As you are all probably intimately aware by now, there has been discussions about 3rd party repositories overriding base Fedora packages and creating their own distribution which causes very serious problems for users and damages the community.
I suggest that we have a comittee (possibly the packaging comittee) create a wiki page which reviews 3rd party repositories for such things as: - Does the repository have a review process - Does the repository abide by Fedora packaging standards - Does the repository override Fedora base packages and possibly other criteria.
This will allow users to see which 3rd party repositories play nice with Fedora, and which ones fork Fedora to create their own distributions, etc.
I FIND IT *VERY* DISTRUBING THAT ONE OF THESE REPOSITORY MAINTAINERS IS ON THE FEDORA PACKAGING COMITTEE!
Here is an example IRC log that was taken just last night illustrating the problem:
20:40:49 RyeBrye | Trying to get yum to install gtk2-devel from the update repo I get 4 dependency errors 20:41:10 RyeBrye | the first one is gtk2 itself 20:41:17 XulChris | RyeBrye: caused by using atrpms no doubt 20:41:25 RyeBrye | hm 20:41:34 RyeBrye | How can I tell which packages I have installed that came from atrpms? 20:42:00 XulChris | RyeBrye: disable the atrpms repo, and run "yum list extras" 20:43:42 RyeBrye | I'm relatively new to Linux - but it seems that the prevailing opinion is that atrpms = the devil 20:44:23 XulChris | RyeBrye: basically yes, we even attempted to tell AT this, but he refused to listen and accused of of using "FUD" techniques against his repository 20:44:40 RyeBrye | Oh. AT is a person? 20:44:46 XulChris | Axel Thimm 20:45:49 XulChris | RyeBrye: we are trying to put measures in place to prevent repositories from doing this 20:45:59 XulChris | see https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=211250 20:50:06 RyeBrye | Sounds like I should get that "priorities" plugin for yum going to prevent atrpms from going buck-wild installing packages 20:50:39 errr | oh boy, more atrpms "fud" :P 20:50:52 XulChris | RyeBrye: yea, or if you just need a single rpm, try installing it by hand using the rpm command instead of yum, and if it doesnt install, get the srpm file and rebuild the rpm from the srpm 20:51:08 XulChris | errr: lol 20:51:27 errr | that was really getting funny in my inbox the other day 20:51:53 XulChris | ya, what did we get? like 80 messages within 24 hours? :) 20:52:00 errr | maybe more 20:52:00 errr | lol 20:52:04 XulChris | hehe 20:52:11 * rewster is pretty sure atrpms' rpm for Myth doesn't build as-is on stock fedora 20:53:51 * RyeBrye is busy removing a whole lot of crap because atrpms installed a gtk2 package - and just about everything he installed from atrpms (mythtv) depended on it 20:53:56 rewster | err srpm. and I think the OP mentioned having AT for myth. 20:54:18 RyeBrye | or rather... yum is busy removing it 20:54:20 rewster | yeah... which is a bummer. A lot of the faqs for myth just say "use atrpms" 22:09:56 RyeBrye | ATRpms really f*'ed up my system... it installed a gtk2 that - when I removed it - removed just about every darn package I had installed because of dependencies... argggggg 22:10:18 EvilBob | RyeBrye: file a bug with them 22:10:24 autopsy | RyeBrye, don't use ATrpms. 22:10:27 siXy | again? 22:10:28 EvilBob | RyeBrye: PLEASE 22:10:42 errr | lol 22:10:49 RyeBrye | I'm learning my lesson... 22:10:56 RyeBrye | Does livna really have mythtv rpms? 22:10:59 siXy | or are you reffering to the first time? 22:11:02 siXy | no 22:11:05 EvilBob | RyeBrye: sorry you got burned but the maintainer insists that there is not a problem with how he does things 22:11:12 errr | does Alex never come to this IRC? 22:11:24 autopsy | You mean Axel. 22:11:26 siXy | axel is an interesting guy... 22:11:28 EvilBob | errr: Alex who? 22:11:34 RyeBrye | Why doesn't he just make his own damned distro! argh 22:11:40 XulChris | EvilBob: axel thimm (AT) 22:11:43 errr | autopsy: oh maybe I do 22:11:45 siXy | some would say he does 22:11:50 EvilBob | RyeBrye: add that to your bug 22:12:20 errr | I dont follow the room all that close but at least once or twice a week I notice someone in here