choice helps anyone.
IANAL, and have not followed this entire thread.
IMHO, posting links to a proprietary software package (vmplayer) and recommending it be used amounts to advertising and is most certainly support of that software.
As an outsider, I would certainly perceive that action as supporting and promoting the software.
Rahul Sundaram wrote:
Hi
Define "community interest"
I'd start with making a new user's first time experience as painless as possible.
What can be done to improve that. Can you provide a specific list. You
[snip]
I can provide a non-specific list with one (1) one element in it:
Keep two copies of ISOs around, one for the initial release, and another which is what one would have if he downloaded the initial release, installed, and then use up2date (or yum) to bring the system up to date with the current revision. When a version is retired, remove the "initial release" version of the ISOs and make the "current revision" the only official copy.
Mike
On 1/22/06, Les Mikesell lesmikesell@gmail.com wrote:
Can we have a show of hands from the people who think a new user should have to download an iso image full of bugs that have been fixed for most of a year that may prevent the install from working and that will require a gig or so of additional downloading to fix if the install succeeds?
There is absolutely no garuntee that respins produced ANYWHERE by ANYONE will not introduce a DIFFERENT set of problems for a different set of hardware. Telling everyone to go out and grab a fc4.2 which uses a different installer codebase and an different installer kernel.. is asking for a whole different set of bugs... bugs that have recieved far less testing in terms of hardware coverage than the original installer did. Just because its a newer version doesn't mean there will not be regressions for some hardware.. as we see with update kernels... we will see it with update installer isos as well. Its time to face facts... no release of the installer is going to be perfect... nor can the release team test all potential hardware combinations. Update isos are not a silver bullet, especially in an aggressive development model where the kernel itself could see hardware-specific regressions with every update. Can you be so very sure that a respin which includes an updated kernel will install without issue on my systems, when the original installer had no problem at all?
Don't get me wrong, I'm all for seeing respins produced in a more formalized way, for a vareity of reasons, by people who are willing to do the work to maintain them and to run point for bug reports that are specific to the respins which are not part of the original installer. But do not make the mistake of overselling the respins as a solution for all the community's ills. Each re-spin will have its own unique problems and regressions which will need manpower to track in bugzilla. Demanding that the fedora core release team build and maintain respins without understanding the amount of effort it would take for them to competently track bugs across the different isosets is unjustified. Sadly, the community so far who have stepped up to produce respins have not been communicating with Fedora representatives, who have made an effort to start a dialog on how to formalize the effort. The work on respins doesn't stop when the respin is built and the torrent file is made public... issues with the respins must be tracked and that will require an additional time commitment.
Would you also count the work going on with Kadischi as being an RHEL inspired subproject? http://fedoraproject.org/wiki/Kadischi I wouldn't.
It looks like a badly-need catch up to all the other popular distributions to me.
What is your point? I notice you didn't deny that this project was a response to "community interest". If you want to see it catch-up faster.. then get invovled. If you want to see it fail so that you can continue to complain about how Fedora doesn't have a livecd.. then don't get invovled. Either way the fate of this piece Fedora technology is primarily in the hands of community contributiors and not a RHEL directive. Whether or not prefer to be an active supporter of those community contributors or whether you'd prefer to sit on the sidelines and say their contributions aren't valued because they are playing catch-up is up to you.
-jef
On Sun, 2006-01-22 at 13:41, Jeff Spaleta wrote:
There is absolutely no garuntee that respins produced ANYWHERE by ANYONE will not introduce a DIFFERENT set of problems for a different set of hardware.
Yes, that's why you do it over after such problems are reported until the problem reports stop.
Would you also count the work going on with Kadischi as being an RHEL inspired subproject? http://fedoraproject.org/wiki/Kadischi I wouldn't.
It looks like a badly-need catch up to all the other popular distributions to me.
What is your point?
Sort of like 'have you stopped beating your wife?'.
I notice you didn't deny that this project was a response to "community interest".
Saying that they respond to community interest slower than the other popular distributions is damning them with faint praise whether you say it your way or mine.
On 1/22/06, Les Mikesell lesmikesell@gmail.com wrote:
Saying that they respond to community interest slower than the other popular distributions is damning them with faint praise whether you say it your way or mine.
Again wtf is your point? how exactly could have this subproject moved faster? Please look over the fedora-livecd-list archives and tell me with specificity where in that discussion, technical decisions could have been made faster.
-jef
On Sun, 2006-01-22 at 14:17, Jeff Spaleta wrote:
On 1/22/06, Les Mikesell lesmikesell@gmail.com wrote:
Saying that they respond to community interest slower than the other popular distributions is damning them with faint praise whether you say it your way or mine.
Again wtf is your point? how exactly could have this subproject moved faster? Please look over the fedora-livecd-list archives and tell me with specificity where in that discussion, technical decisions could have been made faster.
I have no specific complaint about any subproject or the speed that they do or don't complete projects. I'm just pointing out that other distributions seem to have considered it mainstream need instead of waiting for someone else to do it somewhere else.
Les Mikesell wrote:
On Sun, 2006-01-22 at 14:17, Jeff Spaleta wrote:
On 1/22/06, Les Mikesell lesmikesell@gmail.com wrote:
Saying that they respond to community interest slower than the other popular distributions is damning them with faint praise whether you say it your way or mine.
Again wtf is your point? how exactly could have this subproject moved faster? Please look over the fedora-livecd-list archives and tell me with specificity where in that discussion, technical decisions could have been made faster.
I have no specific complaint about any subproject or the speed that they do or don't complete projects. I'm just pointing out that other distributions seem to have considered it mainstream need instead of waiting for someone else to do it somewhere else.
You do realise that Fedora Project funded this effort as a bounty. Dont you?. Refer to http://fedoraproject.org/wiki/FedoraBounties for the details.
On 1/22/06, Les Mikesell lesmikesell@gmail.com wrote:
I have no specific complaint about any subproject or the speed that they do or don't complete projects. I'm just pointing out that other distributions seem to have considered it mainstream need instead of waiting for someone else to do it somewhere else.
And other distributions which pride themselves on being a community project haven't considered a livecd important and have in fact waited for someone else to do it somewhere else. http://www.debian.org/CD/faq/#live-cd
Or has Debian, the gold standard in community driven linux distribution, become irrelevant in your opinion?
Lets all remember kids that knoppix was not part of any official debian initiative and is not formally a part of the debian project at this very moment in time. It along with all the other Debain based livecd's are in fact seperately maintained offshoot projects that live outside of the formal Debian project structure. And there is absolutely nothing wrong with that. In fact there are a lot of Debian offshoots which are not formally a part of Debian, which are popular and make up a large Debian-based ecosystem. Ubuntu itself is actually an offshoot of Debian which leverages the work done by larger community of Debian contributors, but isn't formally a part of the Debian project.
I really don't see why its not acceptible for Fedora to have a similar ecosystem of competing projects which feedback ideas across the ecyosystem over time.. that spreads the development workload around among community members who are willing to suck it up and get the work done to meet specific needs... instead of idly complaining about the allocation of Core's finite resources to satify everyone's infinitely diverse desires.
-jef
On Sun, 2006-01-22 at 14:41 -0500, Jeff Spaleta wrote:
On 1/22/06, Les Mikesell lesmikesell@gmail.com wrote:
Can we have a show of hands from the people who think a new user should have to download an iso image full of bugs that have been fixed for most of a year that may prevent the install from working and that will require a gig or so of additional downloading to fix if the install succeeds?
I think what he is saying here is not only relevant to installer problems. Downloading several hundreds of megs of updates on every fresh install can get very tiresome, time consuming and costly.
Using the updated isos I save about 20-40min on each install here, multiply that with the ~50 machines running them. I think I saved some time here, and time is often important.
There is absolutely no garuntee that respins produced ANYWHERE by ANYONE will not introduce a DIFFERENT set of problems for a different set of hardware. Telling everyone to go out and grab a fc4.2 which uses a different installer codebase and an different installer kernel..
Alternatives is good. This is what the whole linux thing is about.
One wouldn't have to mirror all releases of course. Say the original FC4 and the latest respin. One of those should work for most people. The fedora download page should also make it clear that FC4 is the best tested one, but the FC4.* has fixed some bugs and contain newer packages BUT may contain new bugs.
is asking for a whole different set of bugs... bugs that have recieved far less testing in terms of hardware coverage than the original installer did. Just because its a newer version doesn't mean there will not be regressions for some hardware.. as we see with update kernels... we will see it with update installer isos as well.
Since anaconda is not updated at all in FC4, I'd say fixing the few bugs there would be pretty unlikely to make new ones.
The alternative of saying "wait 6 months, then try our new version" or "you need to use the aging FC3" is not a good solution either.
What about us few who have hardware not yet supported by the original FC4 kernel? Sorry, paperweight. I have installed on several servers here that do not work with the original FC4 kernel due to things like sata/scsi controller too new.
Its time to face facts... no release of the installer is going to be perfect... nor can the release team test all potential hardware combinations.
Agreed.
Most package updates don't break installs, yes you can do something wrong in the spec file for the rpm.
A few can, such as kernel and anaconda. Isn't it worth it to pursue these?
After all, kernel bugs should appear as much in a pre installed as in a new install. And in any case should be fixed one way or another.
Anaconda bugs should be pretty simple fixes since it probably wouldn't need new features / bells & whistles.
Update isos are not a silver bullet, especially in an aggressive development model where the kernel itself could see hardware-specific regressions with every update. Can you be so very sure that a respin which includes an updated kernel will install without issue on my systems, when the original installer had no problem at all?
Updated isos are nowhere near a perfect solution.
But I'd say they are helping.
Don't get me wrong, I'm all for seeing respins produced in a more formalized way, for a vareity of reasons, by people who are willing to do the work to maintain them and to run point for bug reports that are specific to the respins which are not part of the original installer.
Good. But a little cooperation would be nice.
But do not make the mistake of overselling the respins as a solution for all the community's ills. Each re-spin will have its own unique problems and regressions which will need manpower to track in bugzilla.
Indeed.
Demanding that the fedora core release team build and maintain respins without understanding the amount of effort it would take for them to competently track bugs across the different isosets is unjustified.
I'd be glad to help out with that, but I would certainly appreciate help from the official team when I get stuck with a problem I cannot solve myself.
Sadly, the community so far who have stepped up to produce respins have not been communicating with Fedora representatives, who have made an effort to start a dialog on how to formalize the effort.
The one mail I received from Rahul did not strike me as a "I'll help you" mail, but rather a general suggestion. The other one we have found out never reached me.
It was never the lack of willingness.
The work on respins doesn't stop when the respin is built and the torrent file is made public... issues with the respins must be tracked and that will require an additional time commitment.
I have gotten a few bugreports about FC4.1 but none about FC4.2. Excepting of course the installer-boot-menu-needs-garbage and swap-label-contains-garbage bugs. Both of witch could have been fixed had there been official updated rpms.
On 1/23/06, Hans Kristian Rosbach hk@isphuset.no wrote:
I think what he is saying here is not only relevant to installer problems. Downloading several hundreds of megs of updates on every fresh install can get very tiresome, time consuming and costly.
I'm not disagreeing with the idea of update respins. What I am saying is there is absolutely no reason that respins can not be maintained long term by community to spread the manpower around. If the release team say they are not prepared for the potential maintainence burden, that should be respected.. even if they are wrong about what the burden will actually be. The way forward is community effort like yours. The questions before us is identifying the steps to formalized your effort as part of the Fedora project over time.
Alternatives is good. This is what the whole linux thing is about.
Alternative distributiuons get branded with different distribution trademarks in the wider scope of linux in general. I absolutely want to see a full ecosystem of Fedora based alternatives out there, trying out ideas, which may feedback into the Fedora project or may not. How many debian based alternatives are there right now which aren't formally part of the debian project itself? I am hopeful that the new trademark policy will make it much easier for community "alternatives" to say "based on Fedora" without fear of confusion. What I don't want to see is for community to sit on their hands waiting for every good idea to be officially adopted before willing to try it. The respins are not a bad idea, and I absolutely appreciate the fact that you took the initiative to set this up to provide a useful service to other community members.
One wouldn't have to mirror all releases of course. Say the original FC4 and the latest respin.
I'm not going to get into how much content the mirrors are willing to mirror... I simply don't know what is appropriate here.
The fedora download page should also make it clear that FC4 is the best tested one, but the FC4.* has fixed some bugs and contain newer packages BUT may contain new bugs.
I think over time, getting discussion of a "blessed" respin on the download page is not impossible. But I think the extent the Fedora project officially advertises a respin effort will be incremental in nature.
Since anaconda is not updated at all in FC4, I'd say fixing the few bugs there would be pretty unlikely to make new ones.
That's a discussion you should have with the anaconda developers and the fedora release team... since they have the most experience with regard to how sensitive the codebase is.
The alternative of saying "wait 6 months, then try our new version" or "you need to use the aging FC3" is not a good solution either.
There is no perfect solution. Again I'm not saying that respins are an inherently bad. And I very much appreciate your effort. I've pointed people to the respins torrent from irc several times. What we need to do now is figure out the steps to take your effort and begin to formalize it and identify how it can be placed in the Fedora project umbrella officially. It seems Rahul is volunteering to help you with that process.
A few can, such as kernel and anaconda. Isn't it worth it to pursue these?
Its worth pursuing all bugs, the problem is one of manpower. Without more community involvement there isn't enough manpower. Again, the respin effort that you are making I fully support. What I can not support is the demand that the Core release team take on the burden of maintaining respins on their own.
Anaconda bugs should be pretty simple fixes since it probably wouldn't need new features / bells & whistles.
A discussion you'll have to have with the anaconda developers and the release team, since they arguable have the most experienced with these issues.
Good. But a little cooperation would be nice.
I think Rahul is trying to move forward with cooperation.
I'd be glad to help out with that, but I would certainly appreciate help from the official team when I get stuck with a problem I cannot solve myself.
You'll have to cross that bridge when you come to it. Let me remind you in an effort to encourage you that both the ppc and x86_64 bit releases were driven by a non-trivial amount of work by the community with some help by developers in Core. I think willingness to dig in and get your hands dirty before turning to the Core developers for help counts for a lot.
The one mail I received from Rahul did not strike me as a "I'll help you" mail, but rather a general suggestion. The other one we have found out never reached me.
That is an unfortunate turn of events.
It was never the lack of willingness.
Excellent! Hopefully you and Rahul can start a dialog about fedora's relationship with the respin effort.
I have gotten a few bugreports about FC4.1 but none about FC4.2.
I'm glad to here it, but you also don't know how heavily used the fc4.2 sets are so its diffcult to gauge the extent of any new problems that crop up. Sometimes the tracker for the isos are down when people in #fedora attempt to use them, so I don't have an accurate impression from the irc discussion as to how often people are picking these up for installs.
-jef
On Mon, 2006-01-23 at 09:45, Jeff Spaleta wrote:
One wouldn't have to mirror all releases of course. Say the original FC4 and the latest respin.
I'm not going to get into how much content the mirrors are willing to mirror... I simply don't know what is appropriate here.
I'm curious as to why the mirror issue might be a factor for fedora if they don't have a problem with respins of RedHat releases.
Les Mikesell wrote:
On Mon, 2006-01-23 at 09:45, Jeff Spaleta wrote:
One wouldn't have to mirror all releases of course. Say the original FC4 and the latest respin.
I'm not going to get into how much content the mirrors are willing to mirror... I simply don't know what is appropriate here.
I'm curious as to why the mirror issue might be a factor for fedora if they don't have a problem with respins of RedHat releases.
How about you ask in places where people can answer you?. fedora-devel list would be a start.
On 1/23/06, Les Mikesell lesmikesell@gmail.com wrote:
On Mon, 2006-01-23 at 09:45, Jeff Spaleta wrote:
One wouldn't have to mirror all releases of course. Say the original FC4 and the latest respin.
I'm not going to get into how much content the mirrors are willing to mirror... I simply don't know what is appropriate here.
I'm curious as to why the mirror issue might be a factor for fedora if they don't have a problem with respins of RedHat releases.
The good folks at Red Hat don't have a problem with Debian, they don't have a problem with Mandriva, they don't have a problem with FreeBSD, NetBSD, or OpenBSD. I'm pretty sure they privately have a problem with CentOS and WBEL, but they'll never tell. Despite all these "non-problems" they're not hosting downloads of any of this software. I wonder why that is. Could it be because they're not Fedora or Red Hat projects?
Not having a problem with someone else doing something doesn't mean you're throwing open the gates to hosting their project on your own servers. Mirrior maintainers signed up to host Fedora Core and possibly Fedora Extras (and associated updates). They didn't sign up to host someone else's remade ISOs. Now, if they choose to do so, that's their own lookout, they're under no obligation to whatsoever.
-- Chris
"I trust the Democrats to take away my money, which I can afford. I trust the Republicans to take away my freedom, which I cannot."
On Mon, 2006-01-23 at 10:45 -0500, Jeff Spaleta wrote:
On 1/23/06, Hans Kristian Rosbach hk@isphuset.no wrote:
I think what he is saying here is not only relevant to installer problems. Downloading several hundreds of megs of updates on every fresh install can get very tiresome, time consuming and costly.
I'm not disagreeing with the idea of update respins. What I am saying is there is absolutely no reason that respins can not be maintained long term by community to spread the manpower around. If the release team say they are not prepared for the potential maintainence burden, that should be respected.. even if they are wrong about what the burden will actually be. The way forward is community effort like yours. The questions before us is identifying the steps to formalized your effort as part of the Fedora project over time.
Glad to hear you agree :)
*snip a whole lot*
I have gotten a few bugreports about FC4.1 but none about FC4.2.
I'm glad to here it, but you also don't know how heavily used the fc4.2 sets are so its diffcult to gauge the extent of any new problems that crop up. Sometimes the tracker for the isos are down when people in #fedora attempt to use them, so I don't have an accurate impression from the irc discussion as to how often people are picking these up for installs.
As for numbers, the tracker has registered about 600 complete downloads. In addition there are those who don't download all the cds, some use trackerless clients, "broken" clients and some just don't get counted due to communications failing around the time of completion.
Hopefully a fair lot of those did install.
As for currently registered seeds: FC4.2 i386 - 5 seeds FC4.1 i386 - 1 seed FC4.2 x86-64 - 2 seeds
I will follow Rahul's advice and try to make it come under the Fedora umbrella.
-HK
Les Mikesell wrote:
On Sat, 2006-01-21 at 01:21, Rahul Sundaram wrote:
Very unlikely. RHEL is too boring for a community to participate. It is almost stagnant which is great for enterprises but not for the community. If RHEL would have been suitable for everyone Red Hat wouldnt started Fedora or vice versa.
Your argument would be more convincing if you could point out some directions that FC has taken that are more in the community interest than in building the next RHEL. Can you?
I already explained several occasions before. Packages have been moved from Fedora Core to Extras even though RHEL needed it. Packages have been introduced first in RHEL because it fit first into the enterprise space better. Packages are there in Fedora Extras even it would never go into RHEL because the community enabled it to happen. Packages are being maintained in Fedora Legacy long after Red Hat stopped doing it because the community again did the work. They have helped in documentation, bug triaging, websites, infrastructure and so on. Fedora is beyond just Fedora Core.
Rahul Sundaram wrote:
HI
This is total double talk. If Fedora Core is not a testing ground for RHEL features why exactly does it exist? What you are saying makes no sense at all.
Mission to provide a community oriented completely Free and open source operating system.
This answer is perhaps not incorrect, but is at least incomplete, by comparison to what you stated in your long missive. AIUI, there are four (at least) reasons Red Hat is contributing to Fedora...
(1) community good will (2) good will among a possible customer base for RHEL (3) "free" contributions to testing, defect finding, and defect repair (4) building a base of people who are accustomed to using the Red Hat formula for install/maintenance etc, i.e. "brand recognition and familiarity"
So, not all motives are altruistic. That's ok. And I also don't understand why anyone would complain about handing over Fedora Core to a non-profit organization. AFAICT, there is N-O-T-H-I-N-G which obligates Red Hat to continue Fedora Core support at all. From the moment I first investigated it, I considered it to be a free-will kind of thing, which could stop at any moment, for any or even no reason. This change, if and when it takes place (or even if it already has) will not be any problem for me, philosophically, economically, or otherwise.
And I wish people would quit beating up Rahul, who has so far been very polite and soft-spoken.
I am put in mind of the parable of "The Man Who Gave Too Much". Once upon a time, a man of some economic means began sending Christmas Cards to his relatives, each of them containing a $100 bill inside it. He continued this way for ten years, each year sending his relatives $100 in this manner. After this time, he decided that the card was enough to express his feelings, and the next year he sent only a card. To his amazement, all his relatives were angry, demanding to know where was "their" $100.
In this case, the $100 hasn't even been withheld, only the possibility that the $100 would come from someone else, who might not be able to send the entire amount.
My $.02.
Mike
Mike McCarty wrote:
Rahul Sundaram wrote:
Mission to provide a community oriented completely Free and open source operating system.
This answer is perhaps not incorrect, but is at least incomplete, by comparison to what you stated in your long missive. AIUI, there are four (at least) reasons Red Hat is contributing to Fedora...
(1) community good will (2) good will among a possible customer base for RHEL (3) "free" contributions to testing, defect finding, and defect repair (4) building a base of people who are accustomed to using the Red Hat formula for install/maintenance etc, i.e. "brand recognition and familiarity"
So, not all motives are altruistic. That's ok.
If you look at the prior announcements it has always been open that the benefits is going to be mutual. Red Hat has used its business justification to calm down concerns that it is not merely code getting thrown over the wall with Fedora.. As long as there is a good section of the users getting benifitted from Fedora they dont have to complain about Red Hat sharing some of it. Anybody could build a commercial product out of a part of Fedora just like Red Hat has. The foundation makes everyone a equal opportunity player. Red Hat can hope to get more benefits perhaps only by participating in the process more than others. Establishment of the foundation would mean that you dont have to just trust Red Hat's word on it and it provides a legal bond that the community contributions are going to remain free (in both senses) forever among several other defined goals.
<snip>
Hey John, there is no use beating a dead horse. Once a decision is made there is never any chance it will be reversed. I have finally accepted this fact. And if I continue to use FC, I have resigned my self to live with what I am given and just be happy if it works for me, when it doesn't, I figure out how to live with what I can make work myself. It has been made very clear that this is what is expected of FC users. When something breaks, send a bugzilla if there aren't already a bunch for the same problem, then figure out how to live with problem until the package is no longer supported or possibly gets fixed. Remember we are the extended testing ground for the Enterprise products and should be resigned to be treated as such.
By the way FC5t2 seems OK so far, but I will get Ubuntu just in case the final release is as broken as FC4 was, so I have something stable.
I agree this is a dead horse but it should not be. We are either a community or we are not. Red Hat is in business and they can do with there products what they want to. But then don't ask us to put money into a Fedora Foundation where we can pay out money to be ignored.
And when we complain we get the Red Hat shuffle. Go complain on a different list so we can be ignored on two lists at the same time. I am not concerned so much with Fedora or Red Hat not be able to do what we think aught to be done but the fact that it is made obvious that they will not even consider or opinions is intolerable.
I am not going to switch distributions but I resent our opinions being ignored out of hand. And Raul that is exactly what your company is doing no matter what you say.
akonstam@trinity.edu wrote:
I agree this is a dead horse but it should not be. We are either a community or we are not. Red Hat is in business and they can do with there products what they want to. But then don't ask us to put money into a Fedora Foundation where we can pay out money to be ignored.
This Fedora Foundation changeover seems to have happened already in a sufficiently low key way that many folks were not really aware of it, including me. Then this "will you pay money" email seemed to assume that everybody would know what they were being asked to pay money for without spelling out why they might want to. Reading it, I had no idea why I would want to give what still seemed to me "RedHat" my money. Fedora Foundation Folks: many people don't seem know anything about what is going on, get the message saturated out there then talk about donations, and when talking about donations, explain why you would want to donate.
It seems that the new deal is that the Fedora Foundation will decide what is to happen with the general goals of Fedora now, so your description above is outdated now apparently. How it differs I don't really know but I gather it is meant to be a completely standalone creature made up of meriotcratic community folk.
Further nothing stops you joining in what is currently known as Fedora Legacy and backporting fixes for Fedora Core releases Redhat consider too old. You're the "community", right?
And when we complain we get the Red Hat shuffle. Go complain on a different list so we can be ignored on two lists at the same time. I am not concerned so much with Fedora or Red Hat not be able to do what we think aught to be done but the fact that it is made obvious that they will not even consider or opinions is intolerable.
I am not going to switch distributions but I resent our opinions being ignored out of hand. And Raul that is exactly what your company is doing no matter what you say.
Let's give Rahul some slack here: he is a regular visible Redhat presence on the list. I can remember a time before Rahul when I was moaning that there was no regular visible Redhat presence on the list. Just looking at the last couple of day's traffic and imagining if I had that job, I would have blown a fuse by now and be telling people to Mandriva Ubuntu to where the sun don't shine.
