18:04:23 #startmeeting Fedora Release Engineering Meeting 18:04:23 Meeting started Mon Oct 26 18:04:23 2009 UTC. The chair is Oxf13. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:04:23 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:04:37 #meetingname fedora-releng 18:04:37 The meeting name has been set to 'fedora-releng' 18:04:43 #topic roll call 18:05:11 ping: notting jwb spot rdieter lmacken wwoods dgilmore warren poelcat 18:05:23 hm? 18:05:24 oh 18:05:28 here 18:05:29 here i guess 18:05:34 * notting is here 18:05:35 * spot is around for 15 minutes, then has to bail 18:06:15 good enough to get going. 18:06:19 #topic Fedora 12 18:06:38 There really is just one topic, and that's the looming Fedora 12 release. 18:06:46 We had a blocker bug review day or two last week, and trimmed the blocker bug up a bit. Still rather large 18:06:57 We are due to enter RC stage on the 4th of November 18:08:02 * wwoods hereish 18:08:19 so there is going to be a number of tag requests coming in. I'll be making sure they get trough and we have as much testing time as possible. 18:09:30 We also need to figure out when to push updates. 18:09:33 #topic Fedora 12 Updates 18:09:51 the concern with pushing updates too early is that it splits testers, and it also risks the buildroot 18:10:08 as we'd be putting things into the buildroot that aren't necessarily going out in the release, and could lead to broken deps/packages. 18:10:42 so the later we wait to push updates, the less risk there is 18:11:38 yeah 18:11:57 also, no idea if releng2 is ready for mashing or not. asked notting and skvidal about it, but haven't heard back yet 18:12:03 is bodhi open for submitting f12 updates atm ? (i thought no) 18:12:13 rdieter, no. it was, and now it isn't 18:12:20 rdieter: no, we're pushing people toward tagging instead 18:12:20 ok 18:12:23 * dgilmore is here 18:12:26 since we're just rubber stamping non-critpath packages 18:12:48 well, deciding when to push could hinge on when submissions are re-enabled (or it could be vice-versa too) 18:13:18 sure, we have to decide when we want to push, and then open for submissions a bit before that 18:13:44 * skvidal is here 18:13:44 Oxf13: i kinda think we should open up for updates a week before ga 18:13:44 one way around buildroot issues is that we could create a new target/buildroot that is just f12-final content 18:13:44 pushing was going to happen after RC had been finalized 18:13:52 and people building stuff for RC fixes would build specifically using that target 18:14:06 and I'm not sure we have the right bits on releng2 - I'll have to sift what's in the rpm and deltarpm on there 18:14:07 but that seems overkill to me 18:14:17 Oxf13: that would be fine as long as well yell about it 18:14:22 skvidal, that would be appreciated 18:14:24 it would be easy for people to get wrong 18:14:26 I'd lean toward opening for update as soon as we enter "RC" phase 18:14:42 once we get into RC, I only want to tag things that fix problems found in the RC 18:14:51 Oxf13: +1, I was thinking that too 18:16:00 and division of testers isn't as much of a concern at that point 18:16:09 as long as we have a plan, i'm less concerned about the details thereof. +1 18:17:04 #idea Enable submission of and pushing of Fedora 12 updates when Fedora 12 enters RC phase 18:17:08 +1 18:17:26 (I'll take notting's previous +1 into account) 18:18:08 +1 18:18:26 jwb: spot rdieter wwoods lmacken ? 18:18:47 enabling is fine 18:19:10 i probably won't do a push until at least RC2 or 3 though 18:19:10 sure, why not. 18:19:10 ;) 18:19:10 +1 18:19:25 actually some chatter in #fedora-devel reminds me of something 18:19:36 We should have at least a day or 3 in order to fix any "ZOMG signed upates kills packagekit" type bugs 18:19:37 +1 (for posterity, was away for a bit) 18:19:46 so having actual update repos before the RC is gold will help there. 18:20:05 Oxf13, mumble. ok 18:20:14 Oxf13, btw, still need access to the F12 key 18:20:19 craps 18:20:21 jwb: okay - deltarpm sure doesn't look new enough at all on releng2 - so that's going to need some love 18:20:27 Oxf13: yeah I was about to mention that 18:20:31 heh 18:20:34 #action f13 to provide jwb (and notting?) with F12 key access 18:20:43 the ever-terrifying first round of signed packages 18:20:43 skvidal, figured. also think we need a fixed xz 18:20:44 nod 18:21:12 but otherwise yeah, +1 to enabling submission/push of f12 updates at/just before RC 18:21:54 I can start on rebuilding the f12 versions of deltarpm and xz for rhel5 - out of curiosity - if we build deltarpms up to latest for f12 - do we hose them for F11? 