with the ATrpms sobstory 22:12:52 siXy | but anyways - do like i told you and update, then enable atrpms grab myth and its explicit dependancies then delete the atrpms repo 22:13:13 RyeBrye | siXy - yeah, you are right 22:13:31 EvilBob | never ever update with atRPMs enabled 22:13:44 siXy | its important you update fist tho otherwise when you grab myth that will pull some atrpms pacakges that you really dont want 22:14:11 RyeBrye | is there a way to update without it loading a new kernel? I don't feel like rebuilding all my kernel extensions at the moment 22:14:53 siXy | RyeBrye: do you have a copy of that line to remove all atrpms pacakges? was trying to remember it earlier for someone else 22:14:53 chemaja The bz that XulChris pointed us to (https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=211250) will be the key once it is implemented.
Christopher Stone wrote :
As you are all probably intimately aware by now, there has been discussions about 3rd party repositories overriding base Fedora packages and creating their own distribution which causes very serious problems for users and damages the community.
This definitely reads like a personal opinion more than a fact.
I suggest that we have a comittee (possibly the packaging comittee) create a wiki page which reviews 3rd party repositories for such things as:
- Does the repository have a review process
- Does the repository abide by Fedora packaging standards
- Does the repository override Fedora base packages
and possibly other criteria.
This will allow users to see which 3rd party repositories play nice with Fedora, and which ones fork Fedora to create their own distributions, etc.
I don't see how this is possible nor how it would help the Fedora project in any way. Fedora is about Core, Extras and all the orbiting projects. It's not about going hunting for (nor caring about) external stuff.
I FIND IT *VERY* DISTRUBING THAT ONE OF THESE REPOSITORY MAINTAINERS IS ON THE FEDORA PACKAGING COMITTEE!
Again : Your opinion.
I didn't reply in the OT flame fest, and will only reply once here, quoting Jesse :
"Make it stop!!"
It wasn't meant for the maintainers list, I don't see how it suits the packagers list any better. Thanks.
Matthias
On Wed, 18 Oct 2006, Christopher Stone wrote:
I suggest that we have a comittee (possibly the packaging comittee) create a wiki page which reviews 3rd party repositories for such things as:
- Does the repository have a review process
- Does the repository abide by Fedora packaging standards
- Does the repository override Fedora base packages
and possibly other criteria.
This will allow users to see which 3rd party repositories play nice with Fedora, and which ones fork Fedora to create their own distributions, etc.
I FIND IT *VERY* DISTRUBING THAT ONE OF THESE REPOSITORY MAINTAINERS IS ON THE FEDORA PACKAGING COMITTEE!
You know what? I find it very disturbing that, despite the best efforts of everyone to keep this thread reasonably civil, you continue to derail it with nonsense like this. NOBODY IS IMPRESSED WHEN YOU SHOUT.
Do you know why Axel is on the packaging committee?
Because Axel was one of the people who was doing a good job at creating packages for Fedora before the Fedora Extras committee was even vaguely functional.
Because Axel has earned the respect of Fedora packagers all over the world with his knowledge, insights, and hard work.
Because we don't always agree with Axel's decisions, but we also understand that diversity in the decision-making body helps us come to stronger decisions.
Personally? I think we need the mechanism you suggest, Chris. But you make it *insanely* difficult to take your side with this kind of behavior.
--g
------------------------------------------------------------- Greg DeKoenigsberg || Fedora Project || fedoraproject.org Be an Ambassador || http://fedoraproject.org/wiki/Ambassadors -------------------------------------------------------------
Christopher Stone wrote:
I suggest that we have a comittee (possibly the packaging comittee) create a wiki page which reviews 3rd party repositories for such things as:
IMO, this is outside the scope of the packaging committee's mandate/rights/responsibility.