As to opinions being ignored out of hand, this is a wide open Internet Mailing List, some of the (I mean this only in the general) opinions here will only be worth ignoring out of hand. Yeah okay like mine.
-Andy
[All opinions are my own]
Andy Green wrote:
akonstam@trinity.edu wrote:
I agree this is a dead horse but it should not be. We are either a community or we are not. Red Hat is in business and they can do with there products what they want to. But then don't ask us to put money into a Fedora Foundation where we can pay out money to be ignored.
This Fedora Foundation changeover seems to have happened already in a sufficiently low key way that many folks were not really aware of it, including me. Then this "will you pay money" email seemed to assume that everybody would know what they were being asked to pay money for without spelling out why they might want to. Reading it, I had no idea why I would want to give what still seemed to me "RedHat" my money. Fedora Foundation Folks: many people don't seem know anything about what is going on, get the message saturated out there then talk about donations, and when talking about donations, explain why you would want to donate.
It seems that the new deal is that the Fedora Foundation will decide what is to happen with the general goals of Fedora now, so your description above is outdated now apparently. How it differs I don't really know but I gather it is meant to be a completely standalone creature made up of meriotcratic community folk.
Ok. Time to provide some details then... When Red Hat originally launched Red Hat Linux project which later merged with the earlier fedora.us effort into a Fedora Project, Red Hat announced its intention early about opening up its distribution into a community oriented development instead of merely providing a end product under a open source license. The community has very high expectations from Red Hat. This is double edged sword. While it announced its intention, everybody wanted immediate action. It as not originally possible for the community to actually contribute anything beyond bug fixes and stuff that it could have done anyway with Red Hat Linux. It took years for the project to get off its ground with opening up cvs, Fedora Extras, Fedora Legacy and so on. Still there were some concerns with Red Hat as a commercial entity holding all the copyrights etc to the project. While capitalism is all fine and dandy, some of the volunteers within the project are inherently skeptical about commercial entities managing a community project. They believe they will get themselves screwed by it. Some of the concerns are merely perceptions and other genuine. What Red Hat has aimed to do with the Fedora foundation is provide a independent entity that holds all the copyrights, trademarks etc for its own and other community contributions that is different from its commercial open source product - Red Hat Enterprise Linux. The foundation manages and directs all the Fedora projects going forward. It can bless, reject Fedora projects, both established and proposed ones. It can pull of stale projects or entities within any of the Fedora project on extreme occasions (hopefully not required ever). Red Hat by the virtue of its contributions steers ahead one of the Fedora Foundation managed sub projects - Fedora Core. Rest of the Red Hat announced its intention to create such a foundation a while back. Some updates and goals are available at http://fedoraproject.org/wiki/Foundation. While Red Hat has accomplished some of the goals it doest intend to do a more public announcement until all the deliverables are met. It can be said that it learned its lessen against setting too high expectations as it did originally with by prematurely announcing Fedora project. This time around its going through the processes within a relatively small circle of high profile open source contributors especially those with a established history of building such communities. When Red Hat passes the litmus test of acceptance from this group and establishes various other processes it can shoot for a more public announcement. Till then it is sort of a public secret. The Foundation as a legal entity already this though. A board that is community based is in the road map as a high priority item.
Why does Fedora Foundation need money?
Red Hat's revenue source primarily is RHEL. Fedora from a business point of view does not provide any direct revenue being a gratis software. In fact in by itself it reduces it. Red Hat justifies this by the value of its presence within the community and the ability to build a similar code base between Fedora and RHEL which enables community feedback on Fedora to be useful for both Fedora and RHEL. Red Hat intents to continue providing significant funding for Fedora but if it does that is a hindrance to establishes 501(3) status for Fedora foundation as required by the law. 501(3) status is important if the community provides funds needs to be tax deductible. These funds to the Fedora Foundation will establish it to providing funding to projects and community members that is not directly dependent on RHEL. By providing this financial freedom, Red Hat can avoid getting into perceptions of using Fedora as a test bed by enabling the foundation to explore and establish other community projects within the Fedora space. This includes for example the Fedora patents commons project. The potential for this is expected to be immense but we need answers from the community that Elliot Lee asked for.
Would you be willing to donate money to the Fedora Foundation? Why/why not?
Do you live in a country where donations to non-profit organizations may be tax deductible? Does the tax deductability of your contributions affect your willingness to donate?
How does Red Hat's involvement in and funding of Fedora affect your willingness to donate to the Foundation? Your willingness to participate in Fedora?
You can send answers as well as more questions to Elliot Lee (sopwith AT Red Hat.com).
The reason this was asked to each list member individually is to enable them to respond privately about this financial issue which might be too sensitive for them to reveal publictly on the list. Red Hat will not ask for money for Fedora project. It will continue to do what it can provide freely within its financial limits. The Fedora foundation also cannot get any funds at this point. There is a need for the above to be answered by the community to establish the process moving forward to the board as well as the counsel. I request the community at large to take these details into consideration and respond carefully to Elliot. Your responses will affect the future of this project. Feel free to pass on this message to anyone interested in Fedora.
Let's give Rahul some slack here: he is a regular visible Redhat presence on the list. I can remember a time before Rahul when I was moaning that there was no regular visible Redhat presence on the list.
I am not sure everybody knows this but I have been a contributor to Fedora long before I joined Red Hat (can be said that RH hired me because of that though I dont work directly on Fedora for them now). I intend to continue working for the community and getting sponsored by Red Hat in a way.
Just looking at the last couple of day's traffic and imagining if I had that job, I would have blown a fuse by now and be telling people to Mandriva Ubuntu to where the sun don't shine.
As to opinions being ignored out of hand, this is a wide open Internet Mailing List, some of the (I mean this only in the general) opinions here will only be worth ignoring out of hand. Yeah okay like mine.
-Andy
I would request active list members to sign up a report writer and produce weekly reports on this list covering major discussions in a concise summary at http://fedoraproject.org/wiki/Projects/WeeklyReports. It would make my intention to provide a sounding board and would open better communication channels between the community at large and developers. It is increasing hard for me to respond questions in this list due to the amount of hostility expressed and that I have to actually work on projects rather than spending all my volunteer time merely answering questions in this list. Red Hat presence on this list is low due to the high amount of traffic and not that it doesnt want to listen. If you bring up important development discussions to fedora-devel list through detailed proposals and file bug reports and feature enhancements in http://bugzilla.redhat.com, Red Hat will be able to respond efficiently compared to the responses here on the list. Thank you.
At 1:16 PM +0530 1/21/06, Rahul Sundaram wrote:
...Red Hat by the virtue of its contributions steers ahead one of the Fedora Foundation managed sub projects - Fedora Core. ...
Red Hat's contributions of money, or of people, or both? Just asking for clarity. I expect that the answer is "both, and people cost money". ____________________________________________________________________ TonyN.:' mailto:tonynelson@georgeanelson.com ' http://www.georgeanelson.com/
Tony Nelson wrote:
At 1:16 PM +0530 1/21/06, Rahul Sundaram wrote:
...Red Hat by the virtue of its contributions steers ahead one of the Fedora Foundation managed sub projects - Fedora Core. ...
Red Hat's contributions of money, or of people, or both? Just asking for clarity. I expect that the answer is "both, and people cost money". ____________________________________________________________________ TonyN.:' mailto:tonynelson@georgeanelson.com ' http://www.georgeanelson.com/
Yes, it's both, and don't forget the hardware, services, technology, and other resources. Red Hat's contributions are massive, even outside of Fedora Core.
Hi
I agree this is a dead horse but it should not be. We are either a community or we are not. Red Hat is in business and they can do with there products what they want to. But then don't ask us to put money into a Fedora Foundation where we can pay out money to be ignored.
And when we complain we get the Red Hat shuffle. Go complain on a different list so we can be ignored on two lists at the same time. I am not concerned so much with Fedora or Red Hat not be able to do what we think aught to be done but the fact that it is made obvious that they will not even consider or opinions is intolerable.
I am not going to switch distributions but I resent our opinions being ignored out of hand. And Raul that is exactly what your company is doing no matter what you say.
A community oriented project does not mean that that the community demands things and it gets done. Red Hat does not have unlimited resources for maintaining updates forever for something thats given away for free.
See fedora-devel list discussions at https://www.redhat.com/archives/fedora-devel-list/2006-January/msg00744.html. In particular https://www.redhat.com/archives/fedora-devel-list/2006-January/msg00961.html
It just means that the community can participate and improve things. If you care about improving the update lifecycle, you should join the Fedora Legacy project
http://legacy.fedoraproject.org
From: "Rahul Sundaram" sundaram@redhat.com
Hi
I I don't want to join it for the sake of a month; I wish to go from three to five. I prefer not to rejig my yum configurations and/or up2date configurations, and I don't wish to go to four at this late state.
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
"responsibility to contribute."
Jo and I don't contribute? You gotta be joking. We've both been helping out on these lists for years (with a minor break or two when RH left us).
If you would like to get updates from Fedora Legacy project for a longer duration you will have to contribute to that project specifically which is what I meant earlier in context.
You're not listening, Rahul. That leads me to wonder just who constitutes the community that does the influences Fedora policies. I guess folks on this list are basically peons who are not part of the "community" or at least we're not as equal as other members.
We're aware of the policy. We disagree with it. And we're agitating for a change. It appears to affect John directly. It won't affect me until 6 test 2. But I expect I'll get plenty annoyed at that time. And if the annoyance is big enough I can go for some honest autocracy and setup with FreeBSD. (And maybe lose some of the GPL encumbrance problems.)
{^_^}
jdow wrote:
From: "Rahul Sundaram" sundaram@redhat.com
Hi
I I don't want to join it for the sake of a month; I wish to go from three to five. I prefer not to rejig my yum configurations and/or up2date configurations, and I don't wish to go to four at this late state.
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
"responsibility to contribute."
Jo and I don't contribute? You gotta be joking. We've both been helping out on these lists for years (with a minor break or two when RH left us).
If you would like to get updates from Fedora Legacy project for a longer duration you will have to contribute to that project specifically which is what I meant earlier in context.
You're not listening, Rahul. That leads me to wonder just who constitutes the community that does the influences Fedora policies. I guess folks on this list are basically peons who are not part of the "community" or at least we're not as equal as other members.
Hmm. He has suggested you go to the devel area. I don't understand why this is not a reasonable forum for discussing such "policy" related issues, myself. But if it isn't, then it isn't, and so long as there *is* a forum for that type of discussion, then I don't think anyone is being treated badly.
We're aware of the policy. We disagree with it. And we're agitating for a change. It appears to affect John directly. It won't affect me until 6 test 2. But I expect I'll get plenty annoyed at that time. And if the annoyance is big enough I can go for some honest autocracy and setup with FreeBSD. (And maybe lose some of the GPL encumbrance problems.)
I've been seriously considering Debian, actually. I've been having to administer a friend's machine, and the Debian people are pretty helpful in most respects. Debian is at least as easy to administer as FC is. And there is less pressure to "upgrade" in my experience.
But so far, I've been satisfied with Legacy. I don't need or want lots of new "features" (like LVM, or selinux, as examples). But I am interested in "critical updates" and "security fixes".
I've had some experience with FreeBSD, and with Solaris, and on the whole, I'd prefer Solaris to FreeBSD, at least for my home machine.
Mike
Hi
Hmm. He has suggested you go to the devel area. I don't understand why this is not a reasonable forum for discussing such "policy" related issues, myself. But if it isn't, then it isn't, and so long as there *is* a forum for that type of discussion, then I don't think anyone is being treated badly.
Because not all of the developers can follow discussions in this high traffic list and many of them are only subscribed to the fedora development related mailing lists. There is a specific list meant to discuss development issues. A request to update release policy is one such issue. So go ahead, write a detailed rationale and present it in fedora-devel list. Thanks
On Tue, 2006-17-01 at 13:33 +0530, Rahul Sundaram wrote:
Hi
Hmm. He has suggested you go to the devel area. I don't understand why this is not a reasonable forum for discussing such "policy" related issues, myself. But if it isn't, then it isn't, and so long as there *is* a forum for that type of discussion, then I don't think anyone is being treated badly.
Because not all of the developers can follow discussions in this high traffic list and many of them are only subscribed to the fedora development related mailing lists. There is a specific list meant to discuss development issues. A request to update release policy is one such issue. So go ahead, write a detailed rationale and present it in fedora-devel list. Thanks
Except you will get exactly the same answer, and probably from Rahul too.
Guy Fraser wrote:
On Tue, 2006-17-01 at 13:33 +0530, Rahul Sundaram wrote:
Hi
Hmm. He has suggested you go to the devel area. I don't understand why this is not a reasonable forum for discussing such "policy" related issues, myself. But if it isn't, then it isn't, and so long as there *is* a forum for that type of discussion, then I don't think anyone is being treated badly.
Because not all of the developers can follow discussions in this high traffic list and many of them are only subscribed to the fedora development related mailing lists. There is a specific list meant to discuss development issues. A request to update release policy is one such issue. So go ahead, write a detailed rationale and present it in fedora-devel list. Thanks
Except you will get exactly the same answer, and probably from Rahul too.
With detailed explanations of course.
jdow wrote:
You're not listening, Rahul. That leads me to wonder just who constitutes the community that does the influences Fedora policies. I guess folks on this list are basically peons who are not part of the "community" or at least we're not as equal as other members.
If you look at what you're asking it boils down to asking some guys paid by Redhat to spend time and effort they could spend elsewhere continuing to give their high quality security coverage to what is now an older OS, for free. I have a remote server that is FC3, it came with FC3 and the only other choice was Windows, so I am motivated to want security coverage... but... before we huff and puff about 'community' we have to realize we are demanding their work for free in such a case. Redhat can just as easily say, "look, we give you all this great stuff for free and we want the 'community' in the form of the Legacy project to chip in now so we can go on with new stuff".
We're aware of the policy. We disagree with it. And we're agitating for a change. It appears to affect John directly. It won't affect me until 6 test 2. But I expect I'll get plenty annoyed at that time. And if the annoyance is big enough I can go for some honest autocracy and setup with FreeBSD.
Yeah well this is a valid choice... the 'Fedora deal' is that we get high quality, up to date stuff, but we have to be on the dist-upgrade treadmill. If that deal doesn't suit your situation (I sure wish I had CentOS on that server instead of Fedora, but it wasn't a choice) then do consider to move. CentOS makes most sense for many Fedora people in situations where they can't be keeping up with the schedule, since it basically is Fedora with wisdom teeth, but OSS is all about using whatever is going to work for you, not for some particular supplier.
(And maybe lose some of the GPL encumbrance problems.)
I didn't really understand what you meant by this.
-Andy
Andy Green wrote:
[some stuff which I pretty much agree with]
[snip]
jdow wrote:
6 test 2. But I expect I'll get plenty annoyed at that time. And if the annoyance is big enough I can go for some honest autocracy and setup with FreeBSD.
(And maybe lose some of the GPL encumbrance problems.)
I didn't really understand what you meant by this.
FreeBSD is distributed, not under the GPL license. GPL conforms to Richard Stallman's idea about what software should be, and if you use it, then it has a tendency to invade your own software, like an infection. (You understand I'm writing this to illustrate the perspective of those who disagree with R.S.) The license for BSD does not have quite the same tendency for spreading into other people's works.
Mike
Mike McCarty wrote:
(And maybe lose some of the GPL encumbrance problems.)
I didn't really understand what you meant by this.
FreeBSD is distributed, not under the GPL license. GPL
Yes I am reasonably aware of the two licenses. But in the case of Fedora, what are the "GPL encumbrance" problems to a normal user? The lack of patented codecs? Sounds like any commercial enterprise would have to take care in that direction GPL compatability problems or no.
-Andy
On Tue, 2006-01-17 at 11:20, Andy Green wrote:
I didn't really understand what you meant by this.
FreeBSD is distributed, not under the GPL license. GPL
Yes I am reasonably aware of the two licenses. But in the case of Fedora, what are the "GPL encumbrance" problems to a normal user?
Who is a 'normal user'? If you mean one who would would like to make his own choice about paying for an enhanced product or using only what someone is willing to provide for free, the problems are the things that aren't permitted to be available as a choice for you.
The lack of patented codecs? Sounds like any commercial enterprise would have to take care in that direction GPL compatability problems or no.
A rather large number of things cannot be included as part of something with GPL components, *even if* the person who has them wants to give them away and the person who needs them has the appropriate licenses or is willing to get them.
Compare, for example, what comes built into OSX with what Linux distros can include.
From: "Andy Green" andy@warmcat.com
Andy, I am a software developer by trade. I use it to earn my daily bread and board. If I develop in an environment that involves GPL I cannot see a model that will continue to feed me and house me unless I take up a side job asking, "Do you want to supersize that, sir?" or sit at my own help desk all day instead of developing. GPL contaminates things too thoroughly. I am not a lawyer. I just read that document and basically stay away from GPL except for some recreational coding I've done.
{^_^}
jdow wrote:
Andy, I am a software developer by trade. I use it to earn my daily bread and board. If I develop in an environment that involves GPL I cannot see a model that will continue to feed me and house me unless I take up a side job asking, "Do you want to supersize that, sir?" or sit at my own help desk all day instead of developing. GPL contaminates things too thoroughly. I am not a lawyer. I just read that document and basically stay away from GPL except for some recreational coding I've done.
I also design software and hardware... but depending on the field you work in, GPL stuff can bring an awful lot of firepower to the party very cheaply. Again depending on the circumstance, the compataibility, quality, time to market and royalty-free advanatages surrounding that can overwhelm the possible competitive disadvantage of having to open some of your stuff.
The general deal is AIUI if you link with GPL'd stuff -- not LGPL'd, which will not infect what it links to but only changes to itself -- you will have to open your work. It seems that we crossed a threshold now and the signs are that if you generate kernel modules you can expect that sources will be demanded.
However, there's a lot of room in there for usermode apps that just link to LGPL's libraries (like, eg, Oracle) that you for sure won't have to open source unless you want to.
Of course in your particular field if there is no specific benefit to be had from using or being directly compatible with GPL'd stuff then it makes sense to avoid it.
But if its just that you are considering to use, say, Fedora as a platform for scripts or usermode apps then AFAIUI there is zero leakage of the GPL-ness of the OS apps up into your code and BSD or GPL makes no odds there.
-Andy
Andy Green wrote:
jdow wrote:
Andy, I am a software developer by trade. I use it to earn my daily bread and board. If I develop in an environment that involves GPL I cannot see a model that will continue to feed me and house me unless I take up a side job asking, "Do you want to supersize that, sir?" or sit at my own help desk all day instead of developing. GPL contaminates things too thoroughly. I am not a lawyer. I just read that document and basically stay away from GPL except for some recreational coding I've done.
Seconded and re-seconded.
I also design software and hardware... but depending on the field you work in, GPL stuff can bring an awful lot of firepower to the party very cheaply. Again depending on the circumstance, the compataibility, quality, time to market and royalty-free advanatages surrounding that can overwhelm the possible competitive disadvantage of having to open some of your stuff.
No, GPL forces one to open everything.
The general deal is AIUI if you link with GPL'd stuff -- not LGPL'd, which will not infect what it links to but only changes to itself -- you will have to open your work. It seems that we crossed a threshold now and the signs are that if you generate kernel modules you can expect that sources will be demanded.
LGPL is less infective, but even that gets you into being a distributor of source.
[snip]
Of course in your particular field if there is no specific benefit to be had from using or being directly compatible with GPL'd stuff then it makes sense to avoid it.
Unless one just wants to produce GPL or LGPL stuff, and not feed himself with the fruit of his labor (like Joanne and I do) it makes sense to avoid GPL altogether, end of story.
But if its just that you are considering to use, say, Fedora as a platform for scripts or usermode apps then AFAIUI there is zero leakage of the GPL-ness of the OS apps up into your code and BSD or GPL makes no odds there.
Erm, if you compile and link any program which runs on Linux, you end up with at least LGPL stuff in your code.
Mike
From: "Mike McCarty" mike.mccarty@sbcglobal.net
Andy Green wrote:
jdow wrote:
Andy, I am a software developer by trade. I use it to earn my daily bread and board. If I develop in an environment that involves GPL I cannot see a model that will continue to feed me and house me unless I take up a side job asking, "Do you want to supersize that, sir?" or sit at my own help desk all day instead of developing. GPL contaminates things too thoroughly. I am not a lawyer. I just read that document and basically stay away from GPL except for some recreational coding I've done.
Seconded and re-seconded.
I also design software and hardware... but depending on the field you work in, GPL stuff can bring an awful lot of firepower to the party very cheaply. Again depending on the circumstance, the compataibility, quality, time to market and royalty-free advanatages surrounding that can overwhelm the possible competitive disadvantage of having to open some of your stuff.
No, GPL forces one to open everything.
Not true. It forces you to open everything that uses ANYTHING that is itself GPL contaminated and not purely your own work. (You can dual license your own work.)
The general deal is AIUI if you link with GPL'd stuff -- not LGPL'd, which will not infect what it links to but only changes to itself -- you will have to open your work. It seems that we crossed a threshold now and the signs are that if you generate kernel modules you can expect that sources will be demanded.
LGPL is less infective, but even that gets you into being a distributor of source.
At least some source. I've read it. It left me gasping for air. It SEEMS to be intended to allow you to do an Oracle on Linux. But I am not enough of a lawyer to try. It's easier to stay away.
[snip]
Of course in your particular field if there is no specific benefit to be had from using or being directly compatible with GPL'd stuff then it makes sense to avoid it.
Unless one just wants to produce GPL or LGPL stuff, and not feed himself with the fruit of his labor (like Joanne and I do) it makes sense to avoid GPL altogether, end of story.
I've two forays into GPL so far. One is a TINY contribution to the kernel based on specialized knowledge I have. (I helped, in some small way, design the Amiga computer's partitioning scheme. I found a mis- implementation in the kernel and repaired it. Now I can read some of the old horrors I put together for testing when I was doing partitioning software for the Amiga. {^_-}) The other was a "learning and playing around tool" on Winders. I have a ham radio transceiver with fairly good remote control capabilities. So I built a remote control for it. I've sent that to a few people with source and a GPL license on it. It is entirely my work except for the libraries it calls up from Windows. So I also use modified pieces of "gadgets" I put together in it for my other work on the perfectly dual license basis for my own work.
But if its just that you are considering to use, say, Fedora as a platform for scripts or usermode apps then AFAIUI there is zero leakage of the GPL-ness of the OS apps up into your code and BSD or GPL makes no odds there.
Erm, if you compile and link any program which runs on Linux, you end up with at least LGPL stuff in your code.
This is an ongoing discussion on the kernel mailing list that pops up from time to time. It seems "nobody really knows." That makes me even MORE nervous about serious work on Linux. It's a shame, really. Somebody should draw up some REALLY clear guidelines about GPL, LGPL, and Linux. (I've a hunch I know of at least one violation in the lighting control console industry. But *I* am not about the blow the whistle on the guy. He's earned his bread as has the company that employed him to do it. The industry's so small and densely competitive within that "smallness" that open source would utterly kill the company.)
{^_^} Joanne
On Tue, 2006-01-17 at 18:27 -0800, jdow wrote:
From: "Mike McCarty" mike.mccarty@sbcglobal.net
No, GPL forces one to open everything.
Not true. It forces you to open everything that uses ANYTHING that is itself GPL contaminated and not purely your own work. (You can dual license your own work.)
I don't see what the big deal is with the GPL "encumbrance". It's just like any other license - if you can't live with its terms, just don't incorporate software licensed by it into your own work.
It's like seeing some whiz-bang control you might like to use if you're a Visual Basic programmer, except that the license for the control costs $1,000. If you don't want to pay the $1,000, you don't use the control. What's the difference?
Paul.
On Wed, 2006-01-18 at 02:13, Paul Howarth wrote:
On Tue, 2006-01-17 at 18:27 -0800, jdow wrote:
From: "Mike McCarty" mike.mccarty@sbcglobal.net
No, GPL forces one to open everything.
Not true. It forces you to open everything that uses ANYTHING that is itself GPL contaminated and not purely your own work. (You can dual license your own work.)
I don't see what the big deal is with the GPL "encumbrance". It's just like any other license - if you can't live with its terms, just don't incorporate software licensed by it into your own work.
It's like seeing some whiz-bang control you might like to use if you're a Visual Basic programmer, except that the license for the control costs $1,000. If you don't want to pay the $1,000, you don't use the control. What's the difference?
So it wouldn't bother you at all if every time someone asked a question here about a problem that exists in Linux distributions the answer was always to drop Linux and switch to OSX, Windows, Solaris or similar commercial products that include working code under licences not compatible with the GPL? "Don't use it" on the developer side means "not available on Linux" on the end user side.
Les Mikesell wrote:
On Wed, 2006-01-18 at 02:13, Paul Howarth wrote:
On Tue, 2006-01-17 at 18:27 -0800, jdow wrote:
From: "Mike McCarty" mike.mccarty@sbcglobal.net
No, GPL forces one to open everything.
