Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Am Mittwoch, 25. April 2007 schrieb Ciaran McCreesh: On Tue, 24 Apr 2007 12:31:48 -0700 Robin H. Johnson [EMAIL PROTECTED] wrote: printf _rc%d%04d%02d%02d,$RC,$YEAR,$MONTH,$DAY Funnily enough... If we're going by PMS drafts, that's illegal whereas multiple suffixes are legal. PMS permits multiple suffixes, but limits any individual version component to eight digits to avoid problems with integer overflows, floating point precision etc. My point was to avoid providigin existing practice which might need to be respected by either PMS or tree policy. Danny -- Danny van Dyk [EMAIL PROTECTED] Gentoo/AMD64 Project, Gentoo Scientific Project -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Am Mittwoch, 25. April 2007 schrieb Ciaran McCreesh: On Tue, 24 Apr 2007 12:31:48 -0700 Robin H. Johnson [EMAIL PROTECTED] wrote: printf _rc%d%04d%02d%02d,$RC,$YEAR,$MONTH,$DAY Funnily enough... If we're going by PMS drafts, that's illegal whereas multiple suffixes are legal. PMS permits multiple suffixes, but limits any individual version component to eight digits to avoid problems with integer overflows, floating point precision etc. And when PMS specifies that together with a proper way to compare multiple suffixes there will be no problem. This Council decission was to avoid 'existing practice' that might be necessary to include in PMS. Danny -- Danny van Dyk [EMAIL PROTECTED] Gentoo/AMD64 Project, Gentoo Scientific Project -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Ciaran McCreesh wrote: On Tue, 24 Apr 2007 12:31:48 -0700 Robin H. Johnson [EMAIL PROTECTED] wrote: printf _rc%d%04d%02d%02d,$RC,$YEAR,$MONTH,$DAY Funnily enough... If we're going by PMS drafts, that's illegal whereas multiple suffixes are legal. PMS permits multiple suffixes, but limits any individual version component to eight digits to avoid problems with integer overflows, floating point precision etc. Give that all we need for mplayer is a date (as in mmdd) I think we could come up with a good interim workaround. I'd like to have multiple suffixes restored anyway... lu -- Luca Barbato Gentoo/linux Gentoo/PPC http://dev.gentoo.org/~lu_zero -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Wed, 25 Apr 2007 09:57:39 +0200 Danny van Dyk [EMAIL PROTECTED] wrote: Funnily enough... If we're going by PMS drafts, that's illegal whereas multiple suffixes are legal. PMS permits multiple suffixes, but limits any individual version component to eight digits to avoid problems with integer overflows, floating point precision etc. And when PMS specifies that together with a proper way to compare multiple suffixes there will be no problem. PMS *does* specify a proper way of comparing multiple version suffixes (and version specs with a leading zero for that matter). I'm not particularly happy with the wording, but as far as I can see the description is at least correct, even if it isn't clear. -- Ciaran McCreesh signature.asc Description: PGP signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Seemant Kulleen wrote: If I were to guess I'd say people are a little confused that this required action/decision this quickly and outside of a regular council meeting -- for a real emergency situation, you'd probably see a lot less of a hub-bub about it. But, come on, this is a 3-package issue. Perhaps they wanted to make sure it remained a 3-package issue, and thought that it might grow before it could be addressed? This all seems a bit like a tempest in a teapot to me, and I don't mean to single out any individual's contribution to this discussion. The council has stated that multiple version suffixes are to be avoided. I doubt they're going to suspend any developer who hasn't cleaned up their packages by Friday. I'm sure they're happy to see discussion on - -dev regarding pros and cons of various ways of implementing this change before it happens, and in the meantime new packages going into portage will be mindful of the policy from the start. If a particular package needs a month to sort out some really messy issue I'm sure the maintainer would be treated reasonably if they simply emailed a council member about it. Gentoo is a community, and sometimes people in a community don't always agree. Somebody has to make a decision, and we can't make every hill the one we're willing to die on. The council will generally represent the majority opinion of developers, simply due to the fact that it is an elected body. Sometimes in a community cohesiveness is more important than productivity, because it is the ability to mobilize hordes of developers that matters more than the contributions of any individual. Sometimes that means an elegant solution to a problem gets put on the back burner for a few months. Sometimes that means that a developer who is unusually productive is asked to cool down a little. Personally, my feeling has always been that if you want to avoid politics then avoid doing things that create political messes (flamewars, heated discussion, etc). If you disagree passionately with somebody about something, try having a private email conversation where both of you can let down your guard and try to understand each other's concerns. And then don't go quoting each other all over public lists to bolster your arguments... Remember, everybody is here to make Gentoo a better product, because we're all users as well as contributors. When users file poorly-worded bugs they're just trying to help, and when some developer makes an idiotic decision they probably think they're doing the right thing. That means that they're going to be automatically predisposed to helping you out if you just ask nicely. When posting as a user in bugzilla I've been flamed more than a few times (and not just on gentoo). I just try to be as polite as I can - after all I'm the one asking for a little help and somebody else is taking the time to help me. After all, it doesn't cost me anything, and down the road it could pay dividends. The same applies in reverse - being nice to others doesn't really cost you anything, and you never know when you will need their help down the road... -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGLzbBG4/rWKZmVWkRAkjoAKCZnydd8Y6ZFVVIbz5sh/0sryuxoQCeMFh7 sQ+Icf4GdB1dlEezRxdgvpM= =hpaU -END PGP SIGNATURE- smime.p7s Description: S/MIME Cryptographic Signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