I'd suggest you start a new SIG if you feel so strongly about this subject.
-- Rex
On 10/18/06, Rex Dieter rdieter@math.unl.edu wrote:
Christopher Stone wrote:
I suggest that we have a comittee (possibly the packaging comittee) create a wiki page which reviews 3rd party repositories for such things as:
IMO, this is outside the scope of the packaging committee's mandate/rights/responsibility.
I'd suggest you start a new SIG if you feel so strongly about this subject.
I would be perfectly happy with this, even If its a one man SIG consisting of only me. As long as there is a wiki page on the Fedora wiki which has this information and users can be pointed to this page to learn about the consequences of installing another repository I would be happy.
However, I think it would be better to have a comittee review other repositories instead of a single person who might be biased such as myself.
On Wed, 2006-10-18 at 11:02 -0700, Christopher Stone wrote:
On 10/18/06, Rex Dieter rdieter@math.unl.edu wrote:
Christopher Stone wrote:
I suggest that we have a comittee (possibly the packaging comittee) create a wiki page which reviews 3rd party repositories for such things as:
IMO, this is outside the scope of the packaging committee's mandate/rights/responsibility.
I'd suggest you start a new SIG if you feel so strongly about this subject.
I would be perfectly happy with this, even If its a one man SIG consisting of only me. As long as there is a wiki page on the Fedora wiki which has this information and users can be pointed to this page to learn about the consequences of installing another repository I would be happy.
However, I think it would be better to have a comittee review other repositories instead of a single person who might be biased such as myself.
I think that if this is as big of a problem as you claim, then it should be rather trivial for you (or someone else motivated) to install a Fedora box, enable the atrpms repo, and start filing bugs if/when things break. I even think it would be more productive to highlight the FC or FE packages that atrpms is providing overrides for, and start a discussion around why these packages exist, and if there exists the possibility to merge the changes into the FC or FE package and retire the atrpms packages. I'm sure that Axel would welcome that discussion, as less packages means less work for him. :)
Running around screaming isn't really going to motivate anyone.
~spot
On 10/18/06, Tom 'spot' Callaway tcallawa@redhat.com wrote:
On Wed, 2006-10-18 at 11:02 -0700, Christopher Stone wrote:
On 10/18/06, Rex Dieter rdieter@math.unl.edu wrote:
Christopher Stone wrote:
I suggest that we have a comittee (possibly the packaging comittee) create a wiki page which reviews 3rd party repositories for such things as:
IMO, this is outside the scope of the packaging committee's mandate/rights/responsibility.
I'd suggest you start a new SIG if you feel so strongly about this subject.
I would be perfectly happy with this, even If its a one man SIG consisting of only me. As long as there is a wiki page on the Fedora wiki which has this information and users can be pointed to this page to learn about the consequences of installing another repository I would be happy.
However, I think it would be better to have a comittee review other repositories instead of a single person who might be biased such as myself.
I think that if this is as big of a problem as you claim, then it should be rather trivial for you (or someone else motivated) to install a Fedora box, enable the atrpms repo, and start filing bugs if/when things break. I even think it would be more productive to highlight the FC or FE packages that atrpms is providing overrides for, and start a discussion around why these packages exist, and if there exists the possibility to merge the changes into the FC or FE package and retire the atrpms packages. I'm sure that Axel would welcome that discussion, as less packages means less work for him. :)
Fair enough. I will start by filing a bug report against the gtk libs RyeBrye had problems with. I hope this solution works. If it does not, I will re-address this issue here.