Not true. It forces you to open everything that uses ANYTHING that is itself GPL contaminated and not purely your own work. (You can dual license your own work.)
I don't see what the big deal is with the GPL "encumbrance". It's just like any other license - if you can't live with its terms, just don't incorporate software licensed by it into your own work.
It's like seeing some whiz-bang control you might like to use if you're a Visual Basic programmer, except that the license for the control costs $1,000. If you don't want to pay the $1,000, you don't use the control. What's the difference?
So it wouldn't bother you at all if every time someone asked a question here about a problem that exists in Linux distributions the answer was always to drop Linux and switch to OSX, Windows, Solaris or similar commercial products that include working code under licences not compatible with the GPL?
What is it that is GPL'ed in this example? If a developer wants to use someone else's GPLed code in their product (and distribute it), they must GPL their product too. If they don't want to use anyone else's GPL'ed code, they don't.
"Don't use it" on the developer side means "not available on Linux"
on the
end user side.
No, it means "use someone else's library or write your own", just as you would in the proprietary world.
Paul.
Paul Howarth wrote:
What is it that is GPL'ed in this example? If a developer wants to use someone else's GPLed code in their product (and distribute it), they must GPL their product too. If they don't want to use anyone else's GPL'ed code, they don't.
"Use" is too loose here. You can 'use' Linux on your device, the GNU libc, all kinds of coreutils stuff, etc, etc, and your code can be a proprietary usermode app that can even link to LGPL'd libraries, and you will NOT have to "GPL your product", ie, your proprietary app sources. You only have to pass on the sources for the stuff that YOU got for free in such a case.
The specific threshold where it changes is linking in GPL code for userspace, and effectively it seems any kernelmode code unless you can get Linux and the other Linux contributors to turn a blind eye or are willing to risk a future battle over it.
But userspace apps only using LGPL stuff is a huge field with plenty of critical libs to call on, look:
# rpm -qa --queryformat "%{NAME} %{LICENSE}\n" | grep LGPL | sort | uniq
-Andy
Paul Howarth wrote:
Les Mikesell wrote:
[snip]
I don't see what the big deal is with the GPL "encumbrance". It's just like any other license - if you can't live with its terms, just don't incorporate software licensed by it into your own work.
It's like seeing some whiz-bang control you might like to use if you're a Visual Basic programmer, except that the license for the control costs $1,000. If you don't want to pay the $1,000, you don't use the control. What's the difference?
So it wouldn't bother you at all if every time someone asked a question here about a problem that exists in Linux distributions the answer was always to drop Linux and switch to OSX, Windows, Solaris or similar commercial products that include working code under licences not compatible with the GPL?
What is it that is GPL'ed in this example? If a developer wants to use
The entire interface library for C, FORTRAN, etc.
someone else's GPLed code in their product (and distribute it), they must GPL their product too. If they don't want to use anyone else's GPL'ed code, they don't.
No, you see there are parts of the library which any program needs which are loaded at run time. Even though they aren't a part of the program, the GPL claims to exert power over the program because it *might* get linked at load time with a GPL library.
Are you saying we should all re-write open(), close() etc?
"Don't use it" on the developer side means "not available on Linux"
on the
end user side.
No, it means "use someone else's library or write your own", just as you would in the proprietary world.
No, not even MicroSoft says that you can't call open() without making your code theirs. In fact, no one does that except GPL.
(I'm not claiming that open() is GPL, it might be LGPL, but even LGPL is so confusing to read that I'm not sure what it means, exactly, and it certainly has some scary language.)
Mike
Mike McCarty wrote:
No, not even MicroSoft says that you can't call open() without making your code theirs. In fact, no one does that except GPL.
(I'm not claiming that open() is GPL, it might be LGPL, but
Your complaints in the parent post do in fact seem to rely on open() being in a GPL'd library: but Standard C library open() is in glibc which is LGPL.
As I posted earlier, there is quite a large number of core libs which are LGPL, aside from glibc, and so allow a proprietary app writer to include them in his app without forcing the source to be opened.
Please run this:
# rpm -qa --queryformat "%{NAME} %{LICENSE}\n" | grep LGPL | sort | uniq
and satisfy yourself that there is a pretty rich range of libs available in LGPL on Fedora so that you can at least complain accurately about the GPL in your reply, without needing immediately disprovable strawmen like open().
Then to repeat Erwin's point, why should the people who did the work to create, by choice, a GPL'd library or app, have made another choice to facilitate your locking up a proprietary app based on or deriving from it? That is the implication of all this: "If only it was BSD licensed I could do an Apple". The library creators did the work -- they chose the license. If you don't feel able to use their work on their terms then you are free to try to find a proprietary alternative (SCO will probably sublicense a buttload of old junky Unix libs and apps on extremely proprietary terms if that's what you really want), or write your own. But please don't moan about the terrible consequences of the GPL when this is a deliberate feature wielded by the coders that chose the license, it just sounds like you want something for nothing.
even LGPL is so confusing to read that I'm not sure what it means, exactly, and it certainly has some scary language.)
Please read what you wrote above and reconsider if you should be lecturing the list on this.
-Andy
On Wed, 2006-01-18 at 16:37, Andy Green wrote:
Then to repeat Erwin's point, why should the people who did the work to create, by choice, a GPL'd library or app, have made another choice to facilitate your locking up a proprietary app based on or deriving from it?
Once the first little bit is GPL'd, no one has that choice if they want to contribute. For a very large part of the code base, I'd guess that applying the GPL restrictions was not the author's first choice. And from the dual-licensed items like perl, you can tell that some authors go well out of their way to ensure that others do not have their choices taken away.
But please don't moan about the terrible consequences of the GPL when this is a deliberate feature wielded by the coders that chose the license, it just sounds like you want something for nothing.
That would be a fair argument if every person who contributed code were allowed to choose whether to encumber his portion with the GPL or not. That isn't the case, and that's the part of the complaint.
Les Mikesell wrote:
On Wed, 2006-01-18 at 16:37, Andy Green wrote:
Then to repeat Erwin's point, why should the people who did the work to create, by choice, a GPL'd library or app, have made another choice to facilitate your locking up a proprietary app based on or deriving from it?
Once the first little bit is GPL'd, no one has that choice if they want to contribute. For a very large part of the code base, I'd guess that applying the GPL restrictions was not the author's first choice. And from the dual-licensed items like perl, you can tell that some authors go well out of their way to ensure that others do not have their choices taken away.
Fine... but I wondered "why should the people who did the work to create, by choice, a GPL'd library or app, have made another choice to facilitate your locking up a proprietary app based on or deriving from it? ". I couldn't plug your answer into that.
Fact is a GPL project is GPL'd. One can start a new competing project on a more liberal license. That doesn't seem to happen much though, because the best that can be expected from all that struggle to duplicate an existing GPL project in, say, BSD is that the GPL guys will attract patches and improvements from users all around, while the new BSD project will attract proprietary users who will not give back.
But please don't moan about the terrible consequences of the GPL when this is a deliberate feature wielded by the coders that chose the license, it just sounds like you want something for nothing.
That would be a fair argument if every person who contributed code were allowed to choose whether to encumber his portion with the GPL or not. That isn't the case, and that's the part of the complaint.
You need to squint a little and see it from the project's POV. They give this great stuff out for free on the basis you can't lock it up and you can't distribute improvements without giving out the improved source, which usually means giving improved source to the project. That's the "GPL deal". Now you consider only your modifications and groan that they have to be GPL'd also, well that's true but balancing this is the huge benefit of getting given the whole project you modified in the first place. You can't just look at the bit of work you chipped in and mourn the opening of that without also looking at the huge boost you were given -- all the work done by others conscious they were doing it as part of the GPL deal -- in order to have something to modify in the first place.
That's why these complaints seem really churlish to me, ignoring the boon to you and only complaining about your contribution back.
-Andy
On Wed, 2006-01-18 at 17:35, Andy Green wrote:
That's why these complaints seem really churlish to me, ignoring the boon to you and only complaining about your contribution back.
I don't recall mentioning any contributions or lack thereof. My complaint is as an end user, not being able to have certain things in addition to the GPL'd parts for the incremental cost of adding them. Or more realistically, being forced to use Windows or OSX for many tasks because the functionality can't legally be added to Linux. I realize that the choices have already been made - I just don't understand why anyone else thinks it is a good thing.
On Wed, 2006-18-01 at 17:52 -0600, Les Mikesell wrote:
My complaint is as an end user, not being able to have certain things in addition to the GPL'd parts for the incremental cost of adding them. Or more realistically, being forced to use Windows or OSX for many tasks because the functionality can't legally be added to Linux. I realize that the choices have already been made - I just don't understand why anyone else thinks it is a good thing.
Perhaps this is simply a matter of society being used to doing business in one particular way, and not able to fathom that there may actually be a better alternative.
Now, I understand that "better" is subjective. But, again, it's just a matter of not having tried anything else. Nothing remains the same forever, no matter how hard someone may try to keep it that way.
For the record, I like the GPL and everything it represents.
That's about all I'll say on this subject.
Regards,
Ranbir
Les Mikesell wrote:
That's why these complaints seem really churlish to me, ignoring the boon to you and only complaining about your contribution back.
I don't recall mentioning any contributions or lack thereof. My
Man your memory must be failing you. Here you go:
Once the first little bit is GPL'd, no one has that choice if they want to contribute. For a very large part of the code base, I'd
complaint is as an end user, not being able to have certain things in addition to the GPL'd parts for the incremental cost of adding them. Or more realistically, being forced to use Windows or OSX for many tasks because the functionality can't legally be added to Linux. I realize that the choices have already been made - I just don't understand why anyone else thinks it is a good thing.
I did my best to explain to you the point of view of a GPL'd project on this matter and compared it to the experience of an equivalent liberally licensed project without the viral license. If it still doesn't make any sense then I guess you will have to remain puzzled.
-Andy
On Wed, 18 Jan 2006, Les Mikesell wrote:
I don't recall mentioning any contributions or lack thereof. My complaint is as an end user, not being able to have certain things in addition to the GPL'd parts for the incremental cost of adding them. Or more realistically, being forced to use
As an end user, you don't have any encumbrance problem. End users can mix and match as they see fit. Apparently you are not just an end user, you distribute.
If you *really* want to distribute proprietary software that needs GPL software, here is what you do: Get the *documentation* on the GPL software. Write your own code using that documentation. Send your binaries to your customers along with documentation telling them about the GPL stuff they need. Since you never have the GPL software, you do not have to accept the license. In principle, your customers are affected by the license, but since they can't distribute at all, that doesn't matter.
Debugging is left as an exercise for the reader.
Andy Green wrote:
[snip]
That's why these complaints seem really churlish to me, ignoring the boon to you and only complaining about your contribution back.
One issue with this is in deployment in concert with development. Let's say that a company is going to build a new device, and wants to put an OS on an embedded computer, and develop some applications to run on it. This company does not want to divulge its trade secrets to the whole world for copying. Special techniques which may, for example, allow a cell phone to use lower power and still achieve good range, hence not cooking your brain.
Do they chose to deploy Linux+GNU? They do not. Because if they do, then the LGPL will cause them to have to supply, along with the telephone, the source to their product, and give away their trade secrets, and let other people modify and distribute their source. And even if they were willing to do that, they may want to purchase a license for the OMAP (a part of the mobile phone protocol stack) library, rather than redevelop the wheel, thus saving money for themselves, saving money for the customer, saving time to market, and perhaps even making the product possible, given the manpower available.
Oh, but the third party library developer for the OMAP doesn't want his library given away. The good ol' LGPL steps right in and says "YOU CAN'T DO THAT!"
So, instead, they choose to use Solaris, making the new lower- power cellphone possible, and cheaper, and everybody wins.
Er, except Linux deployment. Which suffers.
This example is not as contrived as it sounds. The names have been changed to protect the innocent, but the story is essentially what happened in 2000 or so when I recommended not to use Linux, but to use Solaris in a new VOIP project at Alcatel, which now as a consequence uses Solaris and not Linux.
Mike
Mike McCarty wrote:
That's why these complaints seem really churlish to me, ignoring the boon to you and only complaining about your contribution back.
One issue with this is in deployment in concert with development. Let's say that a company is going to build a new device, and wants to put an OS on an embedded computer, and develop some applications to run on it. This company does not want to divulge its trade secrets to the whole world for copying. Special techniques which may, for example, allow a cell phone to use lower power and still achieve good range, hence not cooking your brain.
Do they chose to deploy Linux+GNU? They do not. Because if they do, then the LGPL will cause them to have to supply, along with
Please read
https://www.redhat.com/archives/fedora-list/2006-January/msg02996.html
Just dynamically link to the LGPL'd libs.
-Andy
Andy Green wrote:
Please read
https://www.redhat.com/archives/fedora-list/2006-January/msg02996.html
Just dynamically link to the LGPL'd libs.
I guess that you are not very well acquainted with embedded development. Dynamic linking and virtual memory are real no-nos in real-time work. You really really really need statically linked images, and a static memory map with no paging.
Mike
Mike McCarty wrote:
Andy Green wrote:
Please read
https://www.redhat.com/archives/fedora-list/2006-January/msg02996.html
Just dynamically link to the LGPL'd libs.
I guess that you are not very well acquainted with embedded development. Dynamic linking and virtual memory are real
Actually I do embedded development for a living.
no-nos in real-time work. You really really really need statically linked images, and a static memory map with no paging.
Here in 2005 we have lovely chips like the 180MHz Atmel AT91RM9200, which I am using with 8MB of flash and 16MB of 90MHz SDRAM. that chip has dual 16kByte cacahes, but it's midrange nowadays for embedded. People are using embedded Pentiums, 600MHz ARMs and so on.
I have a full Linux distro in there with RPM, gdb and it is all dynamically linked, because once you have multiple apps running, running them all with static linked libs is stupidly wasteful of memory.
Don't you agree?
-Andy
On Thu, 2006-01-19 at 03:08 -0600, Mike McCarty wrote:
One issue with this is in deployment in concert with development. Let's say that a company is going to build a new device, and wants to put an OS on an embedded computer, and develop some applications to run on it. This company does not want to divulge its trade secrets to the whole world for copying. Special techniques which may, for example, allow a cell phone to use lower power and still achieve good range, hence not cooking your brain.
Do you really think that competitors can't do that without access to the source code? Taking that example as a case in point, spectrum analysers allow us to check what RF eminates from a device, oscilloscopes allow us to look at waveforms within a device, data analysers allow us to read digital data going about its business in a device, and so on.
Taking other examples: Anybody can buy the circuit diagram to a television set, and from that you can replicate it for less effort than the original designers. Yet TV sets are still sold, and so are the circuits. The same with the electronics that make up the telephone systems.
Car manufacturers buy a car from the opposition, pull it apart, and use what they find to improve their own models. Their competition does the same thing. They know it and don't care, because any of them that produced an exact replica would look pretty silly and get the scorn they deserve.
Closed/open source doesn't make it im/possible to earn a living from producing a product. And whether your competition works out how you built it, or you hand them the information, doesn't make a huge amount of difference. Trade secrets exist more in the mind than reality.
Tim wrote:
On Thu, 2006-01-19 at 03:08 -0600, Mike McCarty wrote:
One issue with this is in deployment in concert with development. Let's say that a company is going to build a new device, and wants to put an OS on an embedded computer, and develop some applications to run on it. This company does not want to divulge its trade secrets to the whole world for copying. Special techniques which may, for example, allow a cell phone to use lower power and still achieve good range, hence not cooking your brain.
Do you really think that competitors can't do that without access to the source code? Taking that example as a case in point, spectrum analysers
Are you arguing with me, or with the large corporate lawyers? What *I* believe is irrelevant on this point, and I'm not going to try to defend or argue about this position. My point is that this *is* the position taken by corporations. End of point.
If you want to argue whether large corporations take this position, then I'm willing to discuss that. But if you accept that they do, then the rest of your argument is irrelevant, because their behavior is dependent on their position, not on whether it makes sense.
And if deployment is a concern for you, then this matters.
[snip]
This reaction reminds me of a conversation I had with a german guy one time. The company I worked for had been bought by an "international corporation", and they started trying to manage american "software engineers" by using european (specifically german) techniques. Memos with the tone of edicts began appearing. Wording like:
All Engineers are ordered to use CHILL programming Languages for all Projects! No Exceptions!
(grammatical errors intentional, to show the nature of the edicts, exclamation points also intentional and present in the type of memos we received, capitals also conform to german orthography)
Now, apart from the fact that there was programming being done which COULD NOT be done in CHILL, a large number of engineers were averse to using CHILL, and didn't want it on their resume. So many of them simply sat back, folded their arms, and said "I won't". These engineers were simply sent memos even more strongly worded, and a bunch of them resigned.
The German I mentioned asked me why. I told him that Americans are free, and are used to freedom, and value their freedom, and resist tyranny in any form. He said "But they shouldn't take it personally." I told him that they don't take it personally, but they don't like it, and they exercise freedom. His response was "Well, they shouldn't *be* like that."
When I told him that they just *are* like that, and that dictatorial european methods like that won't work here, he simply repeated "They shouldn't *be* like that."
This is like deja vu all over again (apologies to Yogi Bera).
Whether corporations should *be* like that is irrelevant.
Mike
Mike McCarty:
One issue with this is in deployment in concert with development. Let's say that a company is going to build a new device, and wants to put an OS on an embedded computer, and develop some applications to run on it. This company does not want to divulge its trade secrets to the whole world for copying. Special techniques which may, for example, allow a cell phone to use lower power and still achieve good range, hence not cooking your brain.
Tim:
Do you really think that competitors can't do that without access to the source code? Taking that example as a case in point, spectrum analysers
Mike McCarty:
Are you arguing with me, or with the large corporate lawyers?
Against the point of view that source code, in itself, is a deadly secret. Against the idea that you can't sell a product without its workings being secret. Against the idea that you can keep its workings secret. And against the idea that suddenly you're out of business if someone else discovers how it works.
Tim wrote:
Mike McCarty:
One issue with this is in deployment in concert with development. Let's say that a company is going to build a new device, and wants to put an OS on an embedded computer, and develop some applications to run on it. This company does not want to divulge its trade secrets to the whole world for copying. Special techniques which may, for example, allow a cell phone to use lower power and still achieve good range, hence not cooking your brain.
Tim:
Do you really think that competitors can't do that without access to the source code? Taking that example as a case in point, spectrum analysers
Mike McCarty:
Are you arguing with me, or with the large corporate lawyers?
Against the point of view that source code, in itself, is a deadly secret. Against the idea that you can't sell a product without its workings being secret. Against the idea that you can keep its workings secret. And against the idea that suddenly you're out of business if someone else discovers how it works.
Then against the lawyers.
Mike
Andy Green wrote:
Mike McCarty wrote:
No, not even MicroSoft says that you can't call open() without making your code theirs. In fact, no one does that except GPL.
(I'm not claiming that open() is GPL, it might be LGPL, but
Your complaints in the parent post do in fact seem to rely on open() being in a GPL'd library: but Standard C library open() is in glibc which is LGPL.
No, they do not. As I have pointed out, clauses 5 and 6 of even LGPL force any *other* libraries I might link with to be distributed in a form which can be reverse engineered and repaired by the end user.
[snip]
even LGPL is so confusing to read that I'm not sure what it means, exactly, and it certainly has some scary language.)
Please read what you wrote above and reconsider if you should be lecturing the list on this.
I think that, until you read clauses 5 and 6 and digest them, you have a way to go. I find that LGPL is quite intrusive enough. I believe that LGPL makes claims which possibly would not hold up in court. IANAL, and I stated above that I find LGPL somewhat confusing to read, because I keep thinking "they can't really make that hold up, can they?". But I certainly am capable of understanding what the claims themselves seem to mean.
In particular, several years ago (2001 or so) I participated in the system requirements and system design of some cutting edge telephony equipment (VOIP stuff) and we considered whether we could use Linux and some GPL and LGPL stuff in our equipment, and after spending weeks reading this stuff, and talking with lawyers, we concluded that we could not.
So Linux went by the wayside, and we used Solaris.
So, on careful reconsideration, yes, I consider that I am probably more competent to lecture on this matter than you are, since you seem not actually to have read the LGPL.
But, IANAL.
Mike
Mike McCarty wrote:
Are you saying we should all re-write open(), close() etc?
...
No, not even MicroSoft says that you can't call open() without making your code theirs. In fact, no one does that except GPL.
(I'm not claiming that open() is GPL, it might be LGPL, but
Your complaints in the parent post do in fact seem to rely on open() being in a GPL'd library: but Standard C library open() is in glibc which is LGPL.
No, they do not.
This is not what you said in the grandparent post, as shown above. See the end of this post for LGPL section 5 and 6.
In particular, several years ago (2001 or so) I participated in the system requirements and system design of some cutting edge telephony equipment (VOIP stuff) and we considered whether we could use Linux and some GPL and LGPL stuff in our equipment, and after spending weeks reading this stuff, and talking with lawyers, we concluded that we could not.
So Linux went by the wayside, and we used Solaris.
And yet Asterisk, Bayonne etc exist and are sold as solutions your expensive Solaris implementation now has to compete with. What were the lawyers actually asked? Can we use Linux and keep everything proprietary?
So, on careful reconsideration, yes, I consider that I am probably more competent to lecture on this matter than you are, since you seem not actually to have read the LGPL.
...
As I have pointed out, clauses 5 and 6 of even LGPL force any *other* libraries I might link with to be distributed in a form which can be reverse engineered and repaired by the end user.
LGPL section 6 begins: "As an exception to the Sections above, you may also...". The reason is that section 5 deals with dynamic linking to an LGPL'd library, and section 6 deals with static linking. The reverse engineering constraint from section 6 therefore only applies to statically linked executables that actually contain the binary from the LGPL'd library. In the (normal, nowadays) case of dynamic linking, it does NOT follow that, as you claim, "LGPL force[s] any *other* libraries I might link with to be distributed in a form which can be reverse engineered and repaired by the end user.".
Here's BSD's summary (you can be sure they cut the [L]GPL no slack):
http://www.freebsd.org/doc/en_US.ISO8859-1/articles/bsdl-gpl/origins-lgpl.ht...
-Andy
Andy Green wrote:
Mike McCarty wrote:
Are you saying we should all re-write open(), close() etc?
...
No, not even MicroSoft says that you can't call open() without making your code theirs. In fact, no one does that except GPL.
(I'm not claiming that open() is GPL, it might be LGPL, but
Your complaints in the parent post do in fact seem to rely on open() being in a GPL'd library: but Standard C library open() is in glibc which is LGPL.
No, they do not.
This is not what you said in the grandparent post, as shown above. See the end of this post for LGPL section 5 and 6.
Maybe I didn't make myself clear. It doesn't matter *which* of GPL, LGPL they are. If one links with them, then the licenses seem to claim that the program produced is some sort of derived work.
In particular, several years ago (2001 or so) I participated in the system requirements and system design of some cutting edge telephony equipment (VOIP stuff) and we considered whether we could use Linux and some GPL and LGPL stuff in our equipment, and after spending weeks reading this stuff, and talking with lawyers, we concluded that we could not.
So Linux went by the wayside, and we used Solaris.
And yet Asterisk, Bayonne etc exist and are sold as solutions your expensive Solaris implementation now has to compete with. What were the lawyers actually asked? Can we use Linux and keep everything proprietary?
Eh? You know of VOIP gateways which run Linux?
So, on careful reconsideration, yes, I consider that I am probably more competent to lecture on this matter than you are, since you seem not actually to have read the LGPL.
...
As I have pointed out, clauses 5 and 6 of even LGPL force any *other* libraries I might link with to be distributed in a form which can be reverse engineered and repaired by the end user.
LGPL section 6 begins: "As an exception to the Sections above, you may also...". The reason is that section 5 deals with dynamic linking to an
Clause 5 deals with both.
LGPL'd library, and section 6 deals with static linking. The reverse engineering constraint from section 6 therefore only applies to statically linked executables that actually contain the binary from the LGPL'd library. In the (normal, nowadays) case of dynamic linking, it does NOT follow that, as you claim, "LGPL force[s] any *other* libraries I might link with to be distributed in a form which can be reverse engineered and repaired by the end user.".
You really need to get your prejudices out of the way and listen to someone who knows his stuff(*), and stop arguing. You obviously know very little about embedded programming. Dynamic linking and virtual memory are real no-nos in real-time programming. So is the use of malloc() in nearly every situation. Static memory maps are the rule of law in almost all cases. And even if not, the LGPL library must be present in the image *somewhere*, and some of the lawyers involved stated that if it is present in the image, then someone might claim it got "linked" into the image. The LGPL is vague on just exactly what constitutes "linking" and "executable".
Look, I don't particularly want to argue. In fact, I had more than enough of that back in 2000 (I guess it was 1998) or so, discussing this with the lawyers.