18:22:31 skvidal: how so? 18:22:50 #agreed We'll enable updates submission and pushing at RC phase 18:23:01 notting: just making sure the deltarpm on f11 could handle the format we might be handing them from the f12 build 18:23:30 there were no deltarpm format changes in f12, afaik. just fixing it to handle xz-compressed payloads if they were encountered 18:23:45 notting: this is why I was checking 18:23:52 hopefully F12 is the last time we have to worry about this at all 18:24:05 jwb: agreed - hopefully we'll have chroots for this stuff 18:24:18 lmacken has it on his TG2 version of bodhi list 18:25:00 ok 18:25:00 hawt. 18:25:31 jwb: wait, we're prototyping a new option to make updates pushes add a step and take more time? :) 18:26:05 notting: you don't get to complain about that, delta boy 18:27:13 notting, yes. at the hope that we can possibly do smart things 18:29:10 ok, anything else on updates? 18:29:33 no? great. 18:29:44 #topic FUDCon Toronto 18:29:51 Just a couple things about fudcon toronto 18:30:10 I'll be there, and I've signed up to work on message bus, autoqa, no frozen rawhide, and SCM replacement. 18:30:48 while doing some incredibly mindless tasks I came up with some thoughts/plans for no frozen rawhide and next gen SCM that I'm going to put somewhere on teh wiki before fudcon 18:31:06 but if any of you are going to be there, and want to join me, please do, or if you have other topics you think we should hack on or talk about in a session, please let me know 18:31:09 i'll be there. i have not signed up to work on anything yet 18:31:20 Oxf13: ill help with scm and message bus 18:31:33 I'llbe there, wearing a facemask and gloves 18:31:38 I'll be there, not sure how much help I could be 18:31:55 im also going to work on epel and secondary arch releng, while tehre 18:32:07 note to self, bring baseball, glove. 18:32:08 cool, sounds like a good set of folks 18:32:16 ? 18:32:32 Oxf13: me or rdieter ? 18:32:38 rdieter: 18:33:09 lame response to skvidal 18:33:25 rdieter: bring purrell 18:33:25 ah 18:33:35 pretty sure i won't be there 18:33:49 note to self, assign all work items to jwb 18:34:03 hee 18:34:15 note to self, realize you don't get paid and laugh manically at Oxf13 18:34:36 yeah, I get paid to take this abuse, what's your excuse? 18:34:51 * Oxf13 watches the meeting go off the rails 18:34:54 i'm a masochist 18:35:01 #topic Hardware 18:35:21 We recently got a 'compose-x862' system from mmcgrath. It's eventually going to replace compose-x86, as it has /lots/ more disk space 18:35:39 so that I don't have to purge Fedora spin isos in order to have enough disk space to make the live images 18:36:01 But since it's new, and since it hasn't been used much, I'm only going to use it for Live image creation for F12 RCs 18:36:13 the Fedora install spin will still be made within chroots on compose-x86 18:36:19 just to limit chances of things going pear shaped 18:37:23 that's all I got. 18:37:28 #topic open floor 18:37:46 anything from anybody else? If not, I'll close the meeting in a few minutes 18:39:27 f13 updates 18:39:46 or really, how they play into NFR 18:39:59 but that can wait i guess 18:41:10 jwb: briefly, my plan for that is when we branch F13 away from rawhide, we'll immediately create an "updates-testing" like repo for F13 18:41:43 anybody who wants to do a build for F13 will have to use bodhi and put it into an updates-testing repo. Karma and/or the maintainer can promote it from testing to stable via bodhi 18:42:00 when it goes "stable" it may just get tagged for f13, and the nightly compose will pick it up 18:42:01 and/or autoqa? 18:42:15 notting: yeah, sorry, autoqa + releng/qa for crit path items play in there too 18:42:22 how is that updates-testing repo going to get mashed? 18:42:36 jwb: we could either do it nightly, or as part of the rest of the updates pushes 18:42:49 i think it'd have to be nightly 18:42:50 if we had autosigning going, nightly would make sense, or even multiple times during a day 18:43:01 right, which was my next question 18:43:25 we should probably brainstorm a bit on autosigning :/ 18:43:36 maybe not in today's meeting, but in general 18:43:43 yeah 18:44:31 and just like this release, when we enter RC stage, promotion to "stable" doesn't mean tag for f13, it means push to updates/ 18:45:54 jwb: does that answer some of your question? 18:46:24 think so. just want it kept in mind during the initial part of F13 and while TG2 bodhi is being worked on 18:46:34 * jwb points lmacken to ^ 18:47:48 kk 18:47:59 alright, lets wrap this thing up, thanks all for coming! 18:48:02 #endmeeting