On 10/18/06, Christopher Stone chris.stone@gmail.com wrote:
On 10/18/06, Tom 'spot' Callaway tcallawa@redhat.com wrote:
On Wed, 2006-10-18 at 11:02 -0700, Christopher Stone wrote:
On 10/18/06, Rex Dieter rdieter@math.unl.edu wrote:
Christopher Stone wrote:
I suggest that we have a comittee (possibly the packaging comittee) create a wiki page which reviews 3rd party repositories for such things as:
IMO, this is outside the scope of the packaging committee's mandate/rights/responsibility.
I'd suggest you start a new SIG if you feel so strongly about this subject.
I would be perfectly happy with this, even If its a one man SIG consisting of only me. As long as there is a wiki page on the Fedora wiki which has this information and users can be pointed to this page to learn about the consequences of installing another repository I would be happy.
However, I think it would be better to have a comittee review other repositories instead of a single person who might be biased such as myself.
I think that if this is as big of a problem as you claim, then it should be rather trivial for you (or someone else motivated) to install a Fedora box, enable the atrpms repo, and start filing bugs if/when things break. I even think it would be more productive to highlight the FC or FE packages that atrpms is providing overrides for, and start a discussion around why these packages exist, and if there exists the possibility to merge the changes into the FC or FE package and retire the atrpms packages. I'm sure that Axel would welcome that discussion, as less packages means less work for him. :)
Fair enough. I will start by filing a bug report against the gtk libs RyeBrye had problems with. I hope this solution works. If it does not, I will re-address this issue here.
I have filed over one-hundred and ten (110+) bugs for conflicts between ATrpms and FC/FE repositories. The tracker bug can be found here:
http://bugzilla.atrpms.net/show_bug.cgi?id=1028
So let's see what becomes of this. AT should be happy because this is probably over 50% of his packages, so his work load should be cut in half.
On 10/18/06, Tom 'spot' Callaway tcallawa@redhat.com wrote:
On Wed, 2006-10-18 at 11:02 -0700, Christopher Stone wrote:
On 10/18/06, Rex Dieter rdieter@math.unl.edu wrote:
Christopher Stone wrote:
I suggest that we have a comittee (possibly the packaging comittee) create a wiki page which reviews 3rd party repositories for such things as:
IMO, this is outside the scope of the packaging committee's mandate/rights/responsibility.
I'd suggest you start a new SIG if you feel so strongly about this subject.
I would be perfectly happy with this, even If its a one man SIG consisting of only me. As long as there is a wiki page on the Fedora wiki which has this information and users can be pointed to this page to learn about the consequences of installing another repository I would be happy.
However, I think it would be better to have a comittee review other repositories instead of a single person who might be biased such as myself.
I even think it would be more productive to highlight the FC or FE packages that atrpms is providing overrides for, and start a discussion around why these packages exist, and if there exists the possibility to merge the changes into the FC or FE package and retire the atrpms packages. I'm sure that Axel would welcome that discussion, as less packages means less work for him. :)
I have filed over one-hundred bugs a month ago, and while some Fedora users made an effort to try and reduce the conflicts, Axel has not made a single response to a single bug report.
May I now proceed with my original idea of making an official wiki page on fedoraproject.org which warns people of the potential dangers from using ATrpms and other items discussed above?
Christopher Stone schrieb:
[...] May I now proceed with my original idea of making an official wiki page on fedoraproject.org which warns people of the potential dangers from using ATrpms and other items discussed above?
No, don't mention ATrpms (or freshrpms, or livna) in the wiki. Not in good ways, but especially not in bad ways, as that would be a sign of bad politics. Axel is trying hard to contribute to Extras (at least until you scare him away with your campaign and then I'll be vary angry on you) and I'm quite confident that those problems you reported will get solved sooner or later, but it all takes time. Side note: I know that Axels time for Atrpms and Extras stuff is quite limited ATM, so please be even more patient.