I'm just telling you how things work in the real world. You don't have to like it, but you do have to live with it. If deployment is important to you, then you need to know that GPL and LGPL deter(**) deployment. They also deter development of interfaces to proprietary hardware. And they prevent the use of proprietary libraries. If you don't care about these things, then that doesn't matter. But if you *do* care about these things, then you need to know that GPL and LGPL are carefully crafted to guard GNU and its derivatives from ever having anything proprietary built into or on top of them. But Solaris doesn't have this characteristic. (At least the versions I recommended. They may have put some GPL or LGPL stuff into it since I last looked.)
IMO, if you really want it free, put it into the public domain.
Otherwise, live with the consequences of [L]GPL, or come up with your own license which is better suited to your purposes.
I've never used either LGPL nor GPL, and likely never will. The stuff I want to be free, I put into the public domain. Other stuff I use licenses which seem to fit. GPL and LGPL do not, and I trow never will, fit my needs.
I've already spent more of my life trying to figure out what the GPL and LGPL really mean (and I am not sure I know that, since some of the claims likely would not hold in court [IANAL]) than I really would have liked to.
(*) in regards to how GPL and LGPL are viewed by american corporations and their lawyers, and how embedded code works
(**) "deter" and "prevent" do not mean the same thing
Mike
Mike McCarty wrote:
Maybe I didn't make myself clear. It doesn't matter *which* of GPL, LGPL they are. If one links with them, then the licenses seem to claim that the program produced is some sort of derived work.
Oh really. What do you think this means then?
''5. A program that contains no derivative of any portion of the Library, but is designed to work with the Library by being compiled or linked with it, is called a "work that uses the Library". Such a work, in isolation, is not a derivative work of the Library, and therefore falls outside the scope of this License. ...''
This is the LGPL dynamically linked case: "outside the scope of this License".
And yet Asterisk, Bayonne etc exist and are sold as solutions your expensive Solaris implementation now has to compete with. What were the lawyers actually asked? Can we use Linux and keep everything proprietary?
Eh? You know of VOIP gateways which run Linux?
Google tells me this seems to exist:
http://www.myphonecall.co.uk/voip/sip_pbx/epygi/default.aspx
From the same place you can buy cards to turn a standard PC into a PABX
with Asterisk.
As I have pointed out, clauses 5 and 6 of even LGPL force any *other* libraries I might link with to be distributed in a form which can be reverse engineered and repaired by the end user.
LGPL section 6 begins: "As an exception to the Sections above, you may also...". The reason is that section 5 deals with dynamic linking to an
Clause 5 deals with both.
Section 5 says that dynamically linked apps "fall outside the scope of this License", then goes on to define what it considers to make up static linking. Section 6 is where you discover the rules to follow if you meet the static linking criteria.
LGPL'd library, and section 6 deals with static linking. The reverse engineering constraint from section 6 therefore only applies to statically linked executables that actually contain the binary from the LGPL'd library. In the (normal, nowadays) case of dynamic linking, it does NOT follow that, as you claim, "LGPL force[s] any *other* libraries I might link with to be distributed in a form which can be reverse engineered and repaired by the end user.".
You really need to get your prejudices out of the way and listen to someone who knows his stuff(*), and stop arguing. You obviously know
Lol... well this will be my last post to you on this subject, since you do not seem to be absorbing anything. Myabe you should email Adobe, Macromedia, Xilinx and so on with your theories and see if you can make them understand they are breaking your reading of the LGPL. Good luck.
very little about embedded programming. Dynamic linking and virtual memory are real no-nos in real-time programming. So is the use of malloc() in nearly every situation. Static memory maps are the rule of law in almost all cases. And even if not, the LGPL library
If you want realtime with low latency, do it in kernelspace. But you will find usually only a small region of your full functionality needs hard realtime. The rest of it can be managed without these hard constraints in usermode... with dynamic linking to improve memory usage and cache locality.
It's good to see though that you are reduced to considering only some esoteric 1990s-style should-be-in-kernelspace "realtime" case as requiring static linking and having problems with the LGPL. I guess that means we agree that everybody else is going to be using dynamic linking and the LGPL is not trying to open their code.
I'm just telling you how things work in the real world. You don't have to like it, but you do have to live with it. If deployment is
I see.. well, thanks for that.
I've never used either LGPL nor GPL, and likely never will.
If you used a copy of linux, of course you used these licenses... or you would not have been able to get the software. If that's the case, given what you say below then you only propose to take from the ?GPL and never give. That's fine... but you should be clear about what your attitude actually is.
The stuff I want to be free, I put into the public domain. Other stuff I use licenses which seem to fit. GPL and LGPL do not, and I trow never will, fit my needs.
Well, speaking for people who do contribute to the ?GPL, we will try to struggle on without your code somehow. God knows it will be hard, but we'll manage.
Have a nice last word, and then a nice day.
-Andy
Andy Green wrote:
Mike McCarty wrote:
Maybe I didn't make myself clear. It doesn't matter *which* of GPL, LGPL they are. If one links with them, then the licenses seem to claim that the program produced is some sort of derived work.
Oh really. What do you think this means then?
''5. A program that contains no derivative of any portion of the Library, but is designed to work with the Library by being compiled or linked with it, is called a "work that uses the Library". Such a work, in isolation, is not a derivative work of the Library, and therefore falls outside the scope of this License. ...''
This is the LGPL dynamically linked case: "outside the scope of this License".
Yes, certainly. And one can do whatever one wants with such kinds of source. But in order for the program actually to be usable, it must eventually be linked with the library. The lawyers were of the opinion that if the library is eventually shipped in one piece with the software, then whether it got hooked up only when it was executed, or while we were building in our offices was irrelevant. It was still being linked. And they were of the opinion that the tool used to do the linking was also irrelevant. It didn't matter that the linking might be done merely by the packager. They were of the opinion that a court might decide that the entire image contained in the ROMs was "an executable". They were of the opinion that putting everything into one or more chips might constitute "linking". They were concerned that distributing the library and then using the loader to link up proprietary code and LGPL code "after the fact" could be considered an attempt to evade the provisions of the license, and might cause a court to view this as devious behavior.
And yet Asterisk, Bayonne etc exist and are sold as solutions your expensive Solaris implementation now has to compete with. What were the lawyers actually asked? Can we use Linux and keep everything proprietary?
Eh? You know of VOIP gateways which run Linux?
Google tells me this seems to exist:
http://www.myphonecall.co.uk/voip/sip_pbx/epygi/default.aspx
This is not a Gateway, nor is it a Gatekeeper. This is just software which talks to VOIP. Sorry, I was using a technical term from VOIP, and didn't make that clear. I wouldn't think there would be any problem just talking to VOIP. That's easy. Well, not easy. But the protocol is pretty clear. But a Gateway and a Gatekeeper need very sophisticated algorithms to track billing and so on, and do hand-offs when failure occur, and track utilization and try to balance loads, etc. They are part of the VOIP backbone, not an end-user kind of thing.
From the same place you can buy cards to turn a standard PC into a PABX
with Asterisk.
As I have pointed out, clauses 5 and 6 of even LGPL force any *other* libraries I might link with to be distributed in a form which can be reverse engineered and repaired by the end user.
LGPL section 6 begins: "As an exception to the Sections above, you may also...". The reason is that section 5 deals with dynamic linking to an
Clause 5 deals with both.
Section 5 says that dynamically linked apps "fall outside the scope of this License", then goes on to define what it considers to make up static linking. Section 6 is where you discover the rules to follow if you meet the static linking criteria.
Read what I wrote above about "devious behavior".
LGPL'd library, and section 6 deals with static linking. The reverse engineering constraint from section 6 therefore only applies to statically linked executables that actually contain the binary from the LGPL'd library. In the (normal, nowadays) case of dynamic linking, it does NOT follow that, as you claim, "LGPL force[s] any *other* libraries I might link with to be distributed in a form which can be reverse engineered and repaired by the end user.".
You really need to get your prejudices out of the way and listen to someone who knows his stuff(*), and stop arguing. You obviously know
Lol... well this will be my last post to you on this subject, since you do not seem to be absorbing anything. Myabe you should email Adobe, Macromedia, Xilinx and so on with your theories and see if you can make them understand they are breaking your reading of the LGPL. Good luck.
I guess I haven't made myself clear at all. The issue I'm addressing is the undeniable fact that there are lawyers who view things the way that I express them, and that this has an effect on where Linux gets deployed, and what software gets written for it. This is not some theory of mine. You may argue all you like whether lawyers *ought* to be concerned like this, the point is that some of them *are*.
Personally, I don't give a rat's behind about the whole matter. I don't care whether Linux lives or dies. But the GPL and the LGPL contain language which people much more adept at reading legalese claim are problematic.
The point is, that the lawyers were *unsure* of their footing, and prudent companies do not spend hundreds of millions of dollars where the ground is unsure, when for a few bucks they can get licensing which *is* known to be sure.
As an aside, Adobe does not ship the libraries they use. But embedded apps do. Also as an aside, there are many many more embedded computers than there are PCs. (But you already know that, being an embedded expert and all.)
very little about embedded programming. Dynamic linking and virtual memory are real no-nos in real-time programming. So is the use of malloc() in nearly every situation. Static memory maps are the rule of law in almost all cases. And even if not, the LGPL library
If you want realtime with low latency, do it in kernelspace. But you will find usually only a small region of your full functionality needs hard realtime. The rest of it can be managed without these hard constraints in usermode... with dynamic linking to improve memory usage and cache locality.
The issues I had in mind deal also with... making sure that there is sufficient memory to load any given app on demand. You don't want the cell phone (or whatever) not to be able to load some interface because memory is full.
It's good to see though that you are reduced to considering only some esoteric 1990s-style should-be-in-kernelspace "realtime" case as requiring static linking and having problems with the LGPL. I guess that means we agree that everybody else is going to be using dynamic linking and the LGPL is not trying to open their code.
I pointed out that the lawyers I talked to stated that whether statically or dynamically linked, the language of the LGPL is vague so that if the library is shipped with the product, then when the linking takes place is dicey. And, in consonance with what you write here, the lawyers disagreed with each other about some of the finer points. But all were agreed that it was risky.
I'm just telling you how things work in the real world. You don't have to like it, but you do have to live with it. If deployment is
I see.. well, thanks for that.
I've never used either LGPL nor GPL, and likely never will.
If you used a copy of linux, of course you used these licenses... or you
Eh? The licenses say things about distribution, not about receiving.
would not have been able to get the software. If that's the case, given what you say below then you only propose to take from the ?GPL and never give. That's fine... but you should be clear about what your attitude actually is.
In what manner have I been less than completely clear? I loaded Linux onto my machine because I was requested to do so by a client, not because I wanted it. Now that I have it, I use it. So?
I see nothing reprehensible in taking something someone freely chooses to offer without expectation of recompense, and giving no recompense in return. I have done that with some of my software, by putting it into the public domain. And I haven't ever groused at not receiving any recompense for software I provided on that basis.
The stuff I want to be free, I put into the public domain. Other stuff I use licenses which seem to fit. GPL and LGPL do not, and I trow never will, fit my needs.
Well, speaking for people who do contribute to the ?GPL, we will try to struggle on without your code somehow. God knows it will be hard, but we'll manage.
I doubt you have authority to speak for anyone but yourself. Sarcasm is unbecoming.
I couldn't give a rat's behind, as I said. But it is at least disingenous for you to assume a stance of moral high ground, when you hamper other people in using your source the way you do.
Have a nice last word, and then a nice day.
Thanks for the well-wishes. I likewise hope you have a nice day.
Mike
Paul Howarth paul@city-fan.org wrote: It's like seeing some whiz-bang control you might like to use if you're a Visual Basic programmer, except that the license for the control costs $1,000. If you don't want to pay the $1,000, you don't use the control. What's the difference? Good example, Paul. There >IS< a 'Gift Debt' that comes from using opensource, which according to my perspectives is the imperative to 'pass it forward'. You cannot directly repay those whose shoulders you stand on, but you can 'pass forward' their gift to the next guy.
If that concept holds one back, then maybe a serious life self examination is in order. Linux is free, but not as in Free Beer. Karma, to me, is a very real thing. One pays for one's decisions , and the actions from them. Make a few bad decisions and see what happens. Recieving a gift on one hand and wanting pay for some augmentation of the gifts on the other just goes against the spirit of the Linux Movement. That concept of 'paying forward' is worthy and just, according to all of the Prophets as well. (waxing spiritually here.) The Age of Aquarius arrives, steadily although slowly, through movements such as this.
You can make money from your labor, offering support, release copies of CD's, shipping & handling. My own project has 6.7 million potential clients, so I figure a buck or so average will enable me some of life's basics and some extra to spread around. Sure, I'll give away the software! That's a no brainer, I don't need to absorb all the money there is in circulation, unlike someone I could name. Just some of it.
As Paul says, the other alternative is to enter the marketplace and pay as you go. No gift-debt there, just monetary consideration. I like the possibility of repaying my 'Gift Debt' with contributions of my own, according to my skill levels and life-experience. I might make a buck or two, and as far as karma goes, I'll make what the universe will provide me, in this lifetime, according to the need to do what I'm about. If it's not worthy, I'll starve! <g>
And so it goes... <Nama Skar> Ric
--------------------------------- Yahoo! Photos Ring in the New Year with Photo Calendars. Add photos, events, holidays, whatever.
On Wed, 2006-01-18 at 11:16, Rickey Moore wrote:
Good example, Paul. There >IS< a 'Gift Debt' that comes from using opensource, which according to my perspectives is the imperative to 'pass it forward'. You cannot directly repay those whose shoulders you stand on, but you can 'pass forward' their gift to the next guy.
Are you suggesting that there is something morally wrong with the people who contributed to the *bsd components whose lack of restrictions allowed them to be used in OSX as well as continuing to be freely available? Or that OSX existing at all is a bad thing?
Les Mikesell lesmikesell@gmail.com wrote:Are you suggesting that there is something morally wrong with the people who contributed to the *bsd components whose lack of restrictions allowed them to be used in OSX as well as continuing to be freely available? Or that OSX existing at all is a bad thing? Morally wrong? Nope, no one HAS to live up to my expectations, that's a given.
My personal take on Mr. Jobs and Apple? <hitches up pants> <deep inhale of breath> <eyeballs Les> I've been largely unimpressed by Steve Jobs since 1978. Woz built a helluva machine. Woz is a Saint. Jobs closed it up, and Woz moved on. My old original Woz built Apple ][ had integer basic in ROM and would drop to a mini-assembler when you hit reset... you could load a 'copy-protected' disk into memory, hit reset and write that sucker straight to tape using assembler level commands... <evil grin> ... and you had it! The 'update' to the roms put floating point basic in and a redirect of the reset (03d0) to reboot the system. No more Wozness... bad users not in charge of their space anymore. Double bad bad to look at how something works. Now Jobs wants open source? If it wasn't for BSD and all of their contributors, Jobs would still be barefoot and fingering his butt, trying to figure a way to lock it all in for his greedy self and forever getting beat like a wet dog by Gates.
Now all these nice wellmeaning people shower Apple with BSD gifts, and it'll be a matter of time that we see a return against the GiftDebt or not. I won't be affected either way as I haven't given them sh*t since I piddled away $2500 in 1978 dollars, to find only several years later new software couldn't run because I needed to replace the entire machine for the latest.. "Then ye shall know the Devil by his works." <chuckles> OSX could be a better thing if it was Fedora'd. The entire thing. Heck, I'd probably use that.
I keep hoping Apple (apart from the OSX community) will just blow up and go to Hell one day, but that line of thinking is REALLY bad for me , my Yama turns a dark shade of maroon. I'll leave them to the Creator of the Universe to deal with. May they recieve, in like manner, as they have given.
You asked, I told ya. <cackles and runs away> Ric
================================================ My father, Victor Moore (Vic) used to say: "There are two Great Sins in the world... ...the Sin of Ignorance, and ...the Sin of Stupidity. Only the former may be overcome." R.I.P. Dad. ================================================ --------------------------------- Yahoo! Photos Ring in the New Year with Photo Calendars. Add photos, events, holidays, whatever.
On Wed, 2006-01-18 at 13:41 -0800, Rickey Moore wrote:
Les Mikesell lesmikesell@gmail.com wrote: Are you suggesting that there is something morally wrong with the people who contributed to the *bsd components whose lack of restrictions allowed them to be used in OSX as well as continuing to be freely available? Or that OSX existing at all is a bad thing? Morally wrong? Nope, no one HAS to live up to my expectations, that's a given.
My personal take on Mr. Jobs and Apple?
<hitches up pants> <deep inhale of breath> <eyeballs Les> I've been largely unimpressed by Steve Jobs since 1978. Woz built a helluva machine. Woz is a Saint. Jobs closed it up, and Woz moved on. My old original Woz built Apple ][ had integer basic in ROM and would drop to a mini-assembler when you hit reset... you could load a 'copy-protected' disk into memory, hit reset and write that sucker straight to tape using assembler level commands... <evil grin> ... and you had it! The 'update' to the roms put floating point basic in and a redirect of the reset (03d0) to reboot the system. No more Wozness... bad users not in charge of their space anymore. Double bad bad to look at how something works. Now Jobs wants open source? If it wasn't for BSD and all of their contributors, Jobs would still be barefoot and fingering his butt, trying to figure a way to lock it all in for his greedy self and forever getting beat like a wet dog by Gates.
Now all these nice wellmeaning people shower Apple with BSD gifts, and it'll be a matter of time that we see a return against the GiftDebt or not. I won't be affected either way as I haven't given them sh*t since I piddled away $2500 in 1978 dollars, to find only several years later new software couldn't run because I needed to replace the ! entire machine for the latest.. "Then ye shall know the Devil by his works." <chuckles> OSX could be a better thing if it was Fedora'd. The entire thing. Heck, I'd probably use that.
I keep hoping Apple (apart from the OSX community) will just blow up and go to Hell one day, but that line of thinking is REALLY bad for me , my Yama turns a dark shade of maroon. I'll leave them to the Creator of the Universe to deal with. May they recieve, in like manner, as they have given.
---- I think their perpetual 3-4% market share is probably akin to the hell you wish upon them. It's not enough for anyone to take them seriously.
I think that it inures to the strength of BSD and BSD license that Apple can make use of some of it's pearls and build a business model upon it. Myself, I find OSX modestly unusable and only effective if I put another 2-5 thousand dollars of software on it.
Craig
Rickey Moore wrote:
[snip]
Good example, Paul. There >IS< a 'Gift Debt' that comes from using
opensource, which according to my perspectives is the imperative to 'pass it forward'. You cannot directly repay those whose shoulders you stand on, but you can 'pass forward' their gift to the next guy.
I completely reject the concept of such a "Gift Debt" due to opensource. I base this on private e-mails I exchanged with Richard Stallman back in 1984 and 1985. Anyone who places his software under [L]GPL does so for his own motives, and makes it available for me to use or not, as I chose, within the parameters of the license.
Period.
Mike
Mike McCarty mike.mccarty@sbcglobal.net wrote: Rickey Moore wrote:
[snip]
Good example, Paul. There >IS< a 'Gift Debt'
I completely reject the concept of such a "Gift Debt" Well! There you have it! <eg> We can agree to disagree. Ric
================================================ My father, Victor Moore (Vic) used to say: "There are two Great Sins in the world... ...the Sin of Ignorance, and ...the Sin of Stupidity. Only the former may be overcome." R.I.P. Dad. ================================================ --------------------------------- Yahoo! Photos Ring in the New Year with Photo Calendars. Add photos, events, holidays, whatever.
From: "Rickey Moore" wayward4now@yahoo.com
Paul Howarth paul@city-fan.org wrote: It's like seeing some whiz-bang control you might like to use if you're a Visual Basic programmer, except that the license for the control costs $1,000. If you don't want to pay the $1,000, you don't use the control. What's the difference? Good example, Paul. There >IS< a 'Gift Debt' that comes from using opensource, which according to my perspectives is the imperative to 'pass it forward'. You cannot directly repay those whose shoulders you stand on, but you can 'pass forward' their gift to the next guy.
If that concept holds one back, then maybe a serious life self examination is in order. Linux is free, but not as in Free Beer. Karma, to me, is a very real thing. One pays for one's decisions , and the actions from them. Make a few bad decisions and see what happens. Recieving a gift on one hand and wanting pay for some augmentation of the gifts on the other just goes against the spirit of the Linux Movement. That concept of 'paying forward' is worthy and just, according to all of the Prophets as well. (waxing spiritually here.) The Age of Aquarius arrives, steadily although slowly, through movements such as this.
You can make money from your labor, offering support, release copies of CD's, shipping & handling. My own project has 6.7 million potential clients, so I figure a buck or so average will enable me some of life's basics and some extra to spread around. Sure, I'll give away the software! That's a no brainer, I don't need to absorb all the money there is in circulation, unlike someone I could name. Just some of it.
THAT is a fatuous comment. If I have to spend all my time offering support then I have no time to develop new stuff. I eventually starve. Basically that statement says I must have another source of income than writing software if I am to be a sole proprietor consultant designing software.
I made the comment as a throwaway not intended as a troll. Rightly or wrongly I perceive two things, it is impossible for me to write code for a GPLed system in such a way as to protect my own work from the accidental side effects of the GPLed system's licenses. I'd need a corps of lawyers to review everything. I cannot afford that. Nor am I allowed by some persons' ethics (particularly RS) to earn money from my work directly. If I write it I "must" release source code or be "unethical". But if I release source code for a purely software project and do not have time to "make my money from support" then some company like RedHat will make the money instead of me. My stomache does not agree with that concept. It insists on periodic infusions of food, the acquisition of which costs the money RS declares I am unethical enough to demand. Scroom, the bastard.
That's enough statement for my purposes. It's off topic here. This discussion should stop now. It's not going to change MY opinion. Nor is it likely to change the opinion of other people who must work for their food and do not want to be tied down to an employer who might manage to RedHat their way through the world.
{^_^}
jdow wrote:
From: "Rickey Moore" wayward4now@yahoo.com
Paul Howarth paul@city-fan.org wrote: It's like seeing some whiz-bang control you might like to use if you're a Visual Basic programmer, except that the license for the control costs $1,000. If you don't want to pay the $1,000, you don't use the control. What's the difference? Good example, Paul. There >IS< a 'Gift Debt' that comes from using opensource, which according to my perspectives is the imperative to 'pass it forward'. You cannot directly repay those whose shoulders you stand on, but you can 'pass forward' their gift to the next guy.
If that concept holds one back, then maybe a serious life self examination is in order. Linux is free, but not as in Free Beer. Karma, to me, is a very real thing. One pays for one's decisions , and the actions from them. Make a few bad decisions and see what happens. Recieving a gift on one hand and wanting pay for some augmentation of the gifts on the other just goes against the spirit of the Linux Movement. That concept of 'paying forward' is worthy and just, according to all of the Prophets as well. (waxing spiritually here.) The Age of Aquarius arrives, steadily although slowly, through movements such as this.
You can make money from your labor, offering support, release copies of CD's, shipping & handling. My own project has 6.7 million potential clients, so I figure a buck or so average will enable me some of life's basics and some extra to spread around. Sure, I'll give away the software! That's a no brainer, I don't need to absorb all the money there is in circulation, unlike someone I could name. Just some of it.
THAT is a fatuous comment. If I have to spend all my time offering support then I have no time to develop new stuff. I eventually starve. Basically that statement says I must have another source of income than writing software if I am to be a sole proprietor consultant designing software.
Thanks, Joanne. I had (as perhaps you may have read in other mails I have sent) e-mail contact with Richard Stallman several years ago. He has a social agenda that many people are just not aware of, and they are likewise not aware of the perhaps less obvious consequences of GPL (and LGPL). Most people are not aware that RS considers capitalism to be evil (at least in regards to creation of software). He has a social agenda of stamping out rewarding creative people for doing creative work. Put plainly, he does not believe in licensing software. He thinks it is morally reprehensible to do so. He believes that one should only sell copies, not license use.
I made the comment as a throwaway not intended as a troll. Rightly or wrongly I perceive two things, it is impossible for me to write code for a GPLed system in such a way as to protect my own work from the accidental side effects of the GPLed system's licenses. I'd need a corps of lawyers
These "side effects" are not accidental, and they are not side effects. RS wants to stamp out capitalism in software development. The GPL was created for the express purpose of killing UNIX and promoting GNU. If it manages to kill all other commercial software along the way, he'd be even happier.
to review everything. I cannot afford that. Nor am I allowed by some persons' ethics (particularly RS) to earn money from my work directly. If I write it I "must" release source code or be "unethical". But if I release source code for a purely software project and do not have time to "make my money from support" then some company like RedHat will make the money instead of me. My stomache does not agree with that concept. It insists on periodic infusions of food, the acquisition of which costs the money RS declares I am unethical enough to demand. Scroom, the bastard.
Well said, agreed to, and seconded. Amen.
That's enough statement for my purposes. It's off topic here. This discussion should stop now. It's not going to change MY opinion. Nor is it likely to change the opinion of other people who must work for their food and do not want to be tied down to an employer who might manage to RedHat their way through the world.
Nor is it likely to change the minds of college kids who live off their parents, and think [L]GPL is a Good Thing(tm).