As usual if you have issues with the council's decision, this is the wrong list to complain on. Try [EMAIL PROTECTED], I here they have popcorn. This is the right list to discuss versioning schemes though. -Alec -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, 2007-04-24 at 21:25 -0400, Seemant Kulleen wrote: On Wed, 2007-04-25 at 00:30 +0200, Danny van Dyk wrote: In my eyes it was a policy issue. Tree-wide policies have to pass the council in one form or the other. So why shouldn't Council care here? My argument is not that Council should not care. My question is: what's the big urgency to rush a half-baked policy through? Except that nobody did that. Read what was done. What was done was a *temporary* block on something that needed further discussion was put in place. Nobody held any emergency meeting. A subset of the Council just used some common sense and said something like hey, maybe we should block this until there is proper discussion and a proper solution is found which makes complete sense to me. I wasn't even involved in the situation and I can see how this happened. As I said, anyone who cannot see just how simple of a thing this was is either blind or specifically looking for something to complain about. I just wonder why several people feel attacked by this decission while the affected parties have no problem with it. I hope you don't mean me here, because I haven't felt attacked at all. My concern isn't a personal one. Rather, it's a question that nobody from the council has actually answered: what was the big hurry to make a decision _NOW_ without even thinking through the migration path, or for that matter without even knowing what is the actual correct way. It's fine to say that _rc_alpha_beta_p is wrong (and I happen to agree). It's another to not say what is actually right. Furthermore, if only 3 packages did the wrong thing where was the emergency? There was no emergency. Nobody from the Council has ever said it was an emergency. I think you were the one that stated that it was. Also, realize that the decision wasn't a solution to the problem. Again, nobody said that it was. The only problem that I see is that we didn't act soon enough. As soon as there was some conflict on how to allow the multiple version suffixes, somebody should have stopped any packages form using them in the tree until a solution was decided. I'm not trying to make you defensive, I just really would like an answer to my question, that is all. I've answered it to the best of my ability and it is hard to not get defensive when every decision your group makes is attacked on multiple fronts by people that put you in the position to make those exact same decisions. It really has made me wonder what the point in being on the Council is if we can't do anything without being assaulted on all sides. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Wed, 2007-04-25 at 07:08 -0400, Richard Freeman wrote: Perhaps they wanted to make sure it remained a 3-package issue, and thought that it might grow before it could be addressed? Exactly. I agree with the rest of what you've said, also. Being on the Council is a thankless job where we try our best to do what's best for Gentoo as a whole. This means that we *will* end up making some decision at some time that you might not agree with. This *is* going to happen. Hell, there have been Council decisions made that *I* don't agree with, but you don't see me running around acting like the Council is doing something wrong when they're doing their job. Honestly, all this has done is made me not want to make any decisions, which will turn us into the previous leadership, which I don't blame for the inaction from their group, as that was all they were given. -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
@council; cross posting to provide the reasoning, if discussion continues on council ml, kindly cc me (unsubscribed long ago). Technical discussion (which should be the basis of why it was banned should be on dev ml imo). On Tue, Apr 24, 2007 at 09:11:44PM +0200, Danny van Dyk wrote: Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. Rules for 'appealing' are a wee bit sparse, but consider this email an appeal to reopen the issue at the next council meeting (and a suggestion to figure out what appealing requires/involves). Offhand, while there has been sqawking, the functionality has been available for over a year (first 2.1 release of portage), pkgcore has long term supported it, paludis will support it in next released version (it's in trunk at least), PMS has the basic comparison rules doc'd out in addition. As others have said, but reiterating in this message- the only 'recent' change for multi-suffix is unlocking it in repoman so folks could use it; nature of backwards compatibility, the support had to be left locked for 6 months to preclude issues from stage releases, only change this side of 2007 was unlocking it. Meanwhile, bug involved which is basically resolved at this point- http://bugs.gentoo.org/166522 If the intention of the subset was to limit things till the allowed permutations of multi-suffix are worked out, please clarify- at least what I've seen thread wise, haven't seen a real explanation for it beyond multi-suffix is icky and robbat2 has a hackish alternative :) This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: Please expand further on this one- no offense meant, but the offered reason is slightly weasely in that it's not really saying anything, what it is saying is pretty obfuscated. Best I can figure, the offered reason is it needs to be blocked before it becomes widespread thus cannot be blocked any further- which isn't much of a reason since the support is long term there already, and doesn't state *why* it needs to be blocked (just states it needs to be blocked). I'm not a mind reader, so lets just assume I'm misreading it. Either way, feel free to expound on the 'why' (either ml or via council appeal). An illegal version specification of media-sound/alsa-driver has already been removed from the tree. I would like to ask the affected package maintainers to move these versions to sane version specifications as soon as possible. Thanks in advance for this. In the future when a subset (or full council, whatever) decides to ban functionality such as this, strongly suggest they ban *further* usage of it- implicit there is that the existing usage is left alone till a full decision can be reached. Y'all banned all usage of it, meaning people have to make changes now. Reasoning is pretty simple; at least for the two versions above, via making it illegal it forces them to transition to a hasty versioning scheme that may (frankly) suck- such as robbats proposal (his proposal works, but it's not human friendly and frankly serves more as a demonstration of why multi-suffix is useful). Joking aside, if the intention is to block further usage till the permutations allowed are ironed out, fair enough- would strongly suggest not decreeing they've got to go now when you're stating in the same breath the decision will (effectively) be revisited a few weeks later. Especially since changes to the versioning scheme can be a royal pain in the ass transitioning away from afterwards. ~harring pgpdYeEDyq3bu.pgp Description: PGP signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Ciaran McCreesh wrote: On Wed, 25 Apr 2007 09:57:39 +0200 Danny van Dyk [EMAIL PROTECTED] wrote: Funnily enough... If we're going by PMS drafts, that's illegal whereas multiple suffixes are legal. PMS permits multiple suffixes, but limits any