Re: [gentoo-dev] Slacker arches
On 1/25/11 1:30 PM, Markos Chandras wrote: QA is not a solution to everything. The problem Tomas is trying to counter here is the idle/slacking arches. If the arch is active but have some concerns regarding the stabilization then let the maintainer deal with it. This is the way we do it now anyway Also, we should have someone to check for stale stabilization bugs. I'm not sure if all reporters are able to take care of that, especially if they have a lot of bugs open. Thats really their problem. Arches can always remove themselves from the bugs. No need to care about stale bugs. If the maintainers don't care then we(arches) don't care. I was mostly thinking about cases like https://bugs.gentoo.org/329633 where indeed arches remove themselves from the bug, but there is a dispute between them and the maintainer about the correct course of action. The usual conflict resolution procedure would be to contact the team lead, and eventually the council. However, I'm not sure whether that would be optimal for stabilization bugs. Paweł signature.asc Description: OpenPGP digital signature
[gentoo-dev] Slacker arches
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi, Given the talk on last council meeting I would like this policy to be in effect over main tree: Every arch teams should stabilise OR write out reason why they can't do so to stable bug in 90 days. If any arch team fails to do so the maintainer can decide to drop their keywords to testing. Given depgraph breakages the maintainer can coordinate with the QA team to drop all reverse dependencies to testing too. Only exception from this rule are toolchain and base-system bugs, since testing-only effectively means that the arch lost stable status as whole. In order to accommodate this goals Arch Teams can generate Arch Testers which can comment on the bugs in their name, where maintainer then can act upon their comments (eg: if ARM at say that everything is ok, maintainer can stabilise it on arm...). (Come on for most stable testing you don't really need to be fully fledged Gentoo dev, but you just need the named hardware and working brain :)) Cheers Tom -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.17 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk0+thsACgkQHB6c3gNBRYfW4wCfXbctXUgKoK53Pd45QuAMgY7r Sy4AoMU6z/oS/JvUum6/29SHYsmuoQBs =LQj9 -END PGP SIGNATURE-
Re: [gentoo-dev] Slacker arches
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Dne 25.1.2011 12:38, Tomáš Chvátal napsal(a): Hi, Given the talk on last council meeting I would like this policy to be in effect over main tree: Every arch teams should stabilise OR write out reason why they can't do so to stable bug in 90 days. If any arch team fails to do so the maintainer can decide to drop their keywords to testing. Given depgraph breakages the maintainer can coordinate with the QA team to drop all reverse dependencies to testing too. Only exception from this rule are toolchain and base-system bugs, since testing-only effectively means that the arch lost stable status as whole. In order to accommodate this goals Arch Teams can generate Arch Testers which can comment on the bugs in their name, where maintainer then can act upon their comments (eg: if ARM at say that everything is ok, maintainer can stabilise it on arm...). (Come on for most stable testing you don't really need to be fully fledged Gentoo dev, but you just need the named hardware and working brain :)) Cheers Tom AAnd as Mark told me I actually didn't say that i want feedback and opinions on those two mails i just sent to this ML, so please tell us how do you feel about it and what would you like to be done differently :) These two mails (slacker arches, eapi usage) are not policies in effect but stuff council would like to decide about and want to know the devhood opinions. Cheers Tom -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.17 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk0+vJMACgkQHB6c3gNBRYc4+wCeLPeysLA/xTacnofptQBbai5z jpEAn0jyipxEV/U/IQylCmzj3IVbe3NZ =LHQi -END PGP SIGNATURE-
Re: [gentoo-dev] Slacker arches
On 1/25/11 12:38 PM, Tomáš Chvátal wrote: Every arch teams should stabilise OR write out reason why they can't do so to stable bug in 90 days. If any arch team fails to do so the maintainer can decide to drop their keywords to testing. Given depgraph breakages the maintainer can coordinate with the QA team to drop all reverse dependencies to testing too. Seconded. Reality++ Be prepared for some issues though. Sometimes maintainers don't agree with reasons arch teams provide that block stabilizations. In those cases, who makes the decision? My suggestion is QA. Also, we should have someone to check for stale stabilization bugs. I'm not sure if all reporters are able to take care of that, especially if they have a lot of bugs open. Paweł signature.asc Description: OpenPGP digital signature
Re: [gentoo-dev] Slacker arches
On Tue, Jan 25, 2011 at 01:20:29PM +0100, Paweł Hajdan, Jr. wrote: On 1/25/11 12:38 PM, Tomáš Chvátal wrote: Every arch teams should stabilise OR write out reason why they can't do so to stable bug in 90 days. If any arch team fails to do so the maintainer can decide to drop their keywords to testing. Given depgraph breakages the maintainer can coordinate with the QA team to drop all reverse dependencies to testing too. Seconded. Reality++ Be prepared for some issues though. Sometimes maintainers don't agree with reasons arch teams provide that block stabilizations. In those cases, who makes the decision? My suggestion is QA. QA is not a solution to everything. The problem Tomas is trying to counter here is the idle/slacking arches. If the arch is active but have some concerns regarding the stabilization then let the maintainer deal with it. This is the way we do it now anyway Also, we should have someone to check for stale stabilization bugs. I'm not sure if all reporters are able to take care of that, especially if they have a lot of bugs open. Paweł Thats really their problem. Arches can always remove themselves from the bugs. No need to care about stale bugs. If the maintainers don't care then we(arches) don't care. Regards, -- Markos Chandras (hwoarang) Gentoo Linux Developer Key ID: B4AFF2C2 Key FP: 660A 0742 84EC 26F1 9EDB F00A FA83 5A15 B4AF F2C2 pgpctzIsQqDcN.pgp Description: PGP signature
Re: [gentoo-dev] Slacker arches
On Tue, 25 Jan 2011 12:38:03 +0100, Tomáš Chvátal wrote: Only exception from this rule are toolchain and base-system bugs, since In both threads you recently started, you used the term base-system bugs but I think you mean @system packages - there are a ton of base-system packages that are not critical and wouldn't apply to either of your threads. Is this an accurate assumption on my part here? -Jeremy
Re: [gentoo-dev] Slacker arches
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Dne 25.1.2011 16:49, Jeremy Olexa napsal(a): On Tue, 25 Jan 2011 12:38:03 +0100, Tomáš Chvátal wrote: Only exception from this rule are toolchain and base-system bugs, since In both threads you recently started, you used the term base-system bugs but I think you mean @system packages - there are a ton of base-system packages that are not critical and wouldn't apply to either of your threads. Is this an accurate assumption on my part here? -Jeremy Yeah you are right :P mea culpa maxima -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.17 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk0+8p4ACgkQHB6c3gNBRYfCHQCggR3NiP+1R/vhHU/tAHSzJC2p ZhAAn0VHw3HaadJstoTpLgLeYG9HL63m =HLLa -END PGP SIGNATURE-