Mike
On 1/19/06, Mike McCarty mike.mccarty@sbcglobal.net wrote:
jdow wrote:
THAT is a fatuous comment. If I have to spend all my time offering support then I have no time to develop new stuff. I eventually starve. Basically that statement says I must have another source of income than writing software if I am to be a sole proprietor consultant designing software.
Thanks, Joanne. I had (as perhaps you may have read in other mails I have sent) e-mail contact with Richard Stallman several years ago. He has a social agenda that many people are just not aware of, and they are likewise not aware of the perhaps less obvious consequences of GPL (and LGPL). Most people are not aware that RS considers capitalism to be evil (at least in regards to creation of software). He has a social agenda of stamping out rewarding creative people for doing creative work. Put plainly, he does not believe in licensing software. He thinks it is morally reprehensible to do so. He believes that one should only sell copies, not license use.
Everyone is aware of his "social agenda". He's very upfront about it. In fact, he often embarasses Tim O'Reily at conventions with his views. He has an entire website devoted to his views, in fact. Perhaps you've heard of it. It's found at http://www.gnu.org/ He doesn't believe capitalism to be evil, he believes the traditional copyright doesn't make sense for something as malleable as software.[1]
I made the comment as a throwaway not intended as a troll. Rightly or wrongly I perceive two things, it is impossible for me to write code for a GPLed system in such a way as to protect my own work from the accidental side effects of the GPLed system's licenses. I'd need a corps of lawyers
These "side effects" are not accidental, and they are not side effects. RS wants to stamp out capitalism in software development. The GPL was created for the express purpose of killing UNIX and promoting GNU. If it manages to kill all other commercial software along the way, he'd be even happier.
No, it wasn't created for the express purpose of killing UNIX (which is more properly written as Unix[2]) and promoting GNU. It was created for the express purpose of continuing the sharing of software and ideas that were prevalent at the MIT AI lab in the late 1970s and early 1980s. RMS is a researcher, a computer scientist, and as such, likes exchanging ideas with other scientists.
Commercial interests decided they could make money off software and came up with the whole "licensing" idea that we're all familiar with today (read: there was no such thing as an EULA back in the day). The only way to force people to pay for software was to make sure folks couldn't make or improve the software on their own. They'd have to buy it, and all the improvements, from the vendor. You do that by locking up the source and calling it a trade secret.
RMS was accustomed to innovation in software and that innovation was founded in people sharing their works with others. In the interests of continuing this collaborative effort, he created the GNU Foundation and the GNU Public License that represents this ideal of share and share alike.
to review everything. I cannot afford that. Nor am I allowed by some persons' ethics (particularly RS) to earn money from my work directly. If I write it I "must" release source code or be "unethical". But if I release source code for a purely software project and do not have time to "make my money from support" then some company like RedHat will make the money instead of me. My stomache does not agree with that concept. It insists on periodic infusions of food, the acquisition of which costs the money RS declares I am unethical enough to demand. Scroom, the bastard.
Well said, agreed to, and seconded. Amen.
A statement that is complete BS. How about MySQL AB? They seem to be doing pretty well for themselves selling commerically licensed versions of otherwise GPL software. They'd still be making money even if they weren't providing support themselves because people wanting to use their software in a proprietary product *have to pay real money for a license to do so.*[3]
That's enough statement for my purposes. It's off topic here. This discussion should stop now. It's not going to change MY opinion. Nor is it likely to change the opinion of other people who must work for their food and do not want to be tied down to an employer who might manage to RedHat their way through the world.
Nor is it likely to change the minds of college kids who live off their parents, and think [L]GPL is a Good Thing(tm).
Because there is no need to change *their* minds. It's *your* mind that's full of Fear, Uncertainty, and Doubt, and it's all based on misinformation and what I can only surmise is sheer pigheadedness. Great way to insult everyone on the list in your parting shot, by the way. Pity your private "OFF LIST" message actually ended up on the list.
As for my "college kid" mind that thinks the "[L]GPL is a Good Thing(tm)", it's not set foot on a college campus in over 10 years. I make my living off people that think cellular phones are a pretty neat idea.
[1] http://www.gnu.org/philosophy/philosophy.html#AboutFreeSoftware [2] http://www.answers.com/topic/unix-1#Hacker_Slang [3] http://www.mysql.com/company/legal/licensing/
-- Chris
"I trust the Democrats to take away my money, which I can afford. I trust the Republicans to take away my freedom, which I cannot."
From: "Christofer C. Bell" christofer.c.bell@gmail.com
Nor is it likely to change the minds of college kids who live off their parents, and think [L]GPL is a Good Thing(tm).
Because there is no need to change *their* minds. It's *your* mind that's full of Fear, Uncertainty, and Doubt, and it's all based on misinformation and what I can only surmise is sheer pigheadedness. Great way to insult everyone on the list in your parting shot, by the way. Pity your private "OFF LIST" message actually ended up on the list.
As for my "college kid" mind that thinks the "[L]GPL is a Good Thing(tm)", it's not set foot on a college campus in over 10 years. I make my living off people that think cellular phones are a pretty neat idea.
A callow youth ye be young fellow. Now, I will admit to setting foot on a college campus in recent decades. However, that was to visit a friend who runs the costume shop at the local JC. I've not been in a college classroom for "decades". Richard Stallman is also a callow youth from my vantage point.
I understand his ideal. I understand his not giving a damn if anybody can earn a living via programming. And I understand he intends to wipe out software copyright and any protections the creator of a body of software might have to the fruits of her work. I still say, "Scroom." I am a phylthy capitalist moneygrubber. When I work I feel I should get paid for it. I'm not trying to get rich. I'm trying to stay fed without taking a side job driving a taxi or serving fast food. Take away Stallman's income from FSF and the like and see how quickly he changes his ideals. It must be nice to get paid for doing nothing.
{^_^}
So much for Christofer's "OFF-LIST". {^_-} ----- Original Message ----- From: "jdow" jdow@earthlink.net
From: "Christofer C. Bell" christofer.c.bell@gmail.com
Nor is it likely to change the minds of college kids who live off their parents, and think [L]GPL is a Good Thing(tm).
Because there is no need to change *their* minds. It's *your* mind that's full of Fear, Uncertainty, and Doubt, and it's all based on misinformation and what I can only surmise is sheer pigheadedness. Great way to insult everyone on the list in your parting shot, by the way. Pity your private "OFF LIST" message actually ended up on the list.
As for my "college kid" mind that thinks the "[L]GPL is a Good Thing(tm)", it's not set foot on a college campus in over 10 years. I make my living off people that think cellular phones are a pretty neat idea.
A callow youth ye be young fellow. Now, I will admit to setting foot on a college campus in recent decades. However, that was to visit a friend who runs the costume shop at the local JC. I've not been in a college classroom for "decades". Richard Stallman is also a callow youth from my vantage point.
I understand his ideal. I understand his not giving a damn if anybody can earn a living via programming. And I understand he intends to wipe out software copyright and any protections the creator of a body of software might have to the fruits of her work. I still say, "Scroom." I am a phylthy capitalist moneygrubber. When I work I feel I should get paid for it. I'm not trying to get rich. I'm trying to stay fed without taking a side job driving a taxi or serving fast food. Take away Stallman's income from FSF and the like and see how quickly he changes his ideals. It must be nice to get paid for doing nothing.
{^_^}
jdow wrote:
So much for Christofer's "OFF-LIST".
Nope, sorry, that was MY foul up.
{^_-}
Mike
On Wed, 2006-01-18 at 21:10 -0800, jdow wrote:
Rightly or wrongly I perceive two things, it is impossible for me to write code for a GPLed system in such a way as to protect my own work from the accidental side effects of the GPLed system's licenses.
Just from a point of curiosity, is it not possible to create Linux software that doesn't depend on other things? Do all compilers "include" things that might contaminate you? And surely that problem might exist on other platforms, too.
Going by some of what I've seen discussed, it does rather look like Linux is set up to be something that's not ripe for exploitation. It exists, but isn't going to be a cash cow. It doesn't owe anybody else a living. And like many things in life, that can be a good thing. There's plenty of other things that can be exploited, I wouldn't like it so that the only way to use a computer, whichever one you went for, had to mean spending lots of money on it.
I certainly appreciate the very different status that Linux has compared to other computer systems, and not just because of the (low) cost of it. I like something that's not under the thumb of a capitalist.
From: "Tim" ignored_mailbox@yahoo.com.au
On Wed, 2006-01-18 at 21:10 -0800, jdow wrote:
Rightly or wrongly I perceive two things, it is impossible for me to write code for a GPLed system in such a way as to protect my own work from the accidental side effects of the GPLed system's licenses.
Just from a point of curiosity, is it not possible to create Linux software that doesn't depend on other things? Do all compilers "include" things that might contaminate you? And surely that problem might exist on other platforms, too.
The point here is that I do not know for sure I can and precisely what I can do to ensure that I will continue to be able to do this. With Windows, for extreme example, the playing field is plainly laid out. With Linux/Gnu it is not nearly so clear. And it appears to be "obscure" by intent. Since I am a small target I expect to be among the first hit if FSF wants to start raking in revenue so that it can afford to attack a giant. I do not like that potential exposure. Protecting myself from it takes too much time I might otherwise spend trying to make what I produce more error free.
Going by some of what I've seen discussed, it does rather look like Linux is set up to be something that's not ripe for exploitation. It exists, but isn't going to be a cash cow. It doesn't owe anybody else a living. And like many things in life, that can be a good thing. There's plenty of other things that can be exploited, I wouldn't like it so that the only way to use a computer, whichever one you went for, had to mean spending lots of money on it.
I certainly appreciate the very different status that Linux has compared to other computer systems, and not just because of the (low) cost of it. I like something that's not under the thumb of a capitalist.
So this is why I do not elect to develop for Linux. I see it's viral licensing as being too big a risk for one person to face alone. That is all I said in that first comment. It's all I continue to say. For *ME* it is not a suitable choice for development efforts that are designed to provide me with my living.
{^_^}
On Thu, 2006-01-19 at 04:43 -0800, jdow wrote:
this is why I do not elect to develop for Linux. I see it's viral licensing as being too big a risk for one person to face alone. That is all I said in that first comment. It's all I continue to say. For *ME* it is not a suitable choice for development efforts that are designed to provide me with my living.
Not knowing anything about what you might be doing, if you considered it worth your while, is this really a problem for you, a lost opportunity for us, or just an intellectual discussion?
There's plenty of other computer OSs to earn your living through, with a wider user base, too. And have we *really* suffered, as Linux users, because we don't have Microsoft Office, et all? Much of what I've used that's commercial hasn't been "better" than what wasn't commercial, just different. Thus far, virtually everything has been an improvement, the more I've moved into using Linux instead of the dark side.
As an abstract concept, I can't really think of a way of making masses of money out of selling fancy writing paper, and aren't concerned that other things don't allow me to do so. I make my money in some other way. Now, if the thing that I couldn't exploit was the only possible way that I could earn a living, that would be a problem. But it isn't.
I tend to view Linux like many other things that are hard to define their worth in dollar terms, but it's existence is worthwhile. Not everything in life should be about money.
Tim wrote:
On Thu, 2006-01-19 at 04:43 -0800, jdow wrote:
this is why I do not elect to develop for Linux. I see it's viral licensing as being too big a risk for one person to face alone. That is all I said in that first comment. It's all I continue to say. For *ME* it is not a suitable choice for development efforts that are designed to provide me with my living.
Not knowing anything about what you might be doing, if you considered it worth your while, is this really a problem for you, a lost opportunity for us, or just an intellectual discussion?
That all depends on perspective, doesn't it? I've given an example of the reasoning of lawyers in large corporations with regards to the [L]GPL. In particular, the LGPL is thought by the lawyers I talked to to have clauses which would not stand in court, specifically the reasoning that linking with a library creates a derivative work.
OTOH, going to court might well mean revealing the source, and demonstrating that none of the code is actually derived kills any trade secret status of the code, because it has been revealed. In any case, companies are unwilling to risk going to court in order to use Linux, when other OS with more interpretable licenses exist, and which are as cheap as Linux when you factor in the support issues which come with the "cost free" Linux versions.
There's plenty of other computer OSs to earn your living through, with a wider user base, too. And have we *really* suffered, as Linux users, because we don't have Microsoft Office, et all? Much of what I've used that's commercial hasn't been "better" than what wasn't commercial, just different. Thus far, virtually everything has been an improvement, the more I've moved into using Linux instead of the dark side.
In part it depends on whether you consider deployment good for an OS. If you don't mind being a 2% market forever, then you don't care whether you get deployment. If you want to grow, then you have to ask whether the deployment you are missing is deployment you want.
It also depends on whether you think that eventually everything you will ever want will eventually be provided in a form which conforms to the GPL, and whether you are willing to wait. In particular, I'm aware that many new devices have proprietary interfaces, so you can't create device drivers for them in a timely manner. This is a topic of discussion here, on a frequent basis. Is that something you consider "suffering", or is it just a minor nuisance? Or is it something that you think you can defeat by refusing to purchase these items? If the last, then 2% of the market is not much muscle, and we wind up with the deployment question again.
Some people seem to *like* the idea that they are suffering in some way, and like to enjoy it being because "other" people are "greedy".
As an abstract concept, I can't really think of a way of making masses of money out of selling fancy writing paper, and aren't concerned that
I don't understand what the allusion is.
other things don't allow me to do so. I make my money in some other way. Now, if the thing that I couldn't exploit was the only possible way that I could earn a living, that would be a problem. But it isn't.
I tend to view Linux like many other things that are hard to define their worth in dollar terms, but it's existence is worthwhile. Not everything in life should be about money.
Hmm, you may be one of the "some people" I mentioned above.
Eating is often about money, and is something I like to do on a regular basis. I love creating software, and I want to have a job doing what I love to do. So long as the GPL applies to GNU (and "Linux" is mostly GNU) then GNU will not be a target of my efforts.
Those things I wish to contribute to others, I don't do by hamstringing them with the GPL or even LGPL. I do it by putting them into the public domain. Then it really is free. GPL and LGPL are, IMO, not the way to contribute to a free effort. I am philosophically opposed to them, as well as practically opposed to them. For those who live in the USA, I recommend investigating Benjamin Franklin's position on patents. He was opposed to having a patent office in these USA, and failed. He therefore indeed got patents, and promptly put them into the public domain, because he was philosophically opposed to profiting from something that could benefit his fellow man. If that is the way you feel, then public domain is the way to go, not GPL or even LGPL.
Mike
Mike McCarty:
OTOH, going to court might well mean revealing the source, and demonstrating that none of the code is actually derived kills any trade secret status of the code, because it has been revealed.
Is it *really* a killer of a product if the source is available? We know how a car is built, but that doesn't put Ford out of business because Holden can do the same thing. How something is built isn't the be-all and end-all of it's existence.
Not to mention just how difficult it can be to make sense of someone else's coding. ;-)
In any case, companies are unwilling to risk going to court in order to use Linux, when other OS with more interpretable licenses exist, and which are as cheap as Linux when you factor in the support issues which come with the "cost free" Linux versions.
Which harks back to my point, do you really care about also coding for this OS as well as all the others? Do you make enough money in other ways?
Tim:
There's plenty of other computer OSs to earn your living through, with a wider user base, too. And have we *really* suffered, as Linux users, because we don't have Microsoft Office, et all? Much of what I've used that's commercial hasn't been "better" than what wasn't commercial, just different. Thus far, virtually everything has been an improvement, the more I've moved into using Linux instead of the dark side.
In part it depends on whether you consider deployment good for an OS. If you don't mind being a 2% market forever, then you don't care whether you get deployment. If you want to grow, then you have to ask whether the deployment you are missing is deployment you want.
That's a big part of it. There's plenty of things out on the market that do quite nicely for themselves, they don't dominate the market, and they don't need to. They earn their creators something that they're content with. Some people are a bit obsessional about growth.
It also depends on whether you think that eventually everything you will ever want will eventually be provided in a form which conforms to the GPL, and whether you are willing to wait. In particular, I'm aware that many new devices have proprietary interfaces, so you can't create device drivers for them in a timely manner. This is a topic of discussion here, on a frequent basis. Is that something you consider "suffering", or is it just a minor nuisance? Or is it something that you think you can defeat by refusing to purchase these items? If the last, then 2% of the market is not much muscle, and we wind up with the deployment question again.
With any computer OS there's always something that you don't have but the other one does. Depending on your needs you get what suits you best. Much as I'd like to be able to just buy any hardware off the shelf and plug it in, I see no point in Linux being the same as the others, you might as well use the other instead. Leading to:
Some people seem to *like* the idea that they are suffering in some way, and like to enjoy it being because "other" people are "greedy".
Well, the world is full of masochists and sadists...
I stopped a lot of my suffering by moving to Linux. ;-)
As an abstract concept, I can't really think of a way of making masses of money out of selling fancy writing paper, and aren't concerned that
I don't understand what the allusion is.
If I can see that *something* isn't a way to make a living, I don't attempt it, I do something else. I don't demand that I can make a living from something just because I want to. More to the point, I don't demand to change the nature of something so that I can exploit it.
I tend to view Linux like many other things that are hard to define their worth in dollar terms, but it's existence is worthwhile. Not everything in life should be about money.
Hmm, you may be one of the "some people" I mentioned above.
Eating is often about money, and is something I like to do on a regular basis. I love creating software, and I want to have a job doing what I love to do. So long as the GPL applies to GNU (and "Linux" is mostly GNU) then GNU will not be a target of my efforts.
Fair enough. For various reasons, I also don't see Linux as somewhere where many people will be able to make a living out of it. There's an ethos about it that's it's open-source, people expect it, and chose it because of that. If you're hoping to extract money from people who chose and OS because they wouldn't have to spend much on it, you're trying to create a market where there isn't one.
It's a bit like hoping that you can make a living from selling luxury cars to hippies. You know they don't want them, aren't willing to splash out the cash. So you concentrate on some other demographic (Linux isn't the only computer OS around, there's plenty of others that you can earn a living through, so you do that).
I spent years working in schools, I taught all sorts of things to people who asked me how something worked. I wasn't a teacher, I was just there, and enjoyed that aspect to helping out, while doing other things (e.g. while you're there to fix the PC, someone asks you how it works, and you explain the basics). Doing so didn't stop other people from being paid to teach the same thing. I wasn't going to demand payment for every little thing that I did; it smacks of prostituting yourself if you turn anything and everything that you can do into a pay as you go activity. So I can certainly appreciate an OS that is more communal in nature than commercial. It's a part of computing that gives back, where the others take. It (nor GPL, et al) doesn't prevent the others from being successful.
Those things I wish to contribute to others, I don't do by hamstringing them with the GPL or even LGPL. I do it by putting them into the public domain. Then it really is free. GPL and LGPL are, IMO, not the way to contribute to a free effort. I am philosophically opposed to them, as well as practically opposed to them.
I don't have too much of a problem with them, because they're not the only way of doing things. You don't like them, you use another model, you're not prevented from doing so (as you've said). But the notion of GPL is something I'm used to - my background's electronics engineering, and I would never consider that the circuit diagrams for anything that I've built are not for the eyes of other people; and I consider anyone that won't supply a diagram for something that I'm trying to repair as a complete pain in the neck. And if they'd built something based on something that I'd made, then told me that there was no way in hell that I could see the schematics for it, I'd think them an utter ass. Their /code/ doesn't need to be kept secret, nor is it really possible.
Going the other way (imposing conditions on users), I find EULAs highly despicable. They put ridiculous burdens on you. Whenever there's some sort of contract imposed on something, there's always going to be some disadvantage. But enough people don't object to GPL for it to continue on. Let those who like the idea use it.
For those who live in the USA, I recommend investigating Benjamin Franklin's position on patents. He was opposed to having a patent office in these USA, and failed. He therefore indeed got patents, and promptly put them into the public domain, because he was philosophically opposed to profiting from something that could benefit his fellow man.
It's a concept that I have sympathy for. I have no sympathy for *rampant* profiteering. A case in point is/was wheelchairs; their design was patented, so the only way to get one was to pay an exhorbitent amount of money for them. Some couldn't afford to. We could get (essentially pointless) computers for disabled kids without any dramas (it was trendy to give grants for that), but we couldn't get wheelchairs for some who needed them (no budget for it, way too expensive).
I think patents are a bit too different than GPL, though.
Tim wrote:
Mike McCarty:
OTOH, going to court might well mean revealing the source, and demonstrating that none of the code is actually derived kills any trade secret status of the code, because it has been revealed.
Is it *really* a killer of a product if the source is available? We know how a car is built, but that doesn't put Ford out of business because Holden can do the same thing. How something is built isn't the be-all and end-all of it's existence.
I'm not an expert on these matters, but I think it probably has something to do with getting venture capital and/or selling stock, when you have to go before the venture capitalist, or go to the board of directors, both of whom are probably completely clueless with regards to what software really is or how it is produced, and then try to extract money from them to start a company, or start a project, when you eventually wind up telling them that you're going to give it all away. (At least that's what it sounds like to them.)
Not to mention just how difficult it can be to make sense of someone else's coding. ;-)
This I certainly second.
In any case, companies are unwilling to risk going to court in order to use Linux, when other OS with more interpretable licenses exist, and which are as cheap as Linux when you factor in the support issues which come with the "cost free" Linux versions.
Which harks back to my point, do you really care about also coding for this OS as well as all the others? Do you make enough money in other ways?
I don't have any *personal* stake in Linux at all, either way. I just don't care. But if you do, then you care, because people like me, talking to lawyers like I did, make decisions about what OS to use.
[snip]
Some people seem to *like* the idea that they are suffering in some way, and like to enjoy it being because "other" people are "greedy".
Well, the world is full of masochists and sadists...
I stopped a lot of my suffering by moving to Linux. ;-)
That has not been my experience.
As an abstract concept, I can't really think of a way of making masses of money out of selling fancy writing paper, and aren't concerned that
I don't understand what the allusion is.
If I can see that *something* isn't a way to make a living, I don't attempt it, I do something else. I don't demand that I can make a
Well, AFAICT, you are reiterating what started this whole thread, which is that a lot of people do not develop for Linux, or include Linux in their products, precisely because so much of it is actually GNU, and using the GPL is detrimental to making a living. The statement you just made, coming from Joanne, Les, me, and perhaps a few others sparked a heated reply.
living from something just because I want to. More to the point, I don't demand to change the nature of something so that I can exploit it.
I haven't seen anyone do this, though I have seen several statements like the one you just made, which seem to imply that someone here is doing that. I have seen no one demand that GPL be changed, or that people not use GPL. I have seen people claim that they do not and will not develop for GPL. Which seems to be what you are recommending.
I tend to view Linux like many other things that are hard to define their worth in dollar terms, but it's existence is worthwhile. Not everything in life should be about money.
Hmm, you may be one of the "some people" I mentioned above.
Eating is often about money, and is something I like to do on a regular basis. I love creating software, and I want to have a job doing what I love to do. So long as the GPL applies to GNU (and "Linux" is mostly GNU) then GNU will not be a target of my efforts.
Fair enough. For various reasons, I also don't see Linux as somewhere where many people will be able to make a living out of it. There's an ethos about it that's it's open-source, people expect it, and chose it because of that. If you're hoping to extract money from people who chose and OS because they wouldn't have to spend much on it, you're trying to create a market where there isn't one.
I don't "hope to extract money" from people, and object to the use of the term wrt me.
I worked for quite some time as a watchmaker, apprenticed to my father. From him I learned a number of things, one of which was how to work in retail. One of the things I learned from him is "Your job is not to repair watches. It is to make friends of everyone who comes in that door, because you don't make money off your enemies, you make it off your friends and relatives."
The term "extract money from people" does not imply a friendly act. Doing business *should* be a friendly act.
It's a bit like hoping that you can make a living from selling luxury cars to hippies. You know they don't want them, aren't willing to splash out the cash. So you concentrate on some other demographic (Linux isn't the only computer OS around, there's plenty of others that you can earn a living through, so you do that).
I only saw three real points being made here and objected to: (1) all versions of GPL are antogonistic to making a living from software development (2) I need to eat, so I don't develop for Linux/GNU (3) Linux might profit from some injection of proprietary software
I recall (3) mostly coming from Les.
I do not, for the life of me, understand why there could possibly be any contention over (1) [though just look at this thread] when it is *obvious* that it is the intent of all versions of GPL.
Part (2) has been stated, but then there are angry reactions (which I also don't understand) demanding that people who don't like GPL and its variants should just not develop for GPL areas!? [Isn't that what I just said?]
I'm not so sure about (3). I think that many *users* of Linux have a philosopical bent such that, they are willing not to have the latest and greatest hardware, in return for what the IRS calls "intangible benefit".
I spent years working in schools, I taught all sorts of things to people who asked me how something worked. I wasn't a teacher, I was just there, and enjoyed that aspect to helping out, while doing other things (e.g. while you're there to fix the PC, someone asks you how it works, and you explain the basics). Doing so didn't stop other people from being paid to teach the same thing. I wasn't going to demand payment for every little thing that I did; it smacks of prostituting yourself if you turn anything and everything that you can do into a pay as you go activity. So I can certainly appreciate an OS that is more communal in nature than commercial. It's a part of computing that gives back, where the others take. It (nor GPL, et al) doesn't prevent the others from being successful.