individual version component to eight digits to avoid problems with integer overflows, floating point precision etc. And when PMS specifies that together with a proper way to compare multiple suffixes there will be no problem. PMS *does* specify a proper way of comparing multiple version suffixes (and version specs with a leading zero for that matter). I'm not particularly happy with the wording, but as far as I can see the description is at least correct, even if it isn't clear. Alright guys, This is enough. PMS is a work in progress its not going to cover everything that users and developers are going to be in some cases boneheaded enough to actually pull off (always have edge conditions). We're continuing to downgrade here and quite frankly the discussions seem be getting into tangents more then the actual topic at hand (you know...the fact about what the proper suffix format is), and that is up to the council to decide. If you have issues with the council, bring it up in the proper channel, as others have mentioned where its at. Now either get it back on topic, take it to private emails to discuss between yourselves, or take up the issues that relate to the council, to the councils mailing-list/members. They are actually you know...alive and willing to talk to you. Annoyed proctor out signature.asc Description: OpenPGP digital signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Wed, 25 Apr 2007 10:06:55 -0700 Joshua Jackson [EMAIL PROTECTED] wrote: This is enough. PMS is a work in progress its not going to cover everything that users and developers are going to be in some cases boneheaded enough to actually pull off (always have edge conditions). No no, you miss the point. If developers are doing something, either PMS needs to allow it or they have to stop doing it. It's entirely relevant to the topic at hand. -- Ciaran McCreesh signature.asc Description: PGP signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Hi. Ciaran McCreesh wrote: On Wed, 25 Apr 2007 10:06:55 -0700 Joshua Jackson [EMAIL PROTECTED] wrote: This is enough. PMS is a work in progress its not going to cover everything that users and developers are going to be in some cases boneheaded enough to actually pull off (always have edge conditions). No no, you miss the point. If developers are doing something, either PMS needs to allow it or they have to stop doing it. It's entirely relevant to the topic at hand. I agree. Also, this issue has arisen from a change in current policy. Even if Portage and repoman now allow the use of multiple suffixes, the devmanual still states that's illegal - so it's illegal in current policy. Instead of people arguing about a decision to uphold the current policy, I think they should be asking that we have a discussion about the current policy and propose alternatives, like is being done on the bug, and in the end submit it to the council for a voting. - -- Regards, Jorge Vicetto (jmbsvicetto) - jmbsvicetto at gentoo dot org Gentoo-forums / Userrel / Proctors -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.3 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGL5HTcAWygvVEyAIRAvNsAJ9FFkIWUbLjmsBHskfaxZbN0Fo7LgCgk5o9 UBuUR5erFfG3rFEktEhNiJ8= =r7Pd -END PGP SIGNATURE- -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Ciaran McCreesh wrote: On Tue, 24 Apr 2007 15:16:38 -0400 Doug Goldstein [EMAIL PROTECTED] wrote: This wouldn't have to be because you have a vested interest in paludis and paludis does not support this syntax and there happens to be no reasonable way to support that? Cut the conspiracy theories. Paludis will support whatever PMS says it should support. Released versions supported what PMS said at that time (which went in line with the Portage documentation), and the next release will support whatever PMS says then (which currently goes against the Portage documentation, but along with Portage behaviour). Ciaran, You missed the bandwagon on trying to use the conspiracy theories phrase already. That happened a full 24 hrs ago. I'm sorry you were off-line. Next time try to come to the party on time, otherwise keep quiet. -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Ciaran, You missed the bandwagon on trying to use the conspiracy theories phrase already. That happened a full 24 hrs ago. I'm sorry you were off-line. Next time try to come to the party on time, otherwise keep quiet. Already been handled as its offtopic, please just let this one drop. signature.asc Description: OpenPGP digital signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Wed, 25 Apr 2007 09:56:02 -0700 Brian Harring [EMAIL PROTECTED] wrote: Best I can figure, the offered reason is it needs to be blocked before it becomes widespread thus cannot be blocked any further- which isn't much of a reason since the support is long term there already, and doesn't state *why* it needs to be blocked (just states it needs to be blocked). It's better stated as we need to put a hold on this so that a reasoned discussion can be had, and a decision made, before use becomes so widespread as to force the issue regardless of what is decided on technical merits. -- [EMAIL PROTECTED] mailing list
[gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: media-viode/mplayer-1.0_rc2_pre20070321-r4 media-video/transcode-1.0.3_rc2_p20070310-r1 An illegal version specification of media-sound/alsa-driver has already been removed from the tree. I would like to ask the affected package maintainers to move these versions to sane version specifications as soon as possible. Thanks in advance for this. Danny -- Danny van Dyk [EMAIL PROTECTED] Gentoo/AMD64 Project, Gentoo Scientific Project -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Am Dienstag, 24. April 2007 schrieb Danny van Dyk:
Hi all,
[CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th,
2007]
A subset of council members decided today that multiple version
suffixes are illegal in the tree pending further notice. This
decission can be appealed at the next Council meeting. If there is
sufficient public demand, an earlier meeting can be held.
This decission has been made to prevent sufficient precedence for
unilateral changes to the tree structure. So far the following
package versions are considered illegal:
media-viode/mplayer-1.0_rc2_pre20070321-r4
media-video/transcode-1.0.3_rc2_p20070310-r1
As requested by Daniel (dsd) on irc, let me state what is wrong with
these versions:
All upstream version suffixes may only be used once. This doesn't affect
the -r1 (ebuild revision) suffix, as that is no upstream suffix but
internal to Gentoo's versioning scheme only.
Examples:
* _alphaX_betaY - illegal
* _rcX_preY - illegal
* _alphaX_preY - illegal
* ...
* _{rc,alpha,beta,...}-rX - legal
The rationale behind this is the following:
* certain combinations of suffixes don't make sense.
* only recent Portage versions support it.
If this feature should be allowed again then we need to document a
sensible subset of suffix-combinations prior to adding them to the
tree.