CU thl
Christopher Stone wrote:
On 10/18/06, Tom 'spot' Callaway tcallawa@redhat.com wrote:
I even think it would be more productive to highlight the FC or FE packages that atrpms is providing overrides for, and start a discussion around why these packages exist, and if there exists the possibility to merge the changes into the FC or FE package and retire the atrpms packages. I'm sure that Axel would welcome that discussion, as less packages means less work for him. :)
I have filed over one-hundred bugs a month ago, and while some Fedora users made an effort to try and reduce the conflicts, Axel has not made a single response to a single bug report.
spot's suggestion was to start a (constructive) dialog regarding this issue. IMO, mass-filing formletter-type bugs is certainly no way to go about that, and I'm not a bit surprised by Axel's (non)response.
-- Rex
On 11/20/06, Rex Dieter rdieter@math.unl.edu wrote:
Christopher Stone wrote:
On 10/18/06, Tom 'spot' Callaway tcallawa@redhat.com wrote:
I even think it would be more productive to highlight the FC or FE packages that atrpms is providing overrides for, and start a discussion around why these packages exist, and if there exists the possibility to merge the changes into the FC or FE package and retire the atrpms packages. I'm sure that Axel would welcome that discussion, as less packages means less work for him. :)
I have filed over one-hundred bugs a month ago, and while some Fedora users made an effort to try and reduce the conflicts, Axel has not made a single response to a single bug report.
spot's suggestion was to start a (constructive) dialog regarding this issue. IMO, mass-filing formletter-type bugs is certainly no way to go about that, and I'm not a bit surprised by Axel's (non)response.
Using bugzilla for discussion is not the way to go?
Wow. This is mind boggling. Just how do you expect me to start a discussion? Is there a ATrpms mailing list or something? Even if there was a mailing list, why would bugzilla be less appropriate? I am totally dumbfounded...
On Mon, 2006-11-20 at 09:59 -0800, Christopher Stone wrote:
On 11/20/06, Rex Dieter rdieter@math.unl.edu wrote:
spot's suggestion was to start a (constructive) dialog regarding this issue. IMO, mass-filing formletter-type bugs is certainly no way to go about that, and I'm not a bit surprised by Axel's (non)response.
Using bugzilla for discussion is not the way to go?
Wow. This is mind boggling. Just how do you expect me to start a discussion? Is there a ATrpms mailing list or something? Even if there was a mailing list, why would bugzilla be less appropriate? I am totally dumbfounded...
Umm, how about filing a couple of bugs initially, instead of carpet-bombing him? It came off as very antagonist filing all those bugs, especially given how you started this whole discussion.
/B
Christopher Stone wrote:
Using bugzilla for discussion is not the way to go?
bugzilla is for *bugs*, not discussion (imo anyway).
Wow. This is mind boggling. Just how do you expect me to start a discussion? Is there a ATrpms mailing list or something?
Yes there is, several: http://lists.atrpms.net/mailman/listinfo/atrpms-users http://lists.atrpms.net/mailman/listinfo/atrpms-devel the latter may be the best place to start.
-- Rex
On 11/21/06, Rex Dieter rdieter@math.unl.edu wrote:
Christopher Stone wrote:
Using bugzilla for discussion is not the way to go?
bugzilla is for *bugs*, not discussion (imo anyway).
Wow. This is mind boggling. Just how do you expect me to start a discussion? Is there a ATrpms mailing list or something?
Yes there is, several: http://lists.atrpms.net/mailman/listinfo/atrpms-users http://lists.atrpms.net/mailman/listinfo/atrpms-devel the latter may be the best place to start.
I still don't see how this would be a better place than bugzilla. If you want to discuss why someone is forking a package, you have to do it on a per-package basis. Having a single thread on a mailing list discussing 100+ different packages is impractical and indeed starting 100 different threads on a mailing list is also impractical.
Christopher Stone wrote:
On 11/21/06, Rex Dieter rdieter@math.unl.edu wrote:
Christopher Stone wrote:
Using bugzilla for discussion is not the way to go?
bugzilla is for *bugs*, not discussion (imo anyway).
Wow. This is mind boggling. Just how do you expect me to start a discussion? Is there a ATrpms mailing list or something?
Yes there is, several: http://lists.atrpms.net/mailman/listinfo/atrpms-users http://lists.atrpms.net/mailman/listinfo/atrpms-devel the latter may be the best place to start.