This at least is sane, understandable, and even congenial. Doesn't put food on the table, though. And I disagree that "other OS" take, but don't give.
Apropos of the retail analogy above, there were many little things I did free of charge when I was a watchmaker, like replacing broken springbars (the little bars which hold watchbands on the watches) without even charging for the springbar itself. I was taught that retail is a *service* for which one gets paid. If one treats customers in a friendly manner, provides curteous service, and doesn't charge for every act one *could* charge for, then in the end one benefits financially. I was taught not to try to make a profit on each sale or other transaction. I was taught to build up a loyal clientele who trusted me not to try to gouge them for every penny I could. I was taught that one does not make profit on any one sale, but on repeat business.
As I said above, business dealings *should* be friendly acts.
Those things I wish to contribute to others, I don't do by hamstringing them with the GPL or even LGPL. I do it by putting them into the public domain. Then it really is free. GPL and LGPL are, IMO, not the way to contribute to a free effort. I am philosophically opposed to them, as well as practically opposed to them.
I don't have too much of a problem with them, because they're not the only way of doing things. You don't like them, you use another model,
Why does this exact statement, coming from the mouth of one who actually applies it to himself, result in recrimination, angry replies, and argumentation?
Joanne, Les, and I make these statements in the form "I find that GPL does not suit my needs, so I don't use it, and don't develop for it, because it makes it hard for me to make a living doing what I like to do." and the response is "You are vile. You are a parasite on the back of GPLd software because you take but do not give back. GPL doesn't mean what you say, and so you *should* develop for Linux. You owe a Gift Debt". (I saw that actual term "Gift Debt" used in this thread, or in its clone.)
You make that statement, and AFAICT, the "Linux community" doesn't complain. I wonder why?
[snip]
Going the other way (imposing conditions on users), I find EULAs highly despicable. They put ridiculous burdens on you. Whenever there's some
Not so. There are good EULAs and there are bad EULAs. I don't participate in bad EULAs any more than I have to. I recall when "copy protection" was so popular back in the '80s. I refused to use "copy protected" software. I recall taking software back on a few occasions and demanding my money back because there was a special floppy with a burnt spot on it to prevent me using the software without the original floppy.
Now we have the same thing on music CDs and DVDs. We can't even make an archival copy of a DVD even though copyright law in the USA specifically states that the publisher/owner of a copyright *must* permit that to be done for any published work. It's a felony in this country to manufacture and/or sell machines which can duplicate protected DVDs without a special license.
sort of contract imposed on something, there's always going to be some disadvantage. But enough people don't object to GPL for it to continue on. Let those who like the idea use it.
That is also not so. Contracts are *supposed* to be beneficial for all parties. You seem to think like a Marxist. Are you married? Are you aware that marriage is a contract? Do you think that marriage is inherently beneficial to one party, and detrimental to the other? Many feminists think so. Marxists believe that in every transaction there is an exploiter and an exploitee. I happen to believe that transactions can be beneficial and profitable to all parties involved. And I try to make sure that my contracts benefit all parties involved.
You know, if you use Fedora Core, you have an EULA. In fact, the GPL is itself an EULA, I think.
For those who live in the USA, I recommend investigating Benjamin Franklin's position on patents. He was opposed to having a patent office in these USA, and failed. He therefore indeed got patents, and promptly put them into the public domain, because he was philosophically opposed to profiting from something that could benefit his fellow man.
It's a concept that I have sympathy for. I have no sympathy for
I do, too. One reason some of my software is in the public domain.
*rampant* profiteering. A case in point is/was wheelchairs; their
Hmm. Emotionally charged phrase "rampant profiteering". Please define it. Somehow, when I try to pin it down, it becomes elusive and vague. Something along the lines of "making a lot more money than I do, because I'm not a rampant profiteer." The problem is that people in Africa or Latin America may have a different opinion of what "rampant profiteering" is from yours. I suspect that there are quite a few people in this world who would consider you to be a rampant profiteer.
design was patented, so the only way to get one was to pay an exhorbitent amount of money for them. Some couldn't afford to. We
My girlfriend happens to be paraplegic, so I have some personal experience with this matter. Mostly, the high cost of wheelchairs (and repair labor on them, which is even higher) is due to lack of market, not "patents". Another thing to bear in mind is that they must be very durable, but also very light. This places special constraints on the materials used in their construction, which materials are generally difficult/expensive to work, not to mention expensive in themselves. As a result of knowing her, I am gradually becoming a wheelchair mechanic, just as when I owned a motorcycle I was perforce something of a motorcycle mechanic, due to the cost, and for the same reasons. Lack of large market (by comparison with automobiles).
could get (essentially pointless) computers for disabled kids without any dramas (it was trendy to give grants for that), but we couldn't get wheelchairs for some who needed them (no budget for it, way too expensive).
I think patents are a bit too different than GPL, though.
I agree that copyright and patent law are significantly different.
Mike
jdow jdow@earthlink.net wrote:
I don't need to absorb all the money there is in circulation, unlike someone I could name. Just some of it.
THAT is a fatuous comment. If I have to spend all my time offering support then I have no time to develop new stuff. I eventually starve. Basically that statement says I must have another source of income than writing software if I am to be a sole proprietor consultant designing software.
That is true... VERY true, >IF< and only if you are a sole proprietor consultant designing software by yourself. You just might starve for real, and I certainly see your point. Linux would have starved as well, if he hadn't offered up his kernel to the net and say "Here tis! Have at it!" He's certainly maintained his 'control', managed to eat fairly well and provide for his family, and recieve back from his initial offering a ton of benefit. As you and I sit in front of our monitors, we are recieving benefit from that initital leap of faith, in the immediate here & now. That is the mindlbower! Nothing like it anywhere else, and I really like it!
Of course if Linus had been meek, and not properly assertive (as opposed to aggressive, withdrawn or passive) he might had gotten run over, robbed blind, and still sitting in his student digs with a very pissed off significant other, and hungry kids. Sure, I do see that and acknowledge the problem. It's not until the development stage is open-sourced, others allowed and encouraged to assist, followed up with asserting intitial rights to authorship and maintaining some semblence of open control of your project that the profit (minus the closed source development cycle) from the income from support, distribution, and releases comes in.
Of course this only applies to larger markets than obscure... if you were developing an industrial control specific to million dollar Milnor tunnel washers, of which there are probably only 100 or so in the world, giving away the farm would serve only to reduce your market share. No brainer. I'd close up the sucker too, if I could and get by with it. Or, if I have the ability to do this kinda thing, then maybe I need to expand my market horizon, and be acknowledged as the go-to guy in industrial controls, not just a developer for Milnor (a very fine company who builds million dollar textile washers!) specific products and hope for the referals.
One of RedHat's most profitable aquisitions was a small company that gave away a open source module designer for industrial controls. I forget their name. But, I remember they made significant coin giving away their product and selling the service that came of it, and I got to meet some of these folks and they were all smiles. That open-source worked for them was quite apparent. It's not for every situation. Nor, can one have it both ways.
So, I think it would be better if you didn't regard my comment as "fatuous" off the cuff. Maybe for your circumstance it is. Maybe, if you changed a perception or two, if wouldn't be, or maybe it still would be. I'm not in your shoes, so I can't say one way or the other, but according to my perceptions, some of which I came by the hard-way, I have my handle on the concept, for my circumstances, once again according to my perceptions. I do care enough to share them, even if my rant comes across as completely in the bucket. <g> Yeah, I have to make allowances for that.
Bob Young didn't do badly and I doubt he, or Donny Barnes, or any of the other original 'living room crew', would ever DREAM where it all would lead. So, from my examples, >I< conclude that Open Source can and does work, within the opportunities that came to those projects, and the leadership with the wit to pursue them. That is the really tough part. Good Management, with a keep eye towards Risk in all of those cases, was in place as well as Development.
Damn, that was almost 10 cents, put it on my tab, Ric
================================================ My father, Victor Moore (Vic) used to say: "There are two Great Sins in the world... ...the Sin of Ignorance, and ...the Sin of Stupidity. Only the former may be overcome." R.I.P. Dad. ================================================ --------------------------------- Yahoo! Photos Showcase holiday pictures in hardcover Photo Books. You design it and well bind it!
Paul Howarth wrote:
On Tue, 2006-01-17 at 18:27 -0800, jdow wrote:
From: "Mike McCarty" mike.mccarty@sbcglobal.net
No, GPL forces one to open everything.
Not true. It forces you to open everything that uses ANYTHING that is itself GPL contaminated and not purely your own work. (You can dual license your own work.)
I don't see what the big deal is with the GPL "encumbrance". It's just like any other license - if you can't live with its terms, just don't incorporate software licensed by it into your own work.
That's all anyone here said.
Why is it that any time someone says "I don't like [L]GPL, so I don't use libraries distributed under it" the villagers rise up with torches, storm the tower, and angrily shout "If you don't like it, then just don't USE it!"
[snip torch bearing villager]
Mike
On Tue, 2006-01-17 at 21:40 +0000, Andy Green wrote:
jdow wrote:
Andy, I am a software developer by trade. I use it to earn my daily bread and board. If I develop in an environment that involves GPL I cannot see a model that will continue to feed me and house me unless I take up a side job asking, "Do you want to supersize that, sir?" or sit at my own help desk all day instead of developing. GPL contaminates things too thoroughly. I am not a lawyer. I just read that document and basically stay away from GPL except for some recreational coding I've done.
I also design software and hardware... but depending on the field you work in, GPL stuff can bring an awful lot of firepower to the party very cheaply. Again depending on the circumstance, the compataibility, quality, time to market and royalty-free advanatages surrounding that can overwhelm the possible competitive disadvantage of having to open some of your stuff.
The general deal is AIUI if you link with GPL'd stuff -- not LGPL'd, which will not infect what it links to but only changes to itself -- you will have to open your work. It seems that we crossed a threshold now and the signs are that if you generate kernel modules you can expect that sources will be demanded.
Not necessarily. Look at the nVidia and ATI video drivers. They are kernel modules and the source is not open. They use proprietary code in the drivers.
However, there's a lot of room in there for usermode apps that just link to LGPL's libraries (like, eg, Oracle) that you for sure won't have to open source unless you want to.
Of course in your particular field if there is no specific benefit to be had from using or being directly compatible with GPL'd stuff then it makes sense to avoid it.
But if its just that you are considering to use, say, Fedora as a platform for scripts or usermode apps then AFAIUI there is zero leakage of the GPL-ness of the OS apps up into your code and BSD or GPL makes no odds there.
-Andy
fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list
From: "Jeff Vian" jvian10@charter.net
On Tue, 2006-01-17 at 21:40 +0000, Andy Green wrote:
jdow wrote:
Andy, I am a software developer by trade. I use it to earn my daily bread and board. If I develop in an environment that involves GPL I cannot see a model that will continue to feed me and house me unless I take up a side job asking, "Do you want to supersize that, sir?" or sit at my own help desk all day instead of developing. GPL contaminates things too thoroughly. I am not a lawyer. I just read that document and basically stay away from GPL except for some recreational coding I've done.
I also design software and hardware... but depending on the field you work in, GPL stuff can bring an awful lot of firepower to the party very cheaply. Again depending on the circumstance, the compataibility, quality, time to market and royalty-free advanatages surrounding that can overwhelm the possible competitive disadvantage of having to open some of your stuff.
The general deal is AIUI if you link with GPL'd stuff -- not LGPL'd, which will not infect what it links to but only changes to itself -- you will have to open your work. It seems that we crossed a threshold now and the signs are that if you generate kernel modules you can expect that sources will be demanded.
Not necessarily. Look at the nVidia and ATI video drivers. They are kernel modules and the source is not open. They use proprietary code in the drivers.
There are some young hotheads who insist that these drivers inherit some GPL headers, at least. Hence they must be themselves GPLed. There has been no serious effort to clarify the situation. It is in a rather unstable balance situation as best I can determine. Rocking the boat could result in no more drivers from manufacturers period. That would not be good for Linux any more than opening their code in the intensely competitive video board market would be good for nVidia or ATI.
It's these young hotheads and the existing "benign neglect" on the part of the majority of kernel developers who COULD bring suit that leaves me nervous. Some of what I have done for the show control industry involves custom device driver-like gadgets designed to look like MIDI ports to applications but in reality do something a little "different" like control PLCs. If I give that code away "poof" I'm out of a market.
{o.o}
jdow wrote:
will have to open your work. It seems that we crossed a threshold now and the signs are that if you generate kernel modules you can expect that sources will be demanded.
Not necessarily. Look at the nVidia and ATI video drivers. They are kernel modules and the source is not open. They use proprietary code in the drivers.
I am working on embedded linux on Arm the last six months, I am quite well aware of the options for argument on this. For various reasons we will be shipping full source for our kernelmode code. Linus has written about this question and he used the test that if there is a substantial base of code intended for crossplatform use then he wouldn't kick up a fuss. For singleplatform kerne1lspace code, he was saying you should open it.
There are some young hotheads who insist that these drivers inherit some GPL headers, at least. Hence they must be themselves GPLed. There has been no serious effort to clarify the situation. It is in a rather
The issue is defined by copyright law AIUI, it is to do with when you have made a work derived from another. People are generally drawing the line as linking stuff together as making a composed work that has to be compatible with the licenses involved. Linus has also said strongly IIRC that syscalling into kernelspace from usermode is NOT linking for these purposes.
gkh would probably be flattered with the young bit, but he does seem to be a hothead about the kernel question and is talking about removing kernel exports from non-GPL licensed kernel code.
Whatever way you look at it, the movement is towards kernelmode code needing to be opened to be consonant with the kernel license, and usermode code only needing to comply with LGPL unless you use GPL libs.
Compliance with the LGPL is not very hard, in the case where you are linking to Fedora-packaged LGPL libs then Redhat are doing the hard work in source archival terms, you just need a link on your site to a matching SRPM at Redhat IMO.
Proper compliance with the GPL for an embedded product involving the kernel + GNU stuff is much harder than it first looks when you consider ongoing versions being generated, needing to capture sources for each... we solved this by using RPM even on our embedded platform, so we capture SRPMs when distributing binaries, basically you're into making your own distro RH-style.
-Andy
On Tue, 2006-01-17 at 18:34 -0800, jdow wrote:
From: "Jeff Vian" jvian10@charter.net
On Tue, 2006-01-17 at 21:40 +0000, Andy Green wrote:
jdow wrote:
Some of what I have done for the show control industry involves custom device driver-like gadgets designed to look like MIDI ports to applications but in reality do something a little "different" like control PLCs. If I give that code away "poof" I'm out of a market.
If you would have to rewrite Linux you would also be *poof* out of the market, wouldn't you ? If you would have to buy a 75 Euro WinXPe for your 50 Euro gadget you would be 75 Euro more expensive than the competition that uses Linux and you would also be *poof* out of the market.
What I don't really understand is how you go *poof* out of the market when you sell gadgets that need a open source driver ? People still would have to by the gadget, wouldn't they ?
- Erwin
Erwin Rol wrote:
If you would have to rewrite Linux you would also be *poof* out of the market, wouldn't you ? If you would have to buy a 75 Euro WinXPe for your 50 Euro gadget you would be 75 Euro more expensive than the competition that uses Linux and you would also be *poof* out of the market.
What I don't really understand is how you go *poof* out of the market when you sell gadgets that need a open source driver ? People still would have to by the gadget, wouldn't they ?
The worry is that when much of what makes the gadget innovative is tied up in the code that must be opened, people will indeed still buy the gadget, but perhaps not from the original author of the opened code... lots of people with financial interests in such projects become highly agitated and concerned when told they must give the sources with the product, although that seems to be gradually going away as we see more and more embedded GPL stuff and so far little or no destruction from it. Companies seem to be using patents to try to put a miasma of doubt on the replication of product categories and other tricks to find comfort even in such a situation.
But I agree, seems to me one cannot overlook the advantages that come in with using the Open stuff... quality is actually one of the foremost ones... eg, the Linux networking stack... being able to dig into *everything* to find bugs, even if that is just placing printfs in some library to help find hardware problems... time to market...
-Andy
Erwin Rol:
What I don't really understand is how you go *poof* out of the market when you sell gadgets that need a open source driver ? People still would have to by the gadget, wouldn't they ?
Andy Green:
The worry is that when much of what makes the gadget innovative is tied up in the code that must be opened, people will indeed still buy the gadget, but perhaps not from the original author of the opened code... lots of people with financial interests in such projects become highly agitated and concerned when told they must give the sources with the product, although that seems to be gradually going away as we see more and more embedded GPL stuff and so far little or no destruction from it.
Nothing new with this. How many different brands of radio can you buy in a shop? Many are copies of others, in concept at least. Whatever you can make, so can someone else; you've just got to make a better product, all round. Or is your television the same brand as the first one ever designed, your radio, your phone, and so on...
Most electronic parts are common, so it's easy enough to open up a gadget and duplicate it. Whether it's worth the effort to directly copy someone else's engineering is another matter.
Occasionally I see manufacturers gouging out identifying marks on components. I hate this, it makes service next to impossible. They want you to buy $1 components from them with a $50 price, and they only keep them available for a couple of years.
If I buy something from someone, I buy what they've built, the product. That's what they sell us, something prebuilt. It's a bit rich to say they us consumers have no right to know how it works. If I decide to crack it open and modify it, I will. If I decide to crack it open and do something similar in something I'm going to use around home or work, I will. I'm sure not going to go around selling something copied that way, though.
When you make everything closed (electronics, software, etc.), you kill off the next generation of experimenters who pull things apart to find out how they work, and build better versions. Few kids getting into computers these days actually have an interest in "computing," they just want a magic trick box. Likewise with electronics, they don't build a radio or stereo amplifier, they just hunt around for prebuilt modules.
On Wed, 2006-01-18 at 06:15, Andy Green wrote:
What I don't really understand is how you go *poof* out of the market when you sell gadgets that need a open source driver ? People still would have to by the gadget, wouldn't they ?
The worry is that when much of what makes the gadget innovative is tied up in the code that must be opened, people will indeed still buy the gadget, but perhaps not from the original author of the opened code...
Or, the gadget depends on code already written and under another license. The GPL is all-or-nothing in this regard so if any component (with some interpretation of components...) needs a non-GPL license, none can have it. Personally I think this is indefensibly anti-competitive. Imagine if Microsoft said that if you used any 3rd party DLLs along with their code the 3rd party code suddenly become controlled by Microsoft's license. I'm not a lawyer, and even if I were I wouldn't risk a business on that interpretation, but there is just something wrong with the concept that you can't combine different products to add value.
Les Mikesell wrote:
Or, the gadget depends on code already written and under another license. The GPL is all-or-nothing in this regard so if any component (with some interpretation of components...) needs a non-GPL license, none can have it. Personally I think this
That isn't right as stated... Fedora contains packages under many different licenses
# rpm -qa --queryformat "%{LICENSE}\n" | sort | uniq AFL/GPL Apacheish Apache License Apache Software License Artistic Artistic or GPL BitTorrent Open Source License Boost Software License BSD BSD-compatible BSD/GPL BSD/GPL dual license BSDish BSD-like BSD-like and LGPL BSD-style BSD style + APSL ...
In a situation where the proprietary app is usermode only, and dynamically links to LGPL stuff, running on top of a mixed / GPL OS like Fedora makes no problems.
If the proprietary code you needed has to go in kernelmode, yes you can have a problem if you want to distribute the resulting binary around.
-Andy
On Wed, 2006-01-18 at 07:40 -0600, Les Mikesell wrote:
On Wed, 2006-01-18 at 06:15, Andy Green wrote:
What I don't really understand is how you go *poof* out of the market when you sell gadgets that need a open source driver ? People still would have to by the gadget, wouldn't they ?
The worry is that when much of what makes the gadget innovative is tied up in the code that must be opened, people will indeed still buy the gadget, but perhaps not from the original author of the opened code...
Or, the gadget depends on code already written and under another license. The GPL is all-or-nothing in this regard so if any component (with some interpretation of components...) needs a non-GPL license, none can have it. Personally I think this is indefensibly anti-competitive. Imagine if Microsoft said that if you used any 3rd party DLLs along with their code the 3rd party code suddenly become controlled by Microsoft's license.
Imagine if Microsoft said you had to pay to use that DLL ? Oh wait they do. Imagine if Microsoft said you where not allowed to use that DLL with open source software, oh wait they do. Imagine if Microsoft said you where not allowed to redistribute the DLL, oh wait they do. Imagine if Microsoft attached a license to the DLL, that forces all your clients to allow Microsoft to audit them, oh wait they do. Imagine if Microsoft said they patented all kind of software and you have to pay them even if you wrote the software yourself, oh wait they do. Imagine if Microsoft said you don't get the source code so you can debug your own software, oh wait the do.
I'm not a lawyer, and even if I were I wouldn't risk a business on that interpretation, but there is just something wrong with the concept that you can't combine different products to add value.
Yep there is something wrong with it, i wanted to add a device driver to Windows and distribute my new Erwindows but Microsoft didn't let me, I wonder why, i mean i used a closed license just like windows, and it added great value.
- Erwin
Erwin Rol wrote:
On Wed, 2006-01-18 at 07:40 -0600, Les Mikesell wrote:
On Wed, 2006-01-18 at 06:15, Andy Green wrote:
What I don't really understand is how you go *poof* out of the market when you sell gadgets that need a open source driver ? People still would have to by the gadget, wouldn't they ?
The worry is that when much of what makes the gadget innovative is tied up in the code that must be opened, people will indeed still buy the gadget, but perhaps not from the original author of the opened code...
Or, the gadget depends on code already written and under another license. The GPL is all-or-nothing in this regard so if any component (with some interpretation of components...) needs a non-GPL license, none can have it. Personally I think this is indefensibly anti-competitive. Imagine if Microsoft said that if you used any 3rd party DLLs along with their code the 3rd party code suddenly become controlled by Microsoft's license.
Imagine if Microsoft said you had to pay to use that DLL ? Oh wait they do. Imagine if Microsoft said you where not allowed to use that DLL with
No, they don't. You apparently are unable to parse English.
Let's be very clear and careful about this so you can follow the logic.
Suppose MicroSoft builds a DLL called MicroSoftDLL, and distributes it under some sort of license.
Suppose WidgitCo builds a DLL called WidgitCoDLL, and distributes it under some sort of license.
Suppose MikeCo builds a program called MikeCoProgram, and it uses MicroSoftDLL and WidgitCoDLL.
Then if WidgitCoDLL is GPL, MikeCo has a problem, because use of WidgitCoDLL is contingent on MicroSoftDLL being GPL as well.
The analogy would be MicroSoft insisting that WidgitCoDLL must be distributed under MicroSoft's license.
Which they do not.
But which GPL does.
[snip rambling illogical nonsense]
I'm not a lawyer, and even if I were I wouldn't risk a business on that interpretation, but there is just something wrong with the concept that you can't combine different products to add value.
Yep there is something wrong with it, i wanted to add a device driver to Windows and distribute my new Erwindows but Microsoft didn't let me, I wonder why, i mean i used a closed license just like windows, and it added great value.
By the argument you just made, we all should avoid [L]GPL libraries in commercial code.
Mike
Les Mikesell wrote:
On Wed, 2006-01-18 at 06:15, Andy Green wrote:
What I don't really understand is how you go *poof* out of the market when you sell gadgets that need a open source driver ? People still would have to by the gadget, wouldn't they ?
The worry is that when much of what makes the gadget innovative is tied up in the code that must be opened, people will indeed still buy the gadget, but perhaps not from the original author of the opened code...
Or, the gadget depends on code already written and under another license. The GPL is all-or-nothing in this regard so if any component (with some interpretation of components...) needs a non-GPL license, none can have it. Personally I think this is indefensibly anti-competitive. Imagine if Microsoft said that if you used any 3rd party DLLs along with their code the 3rd party code suddenly become controlled by Microsoft's license. I'm not a lawyer, and even if I were I wouldn't risk a business on that interpretation, but there is just something wrong with the concept that you can't combine different products to add value.
Thank you, Les. I have tried to say this a few times in this thread, but you have hit it right on the head.
Mike
From: "Erwin Rol" mailinglists@erwinrol.com
On Tue, 2006-01-17 at 18:34 -0800, jdow wrote:
From: "Jeff Vian" jvian10@charter.net
On Tue, 2006-01-17 at 21:40 +0000, Andy Green wrote:
jdow wrote:
Some of what I have done for the show control industry involves custom device driver-like gadgets designed to look like MIDI ports to applications but in reality do something a little "different" like control PLCs. If I give that code away "poof" I'm out of a market.
If you would have to rewrite Linux you would also be *poof* out of the market, wouldn't you ? If you would have to buy a 75 Euro WinXPe for your 50 Euro gadget you would be 75 Euro more expensive than the competition that uses Linux and you would also be *poof* out of the market.