Hope that clarifies it a bit more :-)
Danny
--
Danny van Dyk [EMAIL PROTECTED]
Gentoo/AMD64 Project, Gentoo Scientific Project
--
[EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Danny van Dyk wrote: Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: media-viode/mplayer-1.0_rc2_pre20070321-r4 media-video/transcode-1.0.3_rc2_p20070310-r1 An illegal version specification of media-sound/alsa-driver has already been removed from the tree. I would like to ask the affected package maintainers to move these versions to sane version specifications as soon as possible. Thanks in advance for this. Danny So apparently as little as 1 council member can make a decision and it be binding unless appealed to the entire council at the next meeting. Danny, This wouldn't have to be because you have a vested interest in paludis and paludis does not support this syntax and there happens to be no reasonable way to support that? -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: media-viode/mplayer-1.0_rc2_pre20070321-r4 media-video/transcode-1.0.3_rc2_p20070310-r1 MPlayer needs to be fixed, though it's in the same boat as transcode ... it's a release candidate plus a patch level. Multimedia apps are infamous for rarely having releases, so we are stuck with SVN snapshots. What we really need is a suffix for RCS systems, since that's what they really are. However, if anyone has any suggestions for naming schemes in the meantime, I'm all ears. Steve -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, Apr 24, 2007 at 03:16:38PM -0400, Doug Goldstein wrote: Danny van Dyk wrote: Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: media-viode/mplayer-1.0_rc2_pre20070321-r4 media-video/transcode-1.0.3_rc2_p20070310-r1 An illegal version specification of media-sound/alsa-driver has already been removed from the tree. I would like to ask the affected package maintainers to move these versions to sane version specifications as soon as possible. Thanks in advance for this. Danny So apparently as little as 1 council member can make a decision and it be binding unless appealed to the entire council at the next meeting. Danny, This wouldn't have to be because you have a vested interest in paludis and paludis does not support this syntax and there happens to be no reasonable way to support that? Incidentally paludis *does* support it. -- Fernando J. Pereda Garcimartín 20BB BDC3 761A 4781 E6ED ED0B 0A48 5B0C 60BD 28D4 pgpiD6XCA1LY4.pgp Description: PGP signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, Apr 24, 2007 at 01:15:55PM -0600, Steve Dibb wrote: media-viode/mplayer-1.0_rc2_pre20070321-r4 media-video/transcode-1.0.3_rc2_p20070310-r1 MPlayer needs to be fixed, though it's in the same boat as transcode ... it's a release candidate plus a patch level. Multimedia apps are infamous for rarely having releases, so we are stuck with SVN snapshots. What we really need is a suffix for RCS systems, since that's what they really are. However, if anyone has any suggestions for naming schemes in the meantime, I'm all ears. I was one of the council members to weigh in on this, and the quick improvement that I came up with, was going for a naming scheme like so: printf _rc%d%04d%02d%02d,$RC,$YEAR,$MONTH,$DAY Set the date portions to zero in other builds. Alternatively, follow the example of any ebuild that uses a dated patchset, and just have the date of the patchset in the ebuild, and only increment $PR singly. -- Robin Hugh Johnson Gentoo Linux Developer Council Member E-Mail : [EMAIL PROTECTED] GnuPG FP : 11AC BA4F 4778 E3F6 E4ED F38E B27B 944E 3488 4E85 pgpOTlMUF5lNT.pgp Description: PGP signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Am Dienstag, 24. April 2007 schrieb Doug Goldstein: Danny van Dyk wrote: Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: media-viode/mplayer-1.0_rc2_pre20070321-r4 media-video/transcode-1.0.3_rc2_p20070310-r1 An illegal version specification of media-sound/alsa-driver has already been removed from the tree. I would like to ask the affected package maintainers to move these versions to sane version specifications as soon as possible. Thanks in advance for this. Danny So apparently as little as 1 council member can make a decision and it be binding unless appealed to the entire council at the next meeting. No, that's not correct. 1 council member can't do that. During the council meeting of March 8th 2007 the Council decided that at least 2 members are necessary to act for the whole Council. FYI this decission has been made by 3 Council members, which have been Robin, Bryan and which has been initiated by myself. Further, QA indicated approval prior to this council decission. Danny, This wouldn't have to be because you have a vested interest in paludis and paludis does not support this syntax and there happens to be no reasonable way to support that? Doug, a) Paludis could support arbitrary combinations of multiple version suffixes the same way as Portage currently support this. The Paludis developers chose not to, because b) A very large number of possible suffix combinations aren't sensible. Instead of implicitly allowing every possible combination, one should explicitly allow the sensible subset and explicitly disallow the rest. c) I try very hard to seperate my interest and work on Gentoo and the Council and my interest and work on Paludis. Personally, I would appreciate if you got back to me before you make claims as the ones i just responded to. Both claims are wrong: One evidently so (you can ask kloeri and robbat2), for the other you have to trust either me or ask the other Paludis devs. Danny -- Danny van Dyk [EMAIL PROTECTED] Gentoo/AMD64 Project, Gentoo Scientific Project -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Am Dienstag, 24. April 2007 schrieb Steve Dibb: Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: media-viode/mplayer-1.0_rc2_pre20070321-r4 media-video/transcode-1.0.3_rc2_p20070310-r1 MPlayer needs to be fixed, though it's in the same boat as transcode ... it's a release candidate plus a patch level. Multimedia apps are infamous for rarely having releases, so we are stuck with SVN snapshots. What we really need is a suffix for RCS systems, since that's what they really are. However, if anyone has any suggestions for naming schemes in the meantime, I'm all ears. Only a short response, as I'm a bit in a hurry right now. From #gentoo-council earlier: 18:25 @robbat2 make him covert it to _rc%04d%04d%02d%02d,$RC,$YEAR, $MONTH,$DAY I hope that helps, Danny -- Danny van Dyk [EMAIL PROTECTED] Gentoo/AMD64 Project, Gentoo Scientific Project -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Bryan Østergaard wrote: On Tue, Apr 24, 2007 at 03:49:44PM -0400, Doug Goldstein wrote: Stephen Bennett wrote: On Tue, 24 Apr 2007 15:16:38 -0400 Doug Goldstein [EMAIL PROTECTED] wrote: So apparently as little as 1 council member can make a decision and it be binding unless appealed to the entire council at the next meeting. There were three council members who happened to be around at the time, and those three agreed unanimously. That seems reasonable to me for an interim decision. Is it that serious of an issue that it needed to be done as such and could not wait for a regular council meeting? Granted I understand it's important for you paludis users since