Greetings.
Recently FPC oked the use of rich boolean deps. However, we have run into an issue where the tools used to push updates are not able to correctly handle these new dependencies.
At today's FESCo meeting we decided to ask maintainers to not use rich boolean Requires/Recommends for the time being until tooling can catch up and allow us to push updates with them.
Thanks for your cooperation.
kevin -- https://fedoraproject.org/wiki/Packaging:Guidelines#Rich.2FBoolean_dependenc... https://fedorahosted.org/fpc/ticket/593 https://meetbot.fedoraproject.org/fedora-meeting/2016-04-01/fesco.2016-04-01...
_______________________________________________ devel-announce mailing list devel-announce@lists.fedoraproject.org http://lists.fedoraproject.org/admin/lists/devel-announce@lists.fedoraprojec...
On 04/01/2016 02:51 PM, Kevin Fenzi wrote:
Greetings.
Recently FPC oked the use of rich boolean deps. However, we have run into an issue where the tools used to push updates are not able to correctly handle these new dependencies.
At today's FESCo meeting we decided to ask maintainers to not use rich boolean Requires/Recommends for the time being until tooling can catch up and allow us to push updates with them.
Thanks for your cooperation.
kevin
https://fedoraproject.org/wiki/Packaging:Guidelines#Rich.2FBoolean_dependenc... https://fedorahosted.org/fpc/ticket/593 https://meetbot.fedoraproject.org/fedora-meeting/2016-04-01/fesco.2016-04-01...
To clarify, the use of rich dependencies is banned from use for Requires: and Recommends: dependencies. It is safe for use on Suggests:, Enhances: and Supplements:. In particular, this means that it is still acceptable to generate language subpackages using Supplements.
("Rich dependencies" refers to dependencies that specify "and" or "or" for extended rule-processing).
[1] https://fedoraproject.org/wiki/Packaging:Langpacks
_______________________________________________ devel-announce mailing list devel-announce@lists.fedoraproject.org http://lists.fedoraproject.org/admin/lists/devel-announce@lists.fedoraprojec...
On 04/01/2016 02:51 PM, Kevin Fenzi wrote:
Greetings.
Recently FPC oked the use of rich boolean deps. However, we have run into an issue where the tools used to push updates are not able to correctly handle these new dependencies.
At today's FESCo meeting we decided to ask maintainers to not use rich boolean Requires/Recommends for the time being until tooling can catch up and allow us to push updates with them.
Just to be clear.
Rich deps in Supplements are OK (yum parser ignores those).
And the language pack implementation needs no changes (uses rich deps in supplements).
On Fri, 1 Apr 2016 16:11:24 -0400 Carlos O'Donell carlos@redhat.com wrote:
On 04/01/2016 02:51 PM, Kevin Fenzi wrote:
Greetings.
Recently FPC oked the use of rich boolean deps. However, we have run into an issue where the tools used to push updates are not able to correctly handle these new dependencies.
At today's FESCo meeting we decided to ask maintainers to not use rich boolean Requires/Recommends for the time being until tooling can catch up and allow us to push updates with them.
Just to be clear.
Rich deps in Supplements are OK (yum parser ignores those).
And the language pack implementation needs no changes (uses rich deps in supplements).
Correct.
kevin
"KF" == Kevin Fenzi kevin@scrye.com writes:
KF> At today's FESCo meeting we decided to ask maintainers to not use KF> rich boolean Requires/Recommends for the time being until tooling KF> can catch up and allow us to push updates with them.
Could someone look over the language in https://fedoraproject.org/wiki/Packaging:Guidelines#Rich.2FBoolean_dependenc... to make certain that I've corrected it as necessary to meet the current FESCo guidance? I didn't want to leave incorrect information in the guidelines so I elected to fix this now rather than wait for an FPC meeting.
- J<
On Mon, 04 Apr 2016 18:07:50 -0500 Jason L Tibbitts III tibbs@math.uh.edu wrote:
"KF" == Kevin Fenzi kevin@scrye.com writes:
KF> At today's FESCo meeting we decided to ask maintainers to not use KF> rich boolean Requires/Recommends for the time being until tooling KF> can catch up and allow us to push updates with them.
Could someone look over the language in https://fedoraproject.org/wiki/Packaging:Guidelines#Rich.2FBoolean_dependenc... to make certain that I've corrected it as necessary to meet the current FESCo guidance? I didn't want to leave incorrect information in the guidelines so I elected to fix this now rather than wait for an FPC meeting.