I still don't see how this would be a better place than bugzilla. If you want to discuss why someone is forking a package, you have to do it on a per-package basis. Having a single thread on a mailing list discussing 100+ different packages is impractical and indeed starting 100 different threads on a mailing list is also impractical.
It's clear (to me anyway) your current approach has no realistic chance of getting results.
You have to start somewhere. I'd suggest initiating (constructive!) dialog one package at a time (or by similar pkgs in groups whatever).
-- Rex
I'm sorry to the fellow list members - I did stop defending against Christopher's FUD, but instead of a deescalation and a fade out, I now see him spreading FUD on websites, irc, several fedora lists, more bugzilla reports and probably any other means as well. It looks like a raging personal crusade. But since there are some very false statements placed by him into archives/bugzillas/google I need to rectify some of them at least.
On Wed, Oct 18, 2006 at 10:03:28AM -0700, Christopher Stone wrote:
As you are all probably intimately aware by now, there has been discussions about 3rd party repositories overriding base Fedora packages and creating their own distribution which causes very serious problems for users and damages the community.
I FIND IT *VERY* DISTRUBING THAT ONE OF THESE REPOSITORY MAINTAINERS IS ON THE FEDORA PACKAGING COMITTEE!
Here is an example IRC log that was taken just last night illustrating the problem:
Indeed! Perhaps you'll be surpised to find me agreeing that it illustrates the problem excellently!
20:40:49 RyeBrye | Trying to get yum to install gtk2-devel from the update repo I get 4 dependency errors 20:41:10 RyeBrye | the first one is gtk2 itself 20:41:17 XulChris | RyeBrye: caused by using atrpms no doubt [...] 20:43:42 RyeBrye | I'm relatively new to Linux - but it seems that the prevailing opinion is that atrpms = the devil 20:44:23 XulChris | RyeBrye: basically yes, we even attempted to tell AT this, but he refused to listen and accused of of using "FUD" techniques against his repository [...] 20:53:51 * RyeBrye is busy removing a whole lot of crap because atrpms installed a gtk2 package - and just about everything he installed from atrpms (mythtv) depended on it
Now let's see, someone has an issue with gtk. XulChris (Chistopher Stone?) immediately accuses ATrpms ("no doubt"!). The reporting user believes that w/o questioning since he got some valuable consulting from XulChris who seems to have great insights in this matter.
The culmination is the establishment that ATrpms managed to pull in gtk due to mythtv depending on in. Everybody then goes on in slantering ATrpms and Axel Thimm for doing such bad things.
Now to the simple two facts reverting the "problem"
a) mythtv is a qt application b) ATrpms has no gtk packages
The user probably has some random issue and ATrpms is a good scapegoat. So what is the "problem"? It is
o misinformation o 0 research o blind accusation o in short FUD
It's also self-generating FUD. Tomorrow some of the people in the IRC will hear of someone with problem in his say bind update. They will ask him whether he has ATrpms enables and advise him to reinstall from scratch. The next day the guy doing the last reinstallation will reenforce someone else that ATrpms is bad and that his power issues on the monitor are due to having installed ATrpms. I hope it doesn't end assuming that evils spreading from *.gov and *.mil (and other foreign equivalents) are due to some PC in there having ATrpms enabled. ...
So Chistopher Stone and people like him are not providing a solution to an alleged problem, but instead constitute themselfes great part of the problem. He is building walls where others try to build bridges.
Note: The main theme of the "issues by ATrpms" is the package replacement policy. The ones that are longer here around or are ATrpms' users will know that there *are* efforts (and results) to reduce the overlaps and to provide a repo-side solution. Certainly not due to Chistopher's FUD, though.
packaging@lists.fedoraproject.org
-
Axel Thimm -
Brian Pepple -
Christopher Stone -
Greg DeKoenigsberg -
Matthias Saou -
Rex Dieter -
Thorsten Leemhuis -
Tom Callaway