paludis doesn't support that. But I'm talking about real Gentoo users that use Portage. I think we are setting a VERY dangerous precedent by allowing a subset of council members to make decisions as a whole if they decide to make a decision outside of a normal session. Who were the 3? Already stated in another reply on this thread but the three council members were robbat2, kugelfang and myself. Regards, Bryan Østergaard Bryan, You and Danny have clearly shown your bias towards paludis take over and support of Gentoo. It's fairly poor taste to FORCE this through during a non-regular meeting for something that paludis is lacking. It's AMAZING how fast you guys are to clamor and fix what you call a QA issue and other problems when we've had issues highlighted for years that the council can't move on. But once it's a possible issue with paludis you guys are quick to respond. Very poor taste -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Danny van Dyk wrote: Am Dienstag, 24. April 2007 schrieb Doug Goldstein: Danny van Dyk wrote: Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: media-viode/mplayer-1.0_rc2_pre20070321-r4 media-video/transcode-1.0.3_rc2_p20070310-r1 An illegal version specification of media-sound/alsa-driver has already been removed from the tree. I would like to ask the affected package maintainers to move these versions to sane version specifications as soon as possible. Thanks in advance for this. Danny So apparently as little as 1 council member can make a decision and it be binding unless appealed to the entire council at the next meeting. No, that's not correct. 1 council member can't do that. During the council meeting of March 8th 2007 the Council decided that at least 2 members are necessary to act for the whole Council. FYI this decission has been made by 3 Council members, which have been Robin, Bryan and which has been initiated by myself. Further, QA indicated approval prior to this council decission. Danny, This wouldn't have to be because you have a vested interest in paludis and paludis does not support this syntax and there happens to be no reasonable way to support that? Doug, a) Paludis could support arbitrary combinations of multiple version suffixes the same way as Portage currently support this. The Paludis developers chose not to, because b) A very large number of possible suffix combinations aren't sensible. Instead of implicitly allowing every possible combination, one should explicitly allow the sensible subset and explicitly disallow the rest. c) I try very hard to seperate my interest and work on Gentoo and the Council and my interest and work on Paludis. Personally, I would appreciate if you got back to me before you make claims as the ones i just responded to. Both claims are wrong: One evidently so (you can ask kloeri and robbat2), for the other you have to trust either me or ask the other Paludis devs. Danny QA being spb, who is a noted paludis developer -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Am Dienstag, 24. April 2007 schrieb Petteri Räty: Danny van Dyk kirjoitti: Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: What is the reason this needed an urgent decision? This was first added to the tree little under three months ago so why not just wait for the next council meeting? *alsa-driver-1.0.14_rc2_p3234 (04 Feb 2007) 04 Feb 2007; Diego Pettenò [EMAIL PROTECTED] +alsa-driver-1.0.14_rc2_p3234.ebuild: Add a new snapshot required for kernel 2.6.20. From my POV: * alsa version commited to the tree, * mplayer version has been commited, * alsa version has been removed, * general discussion started on what combinations are allowed * somewhere in between the transcode version was added My rationale was and is to stop people continueing to add such versions w/o prior discussion. Danny -- Danny van Dyk [EMAIL PROTECTED] Gentoo/AMD64 Project, Gentoo Scientific Project -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, Apr 24, 2007 at 04:00:42PM -0400, Doug Goldstein wrote: Bryan Østergaard wrote: On Tue, Apr 24, 2007 at 03:49:44PM -0400, Doug Goldstein wrote: Bryan, You and Danny have clearly shown your bias towards paludis take over and support of Gentoo. It's fairly poor taste to FORCE this through during a non-regular meeting for something that paludis is lacking. It's AMAZING how fast you guys are to clamor and fix what you call a QA issue and other problems when we've had issues highlighted for years that the council can't move on. But once it's a possible issue with paludis you guys are quick to respond. Please stop the conspiracy theories. This has nothing to do with paludis and everything to do with what we consider sane in the tree - no matter which package manager you use. And as stated otherwise paludis already supports multiple suffixes even if it's not in a released version yet so it's not an issue for paludis either. Regards, Bryan Østergaard -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Danny van Dyk wrote: Am Dienstag, 24. April 2007 schrieb Petteri Räty: Danny van Dyk kirjoitti: Hi all, [CC'ing [EMAIL PROTECTED] as requested by GLEP amendment from March 8th, 2007] A subset of council members decided today that multiple version suffixes are illegal in the tree pending further notice. This decission can be appealed at the next Council meeting. If there is sufficient public demand, an earlier meeting can be held. This decission has been made to prevent sufficient precedence for unilateral changes to the tree structure. So far the following package versions are considered illegal: What is the reason this needed an urgent decision? This was first added to the tree little under three months ago so why not just wait for the next council meeting? *alsa-driver-1.0.14_rc2_p3234 (04 Feb 2007) 04 Feb 2007; Diego Pettenò [EMAIL PROTECTED] +alsa-driver-1.0.14_rc2_p3234.ebuild: Add a new snapshot required for kernel 2.6.20. From my POV: * alsa version commited to the tree, * mplayer version has been commited, * alsa version has been removed, * general discussion started on what combinations are allowed * somewhere in between the transcode version was added My rationale was and is to stop people continueing to add such versions w/o prior discussion. Danny If the decision needed to be made quickly after knowing about it for 3 months, there was clearly the opporunity to use the half-impromptu meetings as discussed last year (I believe October-ish) which requires a few days of advance and presence of at least six devs. That should not have been too difficult to use and allowed a little bit more time, warnings and discussion rather then a rush decision. -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Only a short response, as I'm a bit in a hurry right now. From #gentoo-council earlier: 18:25 @robbat2 make him covert it to _rc%04d%04d%02d%02d,$RC,$YEAR, $MONTH,$DAY Let me see if I have this straight: suppose we have package foo-0.1_rc2 released (very outdated) and we're waiting for foo-0.1_rc3. Then example of something between those two would be foo-0.1_rc000220070313? Would that force portage to update to this version? Wouldn't that prevent portage from enforcing update to _rc3 when it's delivered? Of course I might be wrong and if this is the case then excuse me for the whole fuss ;) Best regards, Jurek -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Am Dienstag, 24. April 2007 schrieb Jurek Bartuszek: Only a short response, as I'm a bit in a hurry right now. From #gentoo-council earlier: 18:25 @robbat2 make him covert it to _rc%04d%04d%02d%02d,$RC,$YEAR, $MONTH,$DAY Let me see if I have this straight: suppose we have package foo-0.1_rc2 released (very outdated) and we're waiting for foo-0.1_rc3. Then example of something between those two would be foo-0.1_rc000220070313? Would that force portage to update to this version? Wouldn't that prevent portage from enforcing update to _rc3 when it's delivered? Of course I might be wrong and if this is the case then excuse me for the whole fuss ;) Existing _rcX cases can be handled like this: _rc2-rMMDD Portage will update from _rc2 to a version with revision part 0. Danny -- Danny van Dyk [EMAIL PROTECTED] Gentoo/AMD64 Project, Gentoo Scientific Project -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Jurek Bartuszek wrote: Only a short response, as I'm a bit in a hurry right now. From #gentoo-council earlier: 18:25 @robbat2 make him covert it to _rc%04d%04d%02d%02d,$RC,$YEAR, $MONTH,$DAY Let me see if I have this straight: suppose we have package foo-0.1_rc2 released (very outdated) and we're waiting for foo-0.1_rc3. Then example of something between those two would be foo-0.1_rc000220070313? Would that force portage to update to this version? Wouldn't that prevent portage from enforcing update to _rc3 when it's delivered? Of course I might be wrong and if this is the case then excuse me for the whole fuss ;) Best regards, Jurek This was one of the very valid use cases proposed and is definitely a situation where robbat2's suggestion will not work. -- [EMAIL PROTECTED] mailing list
[PROCTORS] Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Just a general note to everyone in this thread: I haven't had the time to read the posts in this thread, but proctors have received complaints about behaviour within. For the time being, i would ask all people participating to remember the CoC applies here and act accordingly. We will review the posts in this thread for CoC violations as soon as possible. cheers, Wernfried -- Wernfried Haas (amne) - amne at gentoo dot org Gentoo Forums: http://forums.gentoo.org IRC: #gentoo-forums on freenode - email: forum-mods at gentoo dot org pgpNyJFJRpbsB.pgp Description: PGP signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, Apr 24, 2007 at 09:54:20PM +0200, Danny van Dyk wrote: 18:25 @robbat2 make him covert it to _rc%04d%04d%02d%02d,$RC,$YEAR, $MONTH,$DAY In my original email, I also suggested this solution, but it seems that nobody read it: ] Alternatively, follow the example of any ebuild that uses a dated ] patchset, and just have the date of the patchset in the ebuild, and only ] increment $PR singly. This solution already exists in MANY places in the tree, and should probably be preferred over the long $PR or $RC values. -- Robin Hugh Johnson Gentoo Linux Developer Council Member E-Mail : [EMAIL PROTECTED] GnuPG FP : 11AC BA4F 4778 E3F6 E4ED F38E B27B 944E 3488 4E85 pgpxlRX4xyEDb.pgp Description: PGP signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Only a short response, as I'm a bit in a hurry right now. From #gentoo-council earlier: 18:25 @robbat2 make him covert it to _rc%04d%04d%02d%02d,$RC,$YEAR, $MONTH,$DAY Let me see if I have this straight: suppose we have package foo-0.1_rc2 released (very outdated) and we're waiting for foo-0.1_rc3. Then example of something between those two would be foo-0.1_rc000220070313? Would that force portage to update to this version? Wouldn't that prevent portage from enforcing update to _rc3 when it's delivered? Of course I might be wrong and if this is the case then excuse me for the whole fuss ;) I was planning to post the same question and then I reread danny's and robin's mails : what rc3 will actually be named is something like : _rc0003 not a very funky name but like that portage will see it as rc000220070313, otherwise rc3 is rc000220070313 that also means doing some funky $P renamings in the ebuild to catch upstream _rc3 tarball, but that's probably better than allowing such multiple suffixes. [And that'll make us differ from upstream naming scheme for the whole _rc series] Regards, Alexis. pgpwhSsezguRz.pgp Description: PGP signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Existing _rcX cases can be handled like this: _rc2-rMMDD Portage will update from _rc2 to a version with revision part 0. However, _rc2-rMMDD-r1 would *not* be valid anymore, and I think it's quite easy to imagine when this additional -r1 would be neccessary. Regards, Jurek -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Am Dienstag, 24. April 2007 schrieb Jurek Bartuszek: Existing _rcX cases can be handled like this: _rc2-rMMDD Portage will update from _rc2 to a version with revision part 0. However, _rc2-rMMDD-r1 would *not* be valid anymore, and I think it's quite easy to imagine when this additional -r1 would be neccessary. I'd like to refer you that this is kind of 'best-practice' for the tree. Danny -- Danny van Dyk [EMAIL PROTECTED] Gentoo/AMD64 Project, Gentoo Scientific Project -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tuesday 24 of April 2007 22:47:00 Jurek Bartuszek wrote: Let me see if I have this straight: suppose we have package foo-0.1_rc2 released (very outdated) and we're waiting for foo-0.1_rc3. Then example of something between those two would be foo-0.1_rc000220070313? Would that force portage to update to this version? Wouldn't that prevent portage from enforcing update to _rc3 when it's delivered? Of course I might be wrong and if this is the case then excuse me for the whole fuss ;) foo-0.1_rc2 foo-0.1_rc000220070313 foo-0.1_rc3 -- Best Regards, Piotr Jaroszyński -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tuesday 24 of April 2007 23:20:05 Piotr Jaroszyński wrote: foo-0.1_rc2 foo-0.1_rc000220070313 foo-0.1_rc3 err. foo-0.1_rc2 foo-0.1_rc000220070313 foo-0.1_rc000320070512 What I was trying to say is that once you change to the long versions you must stay with them. -- Best Regards, Piotr Jaroszyński -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, 24 Apr 2007 21:54:21 +0200 Fernando J. Pereda [EMAIL PROTECTED] wrote: You mean real Gentoo users that use a Portage version that don't support multiple suffixes, right ? People still using portage 2.0.x have much more serious problems. Marius -- [EMAIL PROTECTED] mailing list
Re: [PROCTORS] Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Wernfried Haas wrote: Just a general note to everyone in this thread: I haven't had the time to read the posts in this thread, but proctors have received complaints about behaviour within. For the time being, i would ask all people participating to remember the CoC applies here and act accordingly. We will review the posts in this thread for CoC violations as soon as possible. cheers, Wernfried Go get em tiger! -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFGLnYh1c+EtXTHkJcRAspLAJ9HUhy/5oegWBYbfX7YEzeDU63bjQCdHn40 qNtMMhg8cm2jtotWYeUMaK8= =azL1 -END PGP SIGNATURE- -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Am Dienstag, 24. April 2007 schrieb Piotr Jaroszyński: foo-0.1_rc2 foo-0.1_rc000220070313 foo-0.1_rc3 Leading zeros are ignored (unless in very special cases in the version spec and since a recent portage version also in the revision part), so the above is incorrect - generally spoken. Danny -- Danny van Dyk [EMAIL PROTECTED] Gentoo/AMD64 Project, Gentoo Scientific Project -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