That looks good to me.
kevin
On 2016-04-04, Jason L Tibbitts III tibbs@math.uh.edu wrote:
"KF" == Kevin Fenzi kevin@scrye.com writes:
KF> At today's FESCo meeting we decided to ask maintainers to not use KF> rich boolean Requires/Recommends for the time being until tooling KF> can catch up and allow us to push updates with them.
Could someone look over the language in https://fedoraproject.org/wiki/Packaging:Guidelines#Rich.2FBoolean_dependenc... to make certain that I've corrected it as necessary to meet the current FESCo guidance?
This changes meaning regarding to F25. Previous text banned rich strong dependencies in F24 only. This current text extends the ban to all Fedoras.
Is that intentional?
-- Petr
On 6 Apr 2016 09:19, "Petr Pisar" ppisar@redhat.com wrote:
On 2016-04-04, Jason L Tibbitts III tibbs@math.uh.edu wrote:
> "KF" == Kevin Fenzi kevin@scrye.com writes:
KF> At today's FESCo meeting we decided to ask maintainers to not use KF> rich boolean Requires/Recommends for the time being until tooling KF> can catch up and allow us to push updates with them.
Could someone look over the language in
https://fedoraproject.org/wiki/Packaging:Guidelines#Rich.2FBoolean_dependenc...
to make certain that I've corrected it as necessary to meet the current FESCo guidance?
This changes meaning regarding to F25. Previous text banned rich strong dependencies in F24 only. This current text extends the ban to all Fedoras.
Is that intentional?
The issue is it's breaking the mash which indeed affects all Fedora versions.
Until the tooling is moved from yum to dnf then it's not safe to use the booleans in requires or recommends.
The reason suggest is left out is that yum currently recognise that tag (but it does the others) so it doesn't cause mash to fail.
"PP" == Petr Pisar ppisar@redhat.com writes:
PP> This changes meaning regarding to F25. Previous text banned rich PP> strong dependencies in F24 only. This current text extends the ban PP> to all Fedoras.
PP> Is that intentional?
It's currently correct according to FESCo's request as I understood it. While f25 might still be able to compose, if the tools aren't fixed before the next branch then we'll have to back things out or end up back in a situation where we can't mash updates.
- J<
On 04/06/2016 04:49 AM, Jason L Tibbitts III wrote:
"PP" == Petr Pisar ppisar@redhat.com writes:
PP> This changes meaning regarding to F25. Previous text banned rich PP> strong dependencies in F24 only. This current text extends the ban PP> to all Fedoras.
PP> Is that intentional?
It's currently correct according to FESCo's request as I understood it. While f25 might still be able to compose, if the tools aren't fixed before the next branch then we'll have to back things out or end up back in a situation where we can't mash updates.
I'd say it's fine for now. Our *hope* is that we will be able to support them for F25 at some point, but composes *right now* probably don't work, so I think this wording is fine.
On Thu, Apr 7, 2016 at 7:37 AM, Stephen Gallagher sgallagh@redhat.com wrote:
On 04/06/2016 04:49 AM, Jason L Tibbitts III wrote:
> "PP" == Petr Pisar ppisar@redhat.com writes:
PP> This changes meaning regarding to F25. Previous text banned rich PP> strong dependencies in F24 only. This current text extends the ban PP> to all Fedoras.
PP> Is that intentional?
It's currently correct according to FESCo's request as I understood it. While f25 might still be able to compose, if the tools aren't fixed before the next branch then we'll have to back things out or end up back in a situation where we can't mash updates.
I'd say it's fine for now. Our *hope* is that we will be able to support them for F25 at some point, but composes *right now* probably don't work, so I think this wording is fine.
So you don't expect that mash could be fixed/replaced before the F24 release in June, then?
On Thu, 7 Apr 2016 07:54:36 -0400 Neal Gompa ngompa13@gmail.com wrote:
On Thu, Apr 7, 2016 at 7:37 AM, Stephen Gallagher sgallagh@redhat.com wrote:
On 04/06/2016 04:49 AM, Jason L Tibbitts III wrote:
>> "PP" == Petr Pisar ppisar@redhat.com writes:
PP> This changes meaning regarding to F25. Previous text banned PP> rich strong dependencies in F24 only. This current text PP> extends the ban to all Fedoras.