What I don't really understand is how you go *poof* out of the market when you sell gadgets that need a open source driver ? People still would have to by the gadget, wouldn't they ?
Gadget is pure software, Erwin. {^_^}
On Tue, 2006-01-17 at 12:20 -0800, jdow wrote:
From: "Andy Green" andy@warmcat.com
Andy, I am a software developer by trade. I use it to earn my daily bread and board. If I develop in an environment that involves GPL I cannot see a model that will continue to feed me and house me unless I take up a side job asking, "Do you want to supersize that, sir?" or sit at my own help desk all day instead of developing. GPL contaminates things too thoroughly. I am not a lawyer. I just read that document and basically stay away from GPL except for some recreational coding I've done.
There are many entities that are making mega$ in providing services in support of GPL software. The GPL license does not mean everything is free and does not prevent use of GPL software in commercial usage. It does have certain limitations on what you can/can not do with it, and certainly switches the main scenario from "pay for the software" to "pay for support" IANAL either, but there are many who have looked at that license and it works for many more.
It is your choice to develop and sell your product, and I applaud that. In your case GPL is not an option according to your business structure. For others it makes their business more profitable.
{^_^}
Andy Green wrote:
Mike McCarty wrote:
(And maybe lose some of the GPL encumbrance problems.)
I didn't really understand what you meant by this.
FreeBSD is distributed, not under the GPL license. GPL
Yes I am reasonably aware of the two licenses. But in the case of Fedora, what are the "GPL encumbrance" problems to a normal user?
Jo did not say "normal user." She writes code, licences are important to her.
Andy Green wrote:
jdow wrote:
You're not listening, Rahul. That leads me to wonder just who constitutes the community that does the influences Fedora policies. I guess folks on this list are basically peons who are not part of the "community" or at least we're not as equal as other members.
If you look at what you're asking it boils down to asking some guys paid by Redhat to spend time and effort they could spend elsewhere continuing to give their high quality security coverage to what is now an older OS, for free. I have a remote server that is FC3, it came with FC3 and the only other choice was Windows, so I am motivated to want security coverage... but... before we huff and puff about 'community' we have to realize we are demanding their work for free in such a case. R
I've been giving my time, freely, for free for almost ten years, and before the Internet, to chess organisations.
I _always_ listened to my chessplayers and tried to give the service they wanted, and I always try to give good advice to computer users.
The value of the time I've given is way more than RH would have got had I merely bought one every retail pack they ever released, and is well comparable to that given by many OSS authors.
So far as I can judge it, much the same is true of Jo, and long timers will recall names such as Tony Nugent, Thom Pain, Jo Klemmer who've also given years of service.
You're not listening, Rahul. That leads me to wonder just who
If you look at what you're asking it boils down to asking some guys paid by Redhat to spend time and effort they could spend elsewhere continuing to give their high quality security coverage to what is now an older OS, for free. I have a remote server that is FC3, it came with FC3 and the
I've been giving my time, freely, for free for almost ten years, and before the Internet, to chess organisations.
The value of the time I've given is way more than RH would have got had
Apologies for missing this reply until now.
So. You work with chess organisations for some years.
Redhat engineers must therefore reprioritize their lives around providing security fixes for older releases of Fedora Core, because that is what you would like, due to your karmic chess-related bank account bursting at the seams.
Man if only I had known about this when I was younger, I could have gotten so many hot dates bending women to my will like toffee with this new exchange rate system between working with chess organisations and getting random people to spend their limited time on the face of our planet doing what I want just because of the supreme moral authority invested in me by my chess activities. Of course in my case, not knowing this special power would accrue, I did not spend any time with chess organizations, but boy I sure would have. You sure kept this to yourself!!! But I'm glad you finally shared the secret so future generations can take advantage of the system. Thank you.
-Andy
On Tue, 2006-17-01 at 14:29 +0000, Andy Green wrote:
jdow wrote:
You're not listening, Rahul. That leads me to wonder just who constitutes the community that does the influences Fedora policies. I guess folks on this list are basically peons who are not part of the "community" or at least we're not as equal as other members.
If you look at what you're asking it boils down to asking some guys paid by Redhat to spend time and effort they could spend elsewhere continuing to give their high quality security coverage to what is now an older OS, for free. I have a remote server that is FC3, it came with FC3 and the only other choice was Windows, so I am motivated to want security coverage... but... before we huff and puff about 'community' we have to realize we are demanding their work for free in such a case. Redhat can just as easily say, "look, we give you all this great stuff for free and we want the 'community' in the form of the Legacy project to chip in now so we can go on with new stuff".
We're aware of the policy. We disagree with it. And we're agitating for a change. It appears to affect John directly. It won't affect me until 6 test 2. But I expect I'll get plenty annoyed at that time. And if the annoyance is big enough I can go for some honest autocracy and setup with FreeBSD.
Yeah well this is a valid choice... the 'Fedora deal' is that we get high quality, up to date stuff, but we have to be on the dist-upgrade treadmill. If that deal doesn't suit your situation (I sure wish I had CentOS on that server instead of Fedora, but it wasn't a choice) then do consider to move. CentOS makes most sense for many Fedora people in situations where they can't be keeping up with the schedule, since it basically is Fedora with wisdom teeth, but OSS is all about using whatever is going to work for you, not for some particular supplier.
(And maybe lose some of the GPL encumbrance problems.)
I didn't really understand what you meant by this.
Not sure how long you have been working with Red Hat Linux Andy.
We used to have an alternative to pay for Red Hat Linux and support which was priced for small business and home use. That option was removed, and now we have to pony up for Red Hat Enterprise if we want a stable alternative from Red Hat.
The whole leach on the resources of Red Hat is totally worn out. We are providing Red Hat with a service, we are the test bed for the next version of Red Hat Enterprise. If it were not for us they would not be able to provide ongoing stable releases of their Enterprise products.
Hi
[Not speaking on behalf of Red Hat or Fedora Project]
Not sure how long you have been working with Red Hat Linux Andy.
We used to have an alternative to pay for Red Hat Linux and support which was priced for small business and home use. That option was removed, and now we have to pony up for Red Hat Enterprise if we want a stable alternative from Red Hat.
If you mean a certified and commercially supported project then yes you do have to pay for it. If you want just the software bits they are available from Red Hat's online mirrors.
The whole leach on the resources of Red Hat is totally worn out. We are providing Red Hat with a service, we are the test bed for the next version of Red Hat Enterprise.
See my other mail on details about this.
If it were not for us they would not be able to provide ongoing stable releases of their Enterprise products.
Its a mutually beneficial relationship. You get a entirely free (in both sense) cutting edge distribution every few months maintained by Red Hat. Red Hat gets some feedback from the community which has a positive effect on both Fedora and RHEL.
Guy Fraser wrote:
Not sure how long you have been working with Red Hat Linux Andy.
Since RH8. I am in fact a current RHCE.
We used to have an alternative to pay for Red Hat Linux and support which was priced for small business and home use. That option was removed, and now we have to pony up for Red Hat Enterprise if we want a stable alternative from Red Hat.
Let's not ignore CentOS and Whitebox. You don't have to pony up anything to benefit from the work Redhat (and of course the others upstream) pour into these OSes. But without Redhat you would not have as smooth a ride as you get with Fedora, CentOS and Whitebox.
The whole leach on the resources of Red Hat is totally worn out. We are providing Red Hat with a service, we are the test bed for
Let's face it, as a whole we provide a patchy grumpy angry unreliable "service". We as a whole Bugzilla stuff if we feel inclined and let most other issues slide because we are busy with our own stuff. Fine! I am not much better than the worst example of this... but unless you are the guy that Bugzillas everything and follows up on it all don't dignify our using the distro for our own benefit as some kind of wonderful service to Redhat so they owe you. In contrast they work hard and send us consistently good stuff with consistently excellent security updates, we just maybe complain when we have time (but benefit all the time).
the next version of Red Hat Enterprise. If it were not for us they would not be able to provide ongoing stable releases of their Enterprise products.
We provide some QA for them when we feel like it is all. They could do betas amongst RHEL customers if they wanted, we are hardly all that stands between them and "not be[ing] able to provide ongoing stable releases" of RHEL.
This FC3 EOL angst coincides with changes in the way Fedora is run that is a but unclear to the users yet, including me. I guess the two are connected and there is some insecurity out there about how the project will go on, the extent of RH involvement going on, what "community" means, how that cuts both ways and generally what the "New Fedora Deal" is going to look like. One thing is for sure though, the more Fedora comes out of direct RH influence, the more "the community" is going to have to put WORK in to keep things going, not just explanations of why they are owed something.
-Andy
On Mon, 2006-01-16 at 23:33 -0800, jdow wrote:
From: "Rahul Sundaram" sundaram@redhat.com
Hi
I I don't want to join it for the sake of a month; I wish to go from three to five. I prefer not to rejig my yum configurations and/or up2date configurations, and I don't wish to go to four at this late state.
Pretty simple. Continue getting updates from Fedora Legacy till Fedora Core 5 is released.
"responsibility to contribute."
Jo and I don't contribute? You gotta be joking. We've both been helping out on these lists for years (with a minor break or two when RH left us).
If you would like to get updates from Fedora Legacy project for a longer duration you will have to contribute to that project specifically which is what I meant earlier in context.
You're not listening, Rahul. That leads me to wonder just who constitutes the community that does the influences Fedora policies. I guess folks on this list are basically peons who are not part of the "community" or at least we're not as equal as other members.
We're aware of the policy. We disagree with it. And we're agitating for a change. It appears to affect John directly. It won't affect me until 6 test 2. But I expect I'll get plenty annoyed at that time. And if the annoyance is big enough I can go for some honest autocracy and setup with FreeBSD. (And maybe lose some of the GPL encumbrance problems.)
Well, let me just jump in say that *I* request that FP not change their policy. I knew going in that I was on a forced march. This has been the policy since Day 1. I started with FC1, and a few weeks after FC2 release upgraded to that, then FC3, then FC4. The point of my use of FC is to experience where Red Hat is going with RHEL, not as a production system (though it is my primary OS on my laptop). I do not want the developers wasting their time on FC3 anymore. I want them spending their time getting FC5 in better shape as crunch time is here to get it ready heading in to test3 then release. I want it fairly stable so that a few weeks after it's release it will be ready for me to upgrade to.
While, in theory, it is not much effort to continue "supporting" the old release, it is a further distraction that *I* would prefer they not spend *any* time on. As a community member I fully endorse the current arrangement.....
--Rob
On Tuesday 17 January 2006 9:22 pm, Robert Locke wrote:
Well, let me just jump in say that *I* request that FP not change their policy. I knew going in that I was on a forced march. This has been the policy since Day 1. I started with FC1, and a few weeks after FC2 release upgraded to that, then FC3, then FC4. The point of my use of FC is to experience where Red Hat is going with RHEL, not as a production system (though it is my primary OS on my laptop). I do not want the developers wasting their time on FC3 anymore. I want them spending their time getting FC5 in better shape as crunch time is here to get it ready heading in to test3 then release. I want it fairly stable so that a few weeks after it's release it will be ready for me to upgrade to.
While, in theory, it is not much effort to continue "supporting" the old release,
I'm not sure about this comment, given what Dave Jones said....
it is a further distraction that *I* would prefer they not spend *any* time on. As a community member I fully endorse the current arrangement.....
I otherwise agree with the above -- as an end user, I appreciate the current pace of development, and when I need stability I use MEPIS or PCLinuxOS, both of which I run extensively in situations where I care more about long-term stability than latest features. Let it be said that there is also 'strong' sentiment to keep things in Fedora-land as they are...
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
FWIW,
Prior to FC 5, there were always two test releases. With FC 5, we have 3 test releases, which pushes the GA a little longer. I think the decision to have 3 test releases, and to have a longer life- cycle is wise. But I would rather see FC 3 being maintained a little longer. Maybe even until FC 5 test 3 is out. You can see it as the same policy: FC X is out when FC X+2 last test release is out.
--- Ezra www.spikesource.com
Ezra Nugroho wrote:
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
FWIW,
Prior to FC 5, there were always two test releases.
Incorrect. See http://fedora.redhat.com/About/schedule/
With FC 5, we have 3 test releases, which pushes the GA a little longer. I think the decision to have 3 test releases, and to have a longer life- cycle is wise. But I would rather see FC 3 being maintained a little longer. Maybe even until FC 5 test 3 is out. You can see it as the same policy: FC X is out when FC X+2 last test release is out.
Its not the same policy since Fedora Core always had three test releases but you still had a relatively long maintenance cycle by the virtue of Fedora Core 5 prolonged development cycle . To put this in perspective,
Fedora Core 2 was released in 18 May 2004 and went into legacy mode in 11th April 2005. Thats 11 months of updates from Fedora Core and ongoing updates from Fedora Legacy which will probably run into nearly 2 years. In comparison, Fedora Core 3 was release in 8 November 2004 and went into legacy mode in 16th January 2006. Thats 1 year and 1 month of updates from Fedora Core. Fedora Legacy now maintains Red Hat Linux 7.3, 9. Fedora Core 1, 2 and 3 and they have committed themselves to continue maintaining Core 1 and subsequent releases as long as there is a community interest in it. So get your hands dirty with the work required in providing security and critical bug fixes if you care about these releases now. You can do package maintenance, QA, documentation, site maintenance etc.
On Tue, 2006-01-17 at 05:10 +0530, Rahul Sundaram wrote:
Ezra Nugroho wrote:
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
FWIW,
Prior to FC 5, there were always two test releases.
Incorrect. See http://fedora.redhat.com/About/schedule/
With FC 5, we have 3 test releases, which pushes the GA a little longer. I think the decision to have 3 test releases, and to have a longer life- cycle is wise. But I would rather see FC 3 being maintained a little longer. Maybe even until FC 5 test 3 is out. You can see it as the same policy: FC X is out when FC X+2 last test release is out.
Its not the same policy since Fedora Core always had three test releases but you still had a relatively long maintenance cycle by the virtue of Fedora Core 5 prolonged development cycle . To put this in perspective,
Fedora Core 2 was released in 18 May 2004 and went into legacy mode in 11th April 2005. Thats 11 months of updates from Fedora Core and ongoing updates from Fedora Legacy which will probably run into nearly 2 years. In comparison, Fedora Core 3 was release in 8 November 2004 and went into legacy mode in 16th January 2006. Thats 1 year and 1 month of updates from Fedora Core. Fedora Legacy now maintains Red Hat Linux 7.3, 9. Fedora Core 1, 2 and 3 and they have committed themselves to continue maintaining Core 1 and subsequent releases as long as there is a community interest in it. So get your hands dirty with the work required in providing security and critical bug fixes if you care about these releases now. You can do package maintenance, QA, documentation, site maintenance etc.
---- Just wanted to say thanks to Rahul for showing patience and dealing with all of us grumpy people who wish things happened differently than they do. I am presently dealing with issues on my other system which was upgraded from 3 to 4 a few weeks ago in anticipation of this day.
Thanks
Craig
Hi
Just wanted to say thanks to Rahul for showing patience and dealing with all of us grumpy people who wish things happened differently than they do. I am presently dealing with issues on my other system which was upgraded from 3 to 4 a few weeks ago in anticipation of this day.
Thanks
Craig
You are welcome. Wishing that the release updates would continue from Fedora Core developers for FC3 for more than a year and two months might be more reasonable if the updates end right there but its not. Fedora Legacy is continuing to provide security and critical bug fixes as long as there is people behind it to do the required work which might run into more years. Everyone within RH involved on this project is already doing tons of work required to sustain the pace of development. Our resources are not unlimited. Adding additional burden by extending the updates cycle within RH is going to affect this project negatively. I am signing up do some of the stuff for Fedora Legacy in this release. Yes, the work required is not very sexy to put it mildly but someone needs to do it if they care about it.
Hi
I am presently dealing with issues on my other system which was upgraded from 3 to 4 a few weeks ago in anticipation of this day.
What issues specifically?
On Mon, 2006-01-16 at 20:28, Rahul Sundaram wrote:
Hi
I am presently dealing with issues on my other system which was upgraded from 3 to 4 a few weeks ago in anticipation of this day.
What issues specifically?
If anyone can jump in with FC4 issues, what about firewire disk recognition? It seems to be broken by the last few FC4 kernels.
Les Mikesell wrote:
On Mon, 2006-01-16 at 20:28, Rahul Sundaram wrote:
Hi
I am presently dealing with issues on my other system which was upgraded from 3 to 4 a few weeks ago in anticipation of this day.
What issues specifically?
If anyone can jump in with FC4 issues, what about firewire disk recognition? It seems to be broken by the last few FC4 kernels.
Bug report?. What about with 2.6.15 in updates-testing?. Is that any better?
On Mon, 2006-01-16 at 21:07, Rahul Sundaram wrote:
If anyone can jump in with FC4 issues, what about firewire disk recognition? It seems to be broken by the last few FC4 kernels.
Bug report?.
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=174537 It's been there a long time.
What about with 2.6.15 in updates-testing?. Is that any better?
I'll try tomorrow. I did like the way FC4 detected firewire devices early enough during booting to notice the md devices on them in the kernels where it worked.
Craig White wrote:
do. I am presently dealing with issues on my other system which was upgraded from 3 to 4 a few weeks ago in anticipation of this day.
I'd rather do it once than twice.
On Tue, 2006-01-17 at 14:40 +0800, John Summerfied wrote:
Craig White wrote:
do. I am presently dealing with issues on my other system which was upgraded from 3 to 4 a few weeks ago in anticipation of this day.
I'd rather do it once than twice.
--- no argument from me - I'm just stating what I did in anticipation of FC-5 test 2 since I knew it meant the transfer to fedora-legacy.
I think it may be unfair to be an alarmist as fedora-legacy has been decent about updates.
Craig
Rahul Sundaram wrote:
[snip]
Fedora Core 2 was released in 18 May 2004 and went into legacy mode in 11th April 2005. Thats 11 months of updates from Fedora Core and ongoing updates from Fedora Legacy which will probably run into nearly 2 years. In comparison, Fedora Core 3 was release in 8 November 2004 and went into legacy mode in 16th January 2006. Thats 1 year and 1 month of updates from Fedora Core. Fedora Legacy now maintains Red Hat Linux 7.3, 9. Fedora Core 1, 2 and 3 and they have committed themselves to continue
I have seen this *discussed* on FC Legacy, but not *commitment* to FC1 in perpetuity.
maintaining Core 1 and subsequent releases as long as there is a community interest in it. So get your hands dirty with the work required in providing security and critical bug fixes if you care about these releases now. You can do package maintenance, QA, documentation, site maintenance etc.
I believe that, in part, this is a matter of chosing the right tool for the job. I was asked to install and use FC2 by a fellow who wanted me to use Linux for a contract job for development of some programs intended to run under Windows, SCO Unix, and Linux.
In retrospect, I would not have chosen FC for that, but rather something like CentOS, given the lifetime of the development, and the "churn" involved with the Fedora Core Project. If you really need some stability and longer-term support, you probably should not be using Fedora. This is not a criticism of the FC Project, in any way, either. FC just isn't the place to go for a stable long-term supported product. It is, as its name says, a PROJECT, not a PRODUCT.
Just my $0.02 worth.
Mike
Mike McCarty wrote:
Rahul Sundaram wrote:
[snip]
Fedora Core 2 was released in 18 May 2004 and went into legacy mode in 11th April 2005. Thats 11 months of updates from Fedora Core and ongoing updates from Fedora Legacy which will probably run into nearly 2 years. In comparison, Fedora Core 3 was release in 8 November 2004 and went into legacy mode in 16th January 2006. Thats 1 year and 1 month of updates from Fedora Core. Fedora Legacy now maintains Red Hat Linux 7.3, 9. Fedora Core 1, 2 and 3 and they have committed themselves to continue
I have seen this *discussed* on FC Legacy, but not *commitment* to FC1 in perpetuity.
Nothing is perpetual as you reminded me a few days back. Fedora Legacy project has made a commitment to continue providing updates to FC1 for now.
I believe that, in part, this is a matter of chosing the right tool for the job. I was asked to install and use FC2 by a fellow who wanted me to use Linux for a contract job for development of some programs intended to run under Windows, SCO Unix, and Linux.
The nature of the project is defined at http://fedora.redhat.com/About/
FC just isn't the place to go for a stable long-term supported product. It is, as its name says, a PROJECT, not a PRODUCT.
As referred to before, The original FAQ still has a description on why it is a project. Allow me quote it for you
" *Q:* Why a project instead of a product?
*A:*
A global steering committee at Red Hat decided that Red Hat Linux was suffering from too many compromises as a retail "product", and that we should redirect our efforts at creating a community-based project. Rather than being run through product management as something that has to appear on retail shelves on a certain date, Fedora Core will be released based on schedules, set by a steering committee, that will be open and accessible to the community, as well as influenced by the community.
"
From: "Rahul Sundaram" sundaram@redhat.com
*Q:* Why a project instead of a product?
*A:*
A global steering committee at Red Hat decided that Red Hat Linux was suffering from too many compromises as a retail "product", and that we should redirect our efforts at creating a community-based project. Rather than being run through product management as something that has to appear on retail shelves on a certain date, Fedora Core will be released based on schedules, set by a steering committee, that will be open and accessible to the community, as well as influenced by the community.
"
We're simply trying to assert some community influence, Rahul. {^_-}
Hi
We're simply trying to assert some community influence, Rahul.
Like I said, post to fedora-devel for development discussions. I am not sure how anything that does not even more additional burden to the development team would be good but maybe with good rationale it will be considered.
On Tue, Jan 17, 2006 at 01:07:46PM +0530, Rahul Sundaram wrote:
Hi
We're simply trying to assert some community influence, Rahul.
Like I said, post to fedora-devel for development discussions. I am not sure how anything that does not even more additional burden to the development team would be good but maybe with good rationale it will be considered.
-- Rahul
Well I guess all this is a matter of opinion. To me the response to a complaint of "that is how we always did it" is puerile. You seem to be the redhat representative in this discussion it seems to me you could spend some of your time communicating to the "powers" about our feelings than just defending the decision blindly.
After all your explanations I can't understand why it would hurt anything or anyone to delay the movement of FC3 to legacy until after FC5 release. People in the "community" are saying that this decision makes it more difficult for them.
This follows multiple decisions the RedHat has made that the list members have felt was wrong and not once have I detected a willingness on Red Hat's part to reconsider the decision. ------------------------------------------- Aaron Konstam Computer Science Trinity University telephone: (210)-999-7484
On Tue, 2006-01-17 at 09:26 -0600, akonstam@trinity.edu wrote:
On Tue, Jan 17, 2006 at 01:07:46PM +0530, Rahul Sundaram wrote:
Hi
We're simply trying to assert some community influence, Rahul.
Like I said, post to fedora-devel for development discussions. I am not sure how anything that does not even more additional burden to the development team would be good but maybe with good rationale it will be considered.
-- Rahul
Well I guess all this is a matter of opinion. To me the response to a complaint of "that is how we always did it" is puerile. You seem to be the redhat representative in this discussion it seems to me you could spend some of your time communicating to the "powers" about our feelings than just defending the decision blindly.
After all your explanations I can't understand why it would hurt anything or anyone to delay the movement of FC3 to legacy until after FC5 release. People in the "community" are saying that this decision makes it more difficult for them.
This follows multiple decisions the RedHat has made that the list members have felt was wrong and not once have I detected a willingness on Red Hat's part to reconsider the decision.
---- He is but one person. The decisions are made by those who are on the fedora-development list. Rahul has stated as much above.
Please take your comments to that list where it will reach those who make the decision rather than pissing in the wind here.
Craig
akonstam@trinity.edu wrote:
Well I guess all this is a matter of opinion. To me the response to a complaint of "that is how we always did it" is puerile. You seem to be the redhat representative in this discussion it seems to me you could spend some of your time communicating to the "powers" about our feelings than just defending the decision blindly.
It doesn't seem to me that this is what Rahul is doing. He's pointing out that this is not the right forum. His claim (IIUI) is that the people who need to see these messages for the most part don't take time to read this forum. He's suggested a place to post where they do read, so...
After all your explanations I can't understand why it would hurt anything or anyone to delay the movement of FC3 to legacy until after FC5 release. People in the "community" are saying that this decision makes it more difficult for them.
This follows multiple decisions the RedHat has made that the list members have felt was wrong and not once have I detected a willingness on Red Hat's part to reconsider the decision.
Did you put any of your concerns where Rahul said they would have effect?
Mike
On Tue, Jan 17, 2006 at 10:27:34AM -0600, Mike McCarty wrote:
akonstam@trinity.edu wrote:
Well I guess all this is a matter of opinion. To me the response to a complaint of "that is how we always did it" is puerile. You seem to be the redhat representative in this discussion it seems to me you could spend some of your time communicating to the "powers" about our feelings than just defending the decision blindly.
It doesn't seem to me that this is what Rahul is doing. He's pointing out that this is not the right forum. His claim (IIUI) is that the people who need to see these messages for the most part don't take time to read this forum. He's suggested a place to post where they do read, so...