that also means doing some funky $P renamings in the ebuild to catch upstream _rc3 tarball, but that's probably better than allowing such multiple suffixes. I disagree, multiple suffixes would be much clearer to read. IMHO renaming _rc3 to _rc0003 is an overkill. Why not simply allow some exceptions from the general rule, such as foo_rcX_preY? It's obvious that mixing most of the suffixes (e.g. _alphaX_betaY) will surely result in bogus versioning system and should be prohibited, although certain combinations would introduce a slight (vast?) simplification. Best regards, Jurek -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, 24 Apr 2007 21:29:37 +0200 Danny van Dyk [EMAIL PROTECTED] wrote: The rationale behind this is the following: * certain combinations of suffixes don't make sense. That's highly subjective. * only recent Portage versions support it. I wouldn't call portage-2.1 recent as it's been stable for almost a year. The only thing that changed recently (like three months ago or so) was that repoman no longer blocks such versions (see bug #165349). Marius PS: Though I disagree with the reasoning I don't have a problem with the decision as long as it's just temporary. -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, 24 Apr 2007 23:20:05 +0200 Piotr Jaroszyński [EMAIL PROTECTED] wrote: On Tuesday 24 of April 2007 22:47:00 Jurek Bartuszek wrote: Let me see if I have this straight: suppose we have package foo-0.1_rc2 released (very outdated) and we're waiting for foo-0.1_rc3. Then example of something between those two would be foo-0.1_rc000220070313? Would that force portage to update to this version? Wouldn't that prevent portage from enforcing update to _rc3 when it's delivered? Of course I might be wrong and if this is the case then excuse me for the whole fuss ;) foo-0.1_rc2 foo-0.1_rc000220070313 foo-0.1_rc3 Nope. The number part of suffixes is compared as integer. Marius -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Robin H. Johnson wrote: In my original email, I also suggested this solution, but it seems that nobody read it: ] Alternatively, follow the example of any ebuild that uses a dated ] patchset, and just have the date of the patchset in the ebuild, and only ] increment $PR singly. This solution already exists in MANY places in the tree, and should probably be preferred over the long $PR or $RC values. Yeah, except revisions are supposed to be for changes to ebuild code, not upstream code. This gets problematic for people trying to report bugs to upstream, because they and upstream have no idea what code they're actually running. Thanks, Donnie -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
err. foo-0.1_rc2 foo-0.1_rc000220070313 foo-0.1_rc000320070512 What I was trying to say is that once you change to the long versions you must stay with them. And there you have another flaw of this system - how am I supposed to predict if I'll ever need the extended _rc versioning in case of that one particular package? I think that massive ebuild renaming is definietly not an option. Best regards, Jurek -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, 2007-04-24 at 13:39 -0700, Ned Ludd wrote: You might be overreacting a little here. To bring you up to speed vapier actually filed the original bug for this after I first noticed one of these atoms creeping into the tree while doing pre release atom compare testing for portage-utils around early February. Till this moment there was no definitive decision of any sort. I think the overreaction here is due to the fact that a seemingly emergency Council meeting was convened to make this decision. And that is a bit confusing (to me, at least). Why the sudden urge to fix this right *now*? I understand that there's a recent addition with ffmpeg and mplayer etc, but this isn't exactly an epidemic in package versioning sweeping through the tree, by any stretch of the imagination. I think a council decision is probably the correct thing (with heavy input from portage and the development community), but an emergency council decision? I'm with Doug on this: it's a little out of place at the moment. Especially when there isn't really an alternative scheme that's been set in stone (the zeroed-out date field idea is one idea -- no offense, Robin, but it does seem a little on the klunky side). I think it'd be nice to first open such alternatives up to discussion before making emergency council decisions and announcements like this. Thanks, Seemant Thanks, Seemant signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Piotr Jaroszyński wrote: On Wednesday 25 of April 2007 00:04:35 Jurek Bartuszek wrote: Hmm, is swiching from _rc2 to _rc0002 trouble-free from user's POV too? Wouldn't he be forced to update from former to latter then? It's the same version. Or am I missing something? There is no need for such a switch, just add new snapshot using the long _rc. Once you do it you will have to keep using it until version bump, e.g. (low to high): 1.0_rc1 1.0_rc000120070101 (newer snapshot of rc1) 1.0_rc0120070102 (leading zeroes are omitted) 1.0_rc000320070201 (here the date doesn't matter as long it's 8 digits long) 1.1_rc1 (yuupi version bump and we can use short _rc again) Where does 1.0_rc2 fall in here? -- Doug Goldstein [EMAIL PROTECTED] http://dev.gentoo.org/~cardoe/ signature.asc Description: OpenPGP digital signature
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Doug Goldstein wrote: Piotr Jaroszyński wrote: There is no need for such a switch, just add new snapshot using the long _rc. Once you do it you will have to keep using it until version bump, e.g. (low to high): 1.0_rc1 1.0_rc000120070101 (newer snapshot of rc1) 1.0_rc0120070102 (leading zeroes are omitted) 1.0_rc000320070201 (here the date doesn't matter as long it's 8 digits long) 1.1_rc1 (yuupi version bump and we can use short _rc again) Where does 1.0_rc2 fall in here? That would be 1.0_rc0002 I suppose. Best regards, Jurek -- [EMAIL PROTECTED] mailing list