PP> Is that intentional?
It's currently correct according to FESCo's request as I understood it. While f25 might still be able to compose, if the tools aren't fixed before the next branch then we'll have to back things out or end up back in a situation where we can't mash updates.
I'd say it's fine for now. Our *hope* is that we will be able to support them for F25 at some point, but composes *right now* probably don't work, so I think this wording is fine.
So you don't expect that mash could be fixed/replaced before the F24 release in June, then?
mash is going to go away soon. Things are moving to koji signed repos (where koji does the repo creation, etc).
However, that is still using yum, so it would need to be ported.
If someone would like to help out and work on that, that would be great!
However, I don't see anyone in releng having the cycles to do so right now. They are busy working on getting rawhide composing again at the moment.
So, no, absent some resources I don't see this being fixed before f24.
kevin
From: "Kevin Fenzi" kevin@scrye.com On Thu, 7 Apr 2016 07:54:36 -0400 Neal Gompa ngompa13@gmail.com wrote:
On Thu, Apr 7, 2016 at 7:37 AM, Stephen Gallagher sgallagh@redhat.com wrote:
On 04/06/2016 04:49 AM, Jason L Tibbitts III wrote:
>>> "PP" == Petr Pisar ppisar@redhat.com writes:
PP> This changes meaning regarding to F25. Previous text banned PP> rich strong dependencies in F24 only. This current text PP> extends the ban to all Fedoras.
PP> Is that intentional?
It's currently correct according to FESCo's request as I understood it. While f25 might still be able to compose, if the tools aren't fixed before the next branch then we'll have to back things out or end up back in a situation where we can't mash updates.
I'd say it's fine for now. Our *hope* is that we will be able to support them for F25 at some point, but composes *right now* probably don't work, so I think this wording is fine.
So you don't expect that mash could be fixed/replaced before the F24 release in June, then?
I've talked to relengs and port of mash [1] will not happen any time soon. In best case it should make composes for F25.
This is more complicated as pungi, mash and signed repos are linked together and the replacement of whole releng tools will not happen in a day. They plan bigger refactoring during the migration of these tools and replace them one by one. We are trying to allocate resources to make this done.
Pungi-5 is almost ready to replace current yum based version and making a library from mash using DNF would probably take 1 man-month.
Honza
On Thursday, April 7, 2016 7:54:36 AM CDT Neal Gompa wrote:
On Thu, Apr 7, 2016 at 7:37 AM, Stephen Gallagher sgallagh@redhat.com
wrote:
On 04/06/2016 04:49 AM, Jason L Tibbitts III wrote:
>> "PP" == Petr Pisar ppisar@redhat.com writes:
PP> This changes meaning regarding to F25. Previous text banned rich PP> strong dependencies in F24 only. This current text extends the ban PP> to all Fedoras.
PP> Is that intentional?
It's currently correct according to FESCo's request as I understood it. While f25 might still be able to compose, if the tools aren't fixed before the next branch then we'll have to back things out or end up back in a situation where we can't mash updates.
I'd say it's fine for now. Our *hope* is that we will be able to support them for F25 at some point, but composes *right now* probably don't work, so I think this wording is fine.
So you don't expect that mash could be fixed/replaced before the F24 release in June, then?
All of the replacements use yum also. The issues may be due to rhel7'a rpm not supporting boolean deps.It is not a small problem to fix and it needs some people to sit down and figure out how to best deal with them.
https://fedoraproject.org/wiki/ReleaseEngineering/PriorityPipeline is some of what we have on our plate to deal with
Dennis
From: "Dennis Gilmore" dennis@ausil.us
On Thursday, April 7, 2016 7:54:36 AM CDT Neal Gompa wrote:
On Thu, Apr 7, 2016 at 7:37 AM, Stephen Gallagher sgallagh@redhat.com
wrote:
On 04/06/2016 04:49 AM, Jason L Tibbitts III wrote:
>>> "PP" == Petr Pisar ppisar@redhat.com writes:
PP> This changes meaning regarding to F25. Previous text banned rich PP> strong dependencies in F24 only. This current text extends the ban PP> to all Fedoras.
PP> Is that intentional?