After all your explanations I can't understand why it would hurt anything or anyone to delay the movement of FC3 to legacy until after FC5 release. People in the "community" are saying that this decision makes it more difficult for them.
This follows multiple decisions the RedHat has made that the list members have felt was wrong and not once have I detected a willingness on Red Hat's part to reconsider the decision.
Did you put any of your concerns where Rahul said they would have effect?
Ok, I can take a hint.
Friends,
I understand up front that I'm getting a 'sorta' RedHat product, which is always going to be cutting edge, and the payback for RedHat is that as a static group of development users, we use the latest and greatest (and most prone to break) as 'beta testers'.
That's their payback for the effort, we get the system, resources, services that we need for a cohesive project. No monetary consideration is involved on either side. If one is profiting dollarwise by the use of Fedora, and wish for stability, support, development, etc. then the deal becomes sorta one-sided, (I make money, you kiss my ass) less than mutually beneficial, as it seems from some posts that require RedHat Fedora to stop in time and maintain for their monetary interests.
People usually pay for that. RedHat Enterprise is all about Rock-Solid, plenty of caution when it comes to updating the distro, everything lacking surprises.
Fedora is about landing gear up, your seatbelt fastened, seatback and tray in the upright and locked position. General Ripper would be happy as a clam. (" -a big bird like a B-52, flying close to the ground, -- WHOOOSH!")
That's what I understood when I installed Fedora. I see it as RedHat attracting willing cutting edge users to their development team, (for free) in exchange for the latest toys to test, and a sense of community for users who are willing to do this for them (RedHat) and themselves (us, the recipients who are in Gift-Debt to RedHat for this <ahem> opportunity) and, ultimately, be of mutual benefit. If stuff blows up, it'll blow up on us first, and not the paying customers. That is properly assertive, as we accepted that deal, possessing true consent as adults, and compromise is certainly an on-going goal (..you can blow us up a little, just not ALOT!).
But, to suggest that RedHat suitcase this deal that we are "gifted" with, and do it "my way or the highway" is aggressive. Not good, according to my perspectives. If anything, it just antagonizes our benefactor, who really doesn't have to do this at all. What's with that? Again, not good, according to my perspectives.
That's my two-cents, my story and I'm sticking to it, your Honor! I'll accept a jury trial. I hope we come to a concensus, here. Ric
__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
On Tue, 17 Jan 2006, Rickey Moore wrote:
Friends,
I understand up front that I'm getting a 'sorta' RedHat product, which is always going to be cutting edge, and the payback for RedHat is that as a static group of development users, we use the latest and greatest (and most prone to break) as 'beta testers'.
<<snip pretty decent argument>>
But, to suggest that RedHat suitcase this deal that we are "gifted" with, and do it "my way or the highway" is aggressive. Not good, according to my perspectives. If anything, it just antagonizes our benefactor, who really doesn't have to do this at all. What's with that? Again, not good, according to my perspectives.
That's my two-cents, my story and I'm sticking to it, your Honor! I'll accept a jury trial. I hope we come to a concensus, here. Ric
I guess I can add my psychotic two cents. I understood Fedora to be seriously beta, testing software, which I was encouraged to use, and potientially break on the rack of weird hardware or hard usage, in return for letting people know about the experience. Tossing FC3 with FC5r2 may or may not fit my particular situation at this particular time, but with the Legacy project I find it hard to to object. There are, after all, fairly viable options available.
Obviously, if Fedora (either the software or the community) causes you trouble, pick a different situation, or seek to set yourself up in a leadership position of the community.
Of course, the fact that there is discussion at all does rather indicate that this _is_ basically a community...
Ok, maybe more like three cents. As always YMWV (Your Millage WILL Vary)
-- ============================================= If you think Education is expensive Try Ignorance Author Unknown ============================================
Hi
I guess I can add my psychotic two cents. I understood Fedora to be seriously beta, testing software, which I was encouraged to use, and potientially break on the rack of weird hardware or hard usage, in return for letting people know about the experience.
There is some pretty serious efforts going on to make sure that is not beta software and it is latest but robust.
Fedora Bug Triaging - > http://fedoraproject.org/wiki/BugZappers
Getting a full time QA person exclusively for Fedora -> https://www.redhat.com/archives/fedora-test-list/2006-January/msg00572.html
I think the current FC5 test 2 looks good. YMMV.
On Wed, 18 Jan 2006, Rahul Sundaram wrote:
Hi
I guess I can add my psychotic two cents. I understood Fedora to be seriously beta, testing software, which I was encouraged to use, and potientially break on the rack of weird hardware or hard usage, in return for letting people know about the experience.
There is some pretty serious efforts going on to make sure that is not beta software and it is latest but robust.
Fedora Bug Triaging - > http://fedoraproject.org/wiki/BugZappers
Getting a full time QA person exclusively for Fedora -> https://www.redhat.com/archives/fedora-test-list/2006-January/msg00572.html
I think the current FC5 test 2 looks good. YMMV.
Ok. Perhaps I should have added quotes around "beta". And I didn't mean to sting some hardworking people who are busy keeping bugs out of the field. Still, I understand Fedora releases to be works in progress with a limited expected lifetime, the potiential for issues which need resolving, and support a user community.
On Tue, 2006-17-01 at 08:41 -0800, Rickey Moore wrote:
Friends,
...
Fedora is about landing gear up, your seatbelt fastened, seatback and tray in the upright and locked position. General Ripper would be happy as a clam. (" -a big bird like a B-52, flying close to the ground, -- WHOOOSH!")
...
That's my two-cents, my story and I'm sticking to it, your Honor! I'll accept a jury trial. I hope we come to a concensus, here. Ric
Thanks for the comedy relief, I hope it was ment that way.
Hi
Well I guess all this is a matter of opinion. To me the response to a complaint of "that is how we always did it" is puerile. You seem to be the redhat representative in this discussion it seems to me you could spend some of your time communicating to the "powers" about our feelings than just defending the decision blindly.
Nope. That speaking on behalf of Red Hat in any of this discussions. All my opinions entirely as always.
After all your explanations I can't understand why it would hurt anything or anyone to delay the movement of FC3 to legacy until after FC5 release. People in the "community" are saying that this decision makes it more difficult for them.
This follows multiple decisions the RedHat has made that the list members have felt was wrong and not once have I detected a willingness on Red Hat's part to reconsider the decision.
Post to fedora-devel list.
akonstam@trinity.edu wrote:
Aaron Konstam Computer Science Trinity University telephone: (210)-999-7484
Another I should have named who'se given his time freely for many years.
From: "John Summerfied" debian@herakles.homelinux.org
akonstam@trinity.edu wrote:
Aaron Konstam Computer Science Trinity University telephone: (210)-999-7484
Another I should have named who'se given his time freely for many years.
And he has this bad habit of being more reasonable about it than you and I.
{^_-}
Rahul Sundaram wrote:
Hi
We're simply trying to assert some community influence, Rahul.
Like I said, post to fedora-devel for development discussions. I am not sure how anything that does not even more additional burden to the development team would be good but maybe with good rationale it will be considered.
We're users, and we're concerned. This is the list for users' concerns. By all means, point some of your developers to it, but don't expect us to enrol en masse in a developer list.
Hi
We're users, and we're concerned. This is the list for users' concerns. By all means, point some of your developers to it, but don't expect us to enrol en masse in a developer list.
I would assume you want me to forward all user concerns to the developers myself? Thats unfeasible. Sorry. Bugs are user concerns too but we have a specific forum for that - bugzilla. Similar we have a fedora-devel list for development discussion. If you dont want to get your concerns across to the developers there, then it is unlikely to get resolved anytime soon. I am off this discussion now.
Rahul Sundaram wrote:
Its not the same policy since Fedora Core always had three test releases but you still had a relatively long maintenance cycle by the virtue of Fedora Core 5 prolonged development cycle . To put this in perspective,
Whether the policy is the same as always is not the point. The point is that it's silly. It forces users to make untimely choices (such as go to a beta release) or to make needless choices and do needless work (such as reconfigure their updating software or install a soon-to-retire release) or to make unsafe choices (such as to use unmaintained software).
Hi
Whether the policy is the same as always is not the point. The point is that it's silly. It forces users to make untimely choices (such as go to a beta release) or to make needless choices and do needless work (such as reconfigure their updating software or install a soon-to-retire release) or to make unsafe choices (such as to use unmaintained software).
Merely install a legacy release repository package is not a huge burden. rpm -ivh <link/fedora-legacy.rpm> or some such. We will likely include legacy repository information by default in subsequent release to save you even that trouble. Like I said already there is no huge issues to be taken care of with the current FC3 packages. Legacy can keep cover the requirements for now. So just wait for FC5 and install it. That you didnt choose to install Fedora Core 4 does not make the update cycle silly IMO.
On Tue, 2006-01-17 at 00:39, John Summerfied wrote:
Whether the policy is the same as always is not the point. The point is that it's silly. It forces users to make untimely choices (such as go to a beta release) or to make needless choices and do needless work (such as reconfigure their updating software or install a soon-to-retire release) or to make unsafe choices (such as to use unmaintained software).
The simple fix for this would be for either the final FC3 update to install a yum repository change so subsequent updates would continue to come in from the legacy repo or just let the legacy team take over maintenance of the existing repository location - or make the existing repo mirror the legacy one.
Failing to do this is going to leave a huge number of machines vulnerable to future security problems even if they had been set up to do automatic updates. That's not a good thing for a distribution's reputation. Letting the Fedora team change focus as planned makes sense. Forcing all users to make a manual configuration change in something that was supposed to help them avoid manual work doesn't make sense.
Les Mikesell wrote:
On Tue, 2006-01-17 at 00:39, John Summerfied wrote:
Whether the policy is the same as always is not the point. The point is that it's silly. It forces users to make untimely choices (such as go to a beta release) or to make needless choices and do needless work (such as reconfigure their updating software or install a soon-to-retire release) or to make unsafe choices (such as to use unmaintained software).
The simple fix for this would be for either the final FC3 update to install a yum repository change so subsequent updates would continue to come in from the legacy repo or just let the legacy team take over maintenance of the existing repository location - or make the existing repo mirror the legacy one.
Failing to do this is going to leave a huge number of machines vulnerable to future security problems even if they had been set up to do automatic updates. That's not a good thing for a distribution's reputation. Letting the Fedora team change focus as planned makes sense. Forcing all users to make a manual configuration change in something that was supposed to help them avoid manual work doesn't make sense.
Maybe, posting an update to Yum that changes the configuration to automatically point to the fedora-legacy servers would be a better idea.
When they get the error messages about not being able to connect to the server they will definitely complain about the problem and then they could also be informed of the change and make the appropriate changes themselves.
Just my 2-cents, James Kosin
From: "James Kosin" jkosin@beta.intcomgrp.com
Les Mikesell wrote:
On Tue, 2006-01-17 at 00:39, John Summerfied wrote:
Whether the policy is the same as always is not the point. The point is that it's silly. It forces users to make untimely choices (such as go to a beta release) or to make needless choices and do needless work (such as reconfigure their updating software or install a soon-to-retire release) or to make unsafe choices (such as to use unmaintained software).
The simple fix for this would be for either the final FC3 update to install a yum repository change so subsequent updates would continue to come in from the legacy repo or just let the legacy team take over maintenance of the existing repository location - or make the existing repo mirror the legacy one.
Failing to do this is going to leave a huge number of machines vulnerable to future security problems even if they had been set up to do automatic updates. That's not a good thing for a distribution's reputation. Letting the Fedora team change focus as planned makes sense. Forcing all users to make a manual configuration change in something that was supposed to help them avoid manual work doesn't make sense.
Maybe, posting an update to Yum that changes the configuration to automatically point to the fedora-legacy servers would be a better idea.
When they get the error messages about not being able to connect to the server they will definitely complain about the problem and then they could also be informed of the change and make the appropriate changes themselves.
Just my 2-cents, James Kosin
That does sound like a quite workable solution, particularly if it's for people in John's situation. (I plan to watch the pain people go through updating from 4 to 5 for a couple months. Then if it seems feasible make the update. So it "might not" affect me.)
{^_^}
On Tue, 2006-01-17 at 09:21 -0500, James Kosin wrote:
Les Mikesell wrote:
On Tue, 2006-01-17 at 00:39, John Summerfied wrote:
Maybe, posting an update to Yum that changes the configuration to automatically point to the fedora-legacy servers would be a better idea.
Great idea. This should make ease transition to Legacy.
When they get the error messages about not being able to connect to the server they will definitely complain about the problem and then they could also be informed of the change and make the appropriate changes themselves.
I wonder if yum can be configured to show "news" feed.
Gilboa
Hi
Maybe, posting an update to Yum that changes the configuration to automatically point to the fedora-legacy servers would be a better idea.
Great idea. This should make ease transition to Legacy.
Yes, that the current process already.
When they get the error messages about not being able to connect to the server they will definitely complain about the problem and then they could also be informed of the change and make the appropriate changes themselves.
I wonder if yum can be configured to show "news" feed.
Check yum-utils package for such tools.
At 6:00 PM +0530 1/18/06, Rahul Sundaram wrote:
Hi
Maybe, posting an update to Yum that changes the configuration to automatically point to the fedora-legacy servers would be a better idea.
Great idea. This should make ease transition to Legacy.
Yes, that the current process already.
OK, so it just hasn't happened yet?
Another announcement when it does happen would make everyone feel better, and then know to look for the update if they haven't already gotten it from the Redhat Network thingy. ____________________________________________________________________ TonyN.:' mailto:tonynelson@georgeanelson.com ' http://www.georgeanelson.com/
Hi
Yes, that the current process already.
OK, so it just hasn't happened yet?
Another announcement when it does happen would make everyone feel better, and then know to look for the update if they haven't already gotten it from the Redhat Network thingy.
Watch fedora-legacy list for details. That would be prudent.
At 2:10 AM +0530 1/19/06, Rahul Sundaram wrote:
Hi
Yes, that the current process already.
OK, so it just hasn't happened yet?
Another announcement when it does happen would make everyone feel better, and then know to look for the update if they haven't already gotten it from the Redhat Network thingy.
Watch fedora-legacy list for details. That would be prudent.
Is it sufficient to watch fedora-legacy-announce? ____________________________________________________________________ TonyN.:' mailto:tonynelson@georgeanelson.com ' http://www.georgeanelson.com/
Tony Nelson wrote:
At 2:10 AM +0530 1/19/06, Rahul Sundaram wrote:
Hi
Yes, that the current process already.
OK, so it just hasn't happened yet?
Another announcement when it does happen would make everyone feel better, and then know to look for the update if they haven't already gotten it from the Redhat Network thingy.
Watch fedora-legacy list for details. That would be prudent.
Is it sufficient to watch fedora-legacy-announce?
Yes but I would recommend taking part in fedora-legacy list discussions too if you are using FL updates.
From: "Les Mikesell" lesmikesell@gmail.com
Failing to do this is going to leave a huge number of machines vulnerable to future security problems even if they had been set up to do automatic updates. That's not a good thing for a distribution's reputation. Letting the Fedora team change focus as planned makes sense. Forcing all users to make a manual configuration change in something that was supposed to help them avoid manual work doesn't make sense.
Every time a server is running fine and up2date , the current running release on the server is going to legacy state and shortly a lot repo's are stopping with that branche. I do run FC3 on some production servers out of the country in a datacenter. I do not want to do live upgrades on such machines and i dont want to go buy a airplane ticket :)
It just pitty that RH choosed for this policy. Why not 1 fedora and just 1 repo and mirrors and update the packages? I see a new RH release as a updated release with newest versions of packages. So why not create something as "yum updaterelease" so it gets the "new" version.
I am staying with all my boxes at FC3. Its running fine, better than FC1 and 2. Since packages are packaged from the opensource versions and the FC3 repo's going to be closed in future then i do uninstall the package that needs a update and install it from the orginal source tarballs. Just simple and up2date are the boxes.
I liked RH and FC much because the easy way of packages installing. I get more and more and more knowlegde of linux. That means that i like other distro's much more now because they use the orginal source tarballs in their repo's.
And the FC3 kernel... verry pitty that is stucks on 2.6.12. Whats the problem to pump it to latest kernel? depencies? put them into the repo too right hehe
We shall see, some time if i buy new HW it stucks becuase the kernel is too old..
Anyway have fun :)
Danny
Hi
And the FC3 kernel... verry pitty that is stucks on 2.6.12. Whats the problem to pump it to latest kernel? depencies? put them into the repo too right hehe
QA, testing, security issues and so on.
Danny Terweij - Net Tuning | Net wrote:
And the FC3 kernel... verry pitty that is stucks on 2.6.12. Whats the problem to pump it to latest kernel? depencies? put them into the repo too right hehe
Dave Jones summarised the problems at http://www.livejournal.com/users/kernelslacker/27433.html
Hope this helps,
James.
James Wilkinson wrote:
Danny Terweij - Net Tuning | Net wrote:
And the FC3 kernel... verry pitty that is stucks on 2.6.12. Whats the problem to pump it to latest kernel? depencies? put them into the repo too right hehe
Dave Jones summarised the problems at http://www.livejournal.com/users/kernelslacker/27433.html
Hope this helps,
James.
I've got kernel.org 2.6.15.1 running fine on a couple of FC3 boxes. No problems the I've detected, but there's nothing exotic about the configuration, just a couple of vanilla Dell boxes.
Regards,
John
Hi
Dave Jones summarised the problems at http://www.livejournal.com/users/kernelslacker/27433.html
Hope this helps,
James.
I've got kernel.org 2.6.15.1 running fine on a couple of FC3 boxes. No problems the I've detected, but there's nothing exotic about the configuration, just a couple of vanilla Dell boxes.
It wouldnt be wise to depend on one user experience to push a new major kernel release into Fedora especially when it goes into a community maintained Fedora legacy project.
On Tue, Jan 17, 2006 at 10:18:10AM -0800, John Wendel wrote:
James Wilkinson wrote:
Danny Terweij - Net Tuning | Net wrote:
And the FC3 kernel... verry pitty that is stucks on 2.6.12. Whats the problem to pump it to latest kernel? depencies? put them into the repo too right hehe
Dave Jones summarised the problems at http://www.livejournal.com/users/kernelslacker/27433.html
Hope this helps,
James.
I've got kernel.org 2.6.15.1 running fine on a couple of FC3 boxes. No problems the I've detected, but there's nothing exotic about the configuration, just a couple of vanilla Dell boxes.
I get this a lot. Building a kernel for your own box, and building one that has to work on as much hardware as possible is a completely different problem. Despite my best efforts, there are *still* always regressions when I push updates, but I do what I can to reduce that, by not rebasing if I know it's going to break a large subset of users.
To use just one example, 2.6.15 needs a new udev, or synaptics touchpads stop working. Even though the 2.6.15 update in updates-testing works just fine for a majority of users, we can't ship that until all the pieces are in place.
Dave
On Tue, 2006-01-17 at 12:30, Dave Jones wrote:
And the FC3 kernel... verry pitty that is stucks on 2.6.12. Whats the problem to pump it to latest kernel? depencies? put them into the repo too right hehe
I've got kernel.org 2.6.15.1 running fine on a couple of FC3 boxes. No problems the I've detected, but there's nothing exotic about the configuration, just a couple of vanilla Dell boxes.
I get this a lot. Building a kernel for your own box, and building one that has to work on as much hardware as possible is a completely different problem. Despite my best efforts, there are *still* always regressions when I push updates, but I do what I can to reduce that, by not rebasing if I know it's going to break a large subset of users.
To use just one example, 2.6.15 needs a new udev, or synaptics touchpads stop working. Even though the 2.6.15 update in updates-testing works just fine for a majority of users, we can't ship that until all the pieces are in place.
If you are running firewire drives, you should be happy that the FC3 kernel hasn't been updated. The 2.6.14 kernels for FC4 and the 2.6.15 in testing won't recognize drives/partitions although the raw ieee1394 layer seems to work.
Hi
If you are running firewire drives, you should be happy that the FC3 kernel hasn't been updated. The 2.6.14 kernels for FC4 and the 2.6.15 in testing won't recognize drives/partitions although the raw ieee1394 layer seems to work.
If you did test 2.6.15, a discussion on fedora-test list might help.
On Tue, 2006-01-17 at 13:21, Rahul Sundaram wrote:
If you are running firewire drives, you should be happy that the FC3 kernel hasn't been updated. The 2.6.14 kernels for FC4 and the 2.6.15 in testing won't recognize drives/partitions although the raw ieee1394 layer seems to work.
If you did test 2.6.15, a discussion on fedora-test list might help.
I did - and updated the bugzilla entry about it. Someone else had already posted more details, but the odd part seems to be that it is recognized as an sg0 device, not an sd?.
Hi
I did - and updated the bugzilla entry about it. Someone else had already posted more details, but the odd part seems to be that it is recognized as an sg0 device, not an sd?.
Weird. I don't have the hardware to check this now unfortunately but firewire in upstream kernel is pretty flaky and pretty much unsupportable.
Hi
The simple fix for this would be for either the final FC3 update to install a yum repository change so subsequent updates would continue to come in from the legacy repo or just let the legacy team take over maintenance of the existing repository location - or make the existing repo mirror the legacy one.
I actually agree with you on this and this has been planned in a future release to my understanding but this discussion should be on the fedora-legacy list instead of here.
On Tue, 2006-01-17 at 11:46, Rahul Sundaram wrote:
The simple fix for this would be for either the final FC3 update to install a yum repository change so subsequent updates would continue to come in from the legacy repo or just let the legacy team take over maintenance of the existing repository location - or make the existing repo mirror the legacy one.
I actually agree with you on this and this has been planned in a future release to my understanding but this discussion should be on the fedora-legacy list instead of here.
I don't see how that makes sense. The necessary update to the yum config would have to be done before transition to fedora-legacy instead of being done by them.
Hi
I don't see how that makes sense. The necessary update to the yum config would have to be done before transition to fedora-legacy instead of being done by them.
Fedora legacy team would have to agree to this change as in their best interests before the core team goes on a tangent and enables this.
On Tue, 2006-01-17 at 11:54, Rahul Sundaram wrote:
Hi
I don't see how that makes sense. The necessary update to the yum config would have to be done before transition to fedora-legacy instead of being done by them.
Fedora legacy team would have to agree to this change as in their best interests before the core team goes on a tangent and enables this.
Errr, I thought that the legacy team had no choice about taking over per that policy you quoted earlier. The choice is in how the core team handles the users and more importantly in how machines are affected when the user does not make any changes other than running updates. This isn't a tangent. It's a necessary security update that can only be done in the core repository. Unless, of course, the core repository becomes a mirror of legacy...
Hi
Fedora legacy team would have to agree to this change as in their best interests before the core team goes on a tangent and enables this.
Errr, I thought that the legacy team had no choice about taking over per that policy you quoted earlier.
Oh yes they do. Its entirely up to them when they decide to support a release, which repositories and on what architectures
The choice is in how the core team handles the users and more importantly in how machines are affected when the user does not make any changes other than running updates. This isn't a tangent. It's a necessary security update that can only be done in the core repository. Unless, of course, the core repository becomes a mirror of legacy...
We can make a change in core packages but I would like to hear the response from the legacy team first.
Rahul Sundaram wrote:
Hi
The simple fix for this would be for either the final FC3 update to install a yum repository change so subsequent updates would continue to come in from the legacy repo or just let the legacy team take over maintenance of the existing repository location - or make the existing repo mirror the legacy one.
I actually agree with you on this and this has been planned in a future release to my understanding but this discussion should be on the fedora-legacy list instead of here.
What? I'm expected to join _another_ new list?
Hi
I actually agree with you on this and this has been planned in a future release to my understanding but this discussion should be on the fedora-legacy list instead of here.
What? I'm expected to join _another_ new list?
Fedora legacy project has its own list. Its entirely your choice if you want to join them or not.
From: "John Summerfied" debian@herakles.homelinux.org
Rahul Sundaram wrote:
Hi
The simple fix for this would be for either the final FC3 update to install a yum repository change so subsequent updates would continue to come in from the legacy repo or just let the legacy team take over maintenance of the existing repository location - or make the existing repo mirror the legacy one.
I actually agree with you on this and this has been planned in a future release to my understanding but this discussion should be on the fedora-legacy list instead of here.
What? I'm expected to join _another_ new list?
Or file a bugzilla report. I'm not sure which is worse. I can understand their attitude, a little.
{^_-}
On Mon, 2006-01-16 at 17:09, Rahul Sundaram wrote:
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
Will the final update add the legacy repository to the yum configuration so people who aren't paying attention continue to get security fixes?
Les Mikesell wrote:
On Mon, 2006-01-16 at 17:09, Rahul Sundaram wrote:
This has always been the policy so far. FC X goes legacy when Fedora Z test 2 gets released. If you care about Fedora X at that point, join the Fedora Legacy team and contribute.
Will the final update add the legacy repository to the yum configuration so people who aren't paying attention continue to get security fixes?
Not sure. Request that in fedora-legacy list.