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, 2007-04-24 at 18:21 -0400, Seemant Kulleen wrote: Seemant: I'd like to continue to discuss the ways of council decission on gentoo-council rather than on gentoo-dev ML. :-) Happy to do that, in the general case. In this specific case, however, it's valid here, because the entire thread revolves around a rather rash council decision (who knew such a thing was even possible?) that affects pretty much all Gentoo developers (and prospective developers). Seemant is half right here. The discussion should stay here, but only if the discussion is about the technical issues involved in multiple suffixes. Anybody who plans on simply making this a bitching session should reconsider sending any replies, at all. How someone can even pretend that this is such an important issue is beyond me. Nobody held some secret emergency meeting. Most likely, three Council members were simply talking, and decided that the best course of action is to block further changes to the tree involving multiple suffixes until a proper decision can be discussed and decided upon. This makes complete sense to me as it seems like something that should be done to keep the changes from spreading for the short term. Seriously, what is up with all the anti-Council conspiracy theories? Are you guys really *that* against the Council trying to do *anything* productive, at all? Should we all just give up and quit? It seems that every time I open my email client, somebody out there is trying to say that by the Council using the powers afforded to them that somehow they're conspiring to take down Gentoo. Yeah... because that's just what the Council wants to do, make Gentoo a steaming pile of rubble so we can be the supreme rulers of... nothing. Now, if only we can get all these pesky developers out of the way, we could rule the world! -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation (Ashamed to call myself a developer these days) signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Wed, 2007-04-25 at 00:30 +0200, Danny van Dyk wrote: I just wonder why several people feel attacked by this decission while the affected parties have no problem with it. There is a growing anti-authority sentiment within the Gentoo developer community. People want to complain about any decision made, even if it doesn't affect them. This is the power of a free and open community. It allows people, even those that have no vested interest in the decision, to voice their opinions and have them heard. The problem here is that people think that their opinion or personal beliefs trump Council decisions, when they do not. As simplistic as it sounds, when the Council makes a decision you have only a very few choices: - Agree, accept it and move on - Disagree, accept it and move on - Disagree and appeal the decision at the next meeting - Disagree and vote out the current Council next election Yes, there are some more variations on the above, but I leave those as an exercise to the reader. The point here is that these constant discussions on how the Council is killing Gentoo simply because they might have made a decision that you personally might not make or they made it in a manner that you personally might not have used are doing nothing more than undermining the authority given to the Council by the developer community at large when they approved GLEP 39. If you disagree with a Council decision, there are procedures in place for you to be heard and for the decision to be appealed. Use those procedures. Trying to make every decision out to be some sort of abuse of power is getting quite old and tired and is going to lend people to not pay attention anymore, which would be sad if there ever *was* a case of the Council abusing its power. Welcome to the story of the developers that cried abuse... -- Chris Gianelloni Release Engineering Strategic Lead Alpha/AMD64/x86 Architecture Teams Games Developer/Council Member/Foundation Trustee Gentoo Foundation signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Tue, 2007-04-24 at 19:31 -0400, Chris Gianelloni wrote: It seems that every time I open my email client, somebody out there is trying to say that by the Council using the powers afforded to them that somehow they're conspiring to take down Gentoo. Yeah... because that's just what the Council wants to do, make Gentoo a steaming pile of rubble so we can be the supreme rulers of... nothing. Now, if only we can get all these pesky developers out of the way, we could rule the world! You're right, there is. For the record, though, my feeling isn't anything about being anti-authority, etc. Quite the opposite, in fact, because the current leadership is actually doing and accomplishing things. As I stated in my original email (agreeing with your own view that this isn't that big an issue): what was the hurry to get an announcement/decision made without even a valid alternative in place? In other words, there was a policy decision without a clear established way to not violate it (yes, the mplayer/ffmpeg maintainers did whatever it is they did to comply, I know that, but it's not a generalisable solution). So, being that this situation is *not* that big, couldn't it just have waited for all the council members to get together and have opportunity to really propose and establish a viable alternative? If I were to guess I'd say people are a little confused that this required action/decision this quickly and outside of a regular council meeting -- for a real emergency situation, you'd probably see a lot less of a hub-bub about it. But, come on, this is a 3-package issue. signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
On Wed, 2007-04-25 at 00:30 +0200, Danny van Dyk wrote: In my eyes it was a policy issue. Tree-wide policies have to pass the council in one form or the other. So why shouldn't Council care here? My argument is not that Council should not care. My question is: what's the big urgency to rush a half-baked policy through? I just wonder why several people feel attacked by this decission while the affected parties have no problem with it. I hope you don't mean me here, because I haven't felt attacked at all. My concern isn't a personal one. Rather, it's a question that nobody from the council has actually answered: what was the big hurry to make a decision _NOW_ without even thinking through the migration path, or for that matter without even knowing what is the actual correct way. It's fine to say that _rc_alpha_beta_p is wrong (and I happen to agree). It's another to not say what is actually right. Furthermore, if only 3 packages did the wrong thing where was the emergency? Anybody who attends the regular Council meetings and/or reads their logs/summaries knew that this kind of decission is possible. To paraphrase something I've said to people on this list: just because you can does not necessarily mean that you _should_. I probably have more council related commentary, but I'll save that for the appropriate mailing list :) I'm not trying to make you defensive, I just really would like an answer to my question, that is all. Thanks, Seemant signature.asc Description: This is a digitally signed message part
Re: [gentoo-dev] [ANN] Multiple version suffixes illegal in gentoo-x86
Donnie Berkholz wrote: Robin H. Johnson wrote: In my original email, I also suggested this solution, but it seems that nobody read it: ] Alternatively, follow the example of any ebuild that uses a dated ] patchset, and just have the date of the patchset in the ebuild, and only ] increment $PR singly. This solution already exists in MANY places in the tree, and should probably be preferred over the long $PR or $RC values. Yeah, except revisions are supposed to be for changes to ebuild code, not upstream code. This gets problematic for people trying to report bugs to upstream, because they and upstream have no idea what code they're actually running. Thanks, Donnie +1 I agree -r# is for ebuild changes not code changes. I remember a while back Portage would constantly use -r# instead of a 4th number and we worked at that to change that behavior since it was firmly established that -r# was for ebuild changes only. Not bumps in the code. -- Doug Goldstein [EMAIL PROTECTED] http://dev.gentoo.org/~cardoe/ signature.asc Description: OpenPGP digital signature