It's currently correct according to FESCo's request as I understood it. While f25 might still be able to compose, if the tools aren't fixed before the next branch then we'll have to back things out or end up back in a situation where we can't mash updates.
I'd say it's fine for now. Our *hope* is that we will be able to support them for F25 at some point, but composes *right now* probably don't work, so I think this wording is fine.
So you don't expect that mash could be fixed/replaced before the F24 release in June, then?
All of the replacements use yum also. The issues may be due to rhel7'a rpm not supporting boolean deps.It is not a small problem to fix and it needs some people to sit down and figure out how to best deal with them.
We've built whole DNF stack supporting rich dependencies for this purpose. We are in touch with relengs. Hopefully we'll make this done to F25.
Honza
On Thursday, April 14, 2016 11:34:08 AM CDT Honza Šilhan wrote:
From: "Dennis Gilmore" dennis@ausil.us
On Thursday, April 7, 2016 7:54:36 AM CDT Neal Gompa wrote:
On Thu, Apr 7, 2016 at 7:37 AM, Stephen Gallagher sgallagh@redhat.com
wrote:
On 04/06/2016 04:49 AM, Jason L Tibbitts III wrote:
>>>> "PP" == Petr Pisar ppisar@redhat.com writes:
PP> This changes meaning regarding to F25. Previous text banned rich PP> strong dependencies in F24 only. This current text extends the ban PP> to all Fedoras.
PP> Is that intentional?
It's currently correct according to FESCo's request as I understood it. While f25 might still be able to compose, if the tools aren't fixed before the next branch then we'll have to back things out or end up back in a situation where we can't mash updates.
I'd say it's fine for now. Our *hope* is that we will be able to support them for F25 at some point, but composes *right now* probably don't work, so I think this wording is fine.
So you don't expect that mash could be fixed/replaced before the F24 release in June, then?
All of the replacements use yum also. The issues may be due to rhel7'a rpm not supporting boolean deps.It is not a small problem to fix and it needs some people to sit down and figure out how to best deal with them.
We've built whole DNF stack supporting rich dependencies for this purpose. We are in touch with relengs. Hopefully we'll make this done to F25.
Honza
I am afraid you are not in touch with relengs, I have not had any discussions with you.
Dennis
On 1. 4. 2016 at 12:51:31, Kevin Fenzi wrote:
Greetings.
Recently FPC oked the use of rich boolean deps. However, we have run into an issue where the tools used to push updates are not able to correctly handle these new dependencies.
At today's FESCo meeting we decided to ask maintainers to not use rich boolean Requires/Recommends for the time being until tooling can catch up and allow us to push updates with them.
Thanks for your cooperation.
kevin
https://fedoraproject.org/wiki/Packaging:Guidelines#Rich.2FBoolean_dependenc ies https://fedorahosted.org/fpc/ticket/593 https://meetbot.fedoraproject.org/fedora-meeting/2016-04-01/fesco.2016-04-01 -17.00.log.html
Is there an estimate as to when the tooling (doesn't matter if it's mash or pungi) might be ready? Looking at the IRC log, it doesn't say what is the plan.
Thanks, Jan
On Tue, 05 Apr 2016 08:27:36 +0200 Jan Zelený jzeleny@redhat.com wrote:
Is there an estimate as to when the tooling (doesn't matter if it's mash or pungi) might be ready? Looking at the IRC log, it doesn't say what is the plan.
I have no idea. :)
The two things that likely should be looked at are pungi and now the signed koji repos (which bodhi/updates are going to switch to before too long from mash).
Are there any resources that might help with this work?
pungi is at: https://pagure.io/pungi
signed repos is still being worked on and will be in koji: https://pagure.io/koji/pull-requests
kevin
On 5. 4. 2016 at 10:08:20, Kevin Fenzi wrote:
On Tue, 05 Apr 2016 08:27:36 +0200
Jan Zelený jzeleny@redhat.com wrote:
Is there an estimate as to when the tooling (doesn't matter if it's mash or pungi) might be ready? Looking at the IRC log, it doesn't say what is the plan.
I have no idea. :)
The two things that likely should be looked at are pungi and now the signed koji repos (which bodhi/updates are going to switch to before too long from mash).
Are there any resources that might help with this work?
pungi is at: https://pagure.io/pungi
signed repos is still being worked on and will be in koji: https://pagure.io/koji/pull-requests
I will ask around and see what I can do.
Thanks, Jan