18:00:38 #startmeeting FESCO (YYYY-MM-DD) 18:00:38 Meeting started Wed Jan 9 18:00:38 2013 UTC. The chair is jwb. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:38 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:00:38 #meetingname fesco 18:00:38 The meeting name has been set to 'fesco' 18:00:38 #chair abadger1999 jwb mitr mmaslano notting nirik pjones t8m sgallagh 18:00:38 Current chairs: abadger1999 jwb mitr mmaslano nirik notting pjones sgallagh t8m 18:00:39 #topic init process 18:00:48 morning everyone. 18:00:48 uh, heh 18:00:50 hi 18:00:53 Hello 18:00:55 hello 18:00:55 hola 18:01:19 oh well. i'll correct the YYYY-MM-DD thing manually 18:01:20 Hello 18:01:45 happy new year YYYY! 18:02:10 notting is here. he's just not paying attention 18:02:14 I keep writing YYYX on all my checks. ;) 18:02:15 let's get rolling 18:02:32 #topic #896 Refine Feature process 18:02:33 .fesco 896 18:02:33 https://fedorahosted.org/fesco/ticket/896 18:02:35 jwb: #896 (Refine Feature process) – FESCo - https://fedorahosted.org/fesco/ticket/896 18:02:52 ok, so i think the biggest follow up was to look at mitr's templates 18:03:01 i made a few comments on the systemwide feature one 18:03:05 did anyone else look at them? 18:03:16 * nirik failed. Can look more now. 18:04:09 well, i guess there's also the implied thing too. the templates themselves aren't all that big a deal, but they imply we'd adopt a model where we have self-contained and systemwide features 18:04:39 yeah. 18:04:41 hm. some of the sections from the system-wide features could still apply to self-contained ones if the submitter wants them to 18:04:41 i was kind of hoping we'd defer the broader implications until FUDCon, as we have with other things 18:05:21 notting: The idea is that upstream projects have (or should have) similar concerns (compatibility, migration, user notification) and thus it would be mostly taken care of upstream. 18:05:46 For packages where upstreams don't do it or that are especially important, we could bump them into the systemwide category. 18:06:05 mitr, that seems too big bump to me 18:06:07 jwb: yeah, that was my hope too. I know others wanted to have something in place for f19, but I really don't think thats going to be easy to do... 18:07:16 nirik, perhaps we could for F19 just change the process that some non-controversial features will be voted on en bloc? 18:07:27 mitr: true. but i suspect even self-contained ones may have *some* dependencies 18:07:44 t8m: 'en bloc'? 18:07:56 notting: If anything depends on them (and that $anything is not a part of the feature), it doesn't qualify as self-contained. 18:08:15 nirik: "which ones does anybody have objections to? alright, proposal to approve all the others" 18:08:31 which was argued for this time but we just decided to charge all through them anyway 18:08:32 sure, I would be ok with that. 18:08:40 nirik, just vote that we accept feature A B C in one vote - someone would have to propose which are uncontroversial though 18:09:07 i'd be OK with that as a stop-gap for f19. particularly with the devel-announce thing we're doing now 18:09:19 * pjones would as well 18:09:32 fine 18:09:36 ok 18:09:53 sounds good 18:10:06 BTW, I think the devel-announce thing has been a success so far. 18:10:16 nirik: 18:10:17 more discussion from people and interaction from feature owners. 18:10:41 nirik, yep 18:10:55 sgallagh, nirik, abadger1999, mitr, are you OK with voting en bloc for features nobody explicitly questions? 18:11:03 * nirik is 18:11:15 Yes 18:11:29 yeah 18:11:31 * abadger1999 not sure he likes it but is willing to try it this cycle :-) 18:12:27 jwb: count me as 0, undecided. Adds a hurdle to speaking up 18:12:53 #agreed As a stop-gap for Fedora 19, FESCo will vote en bloc on Features that do not have an explicit question/objection (+:8,0:1,-:0) 18:13:22 I'd note that if later anyone has questions about a approved feature we should be able to revisit it. 18:13:38 sure 18:13:59 #info Features can be revisited if later questions arise 18:13:59 which is always true. 18:14:21 ok, propose we defer more discussion on this as we have in previous meetings 18:14:23 * nirik nods. 18:14:27 +1 18:14:32 +1 18:14:42 +1 18:14:49 +1 18:14:58 +1 18:15:05 ok, moving on then 18:15:11 #topic #963 change of names of configuration files 18:15:12 .fesco 963 18:15:12 https://fedorahosted.org/fesco/ticket/963 18:15:14 jwb: #963 (change of names of configuration files) – FESCo - https://fedorahosted.org/fesco/ticket/963 18:15:26 ok, we said we were only really worried about firstboot for release 18:15:36 and i _think_ that is fixed with bug 892097 18:15:44 yeah. 18:15:47 fwiw the grubby change is in updates-testing now, which should be good enough for it. 18:15:55 testing would be appreciated ;) 18:16:12 proposal: close ticket and let rest be fixed by maintainers in the normal cycle? 18:16:17 +1 18:16:18 +1 18:16:21 +1 18:16:29 +1 18:16:30 nirik: +1 since it seems like that has already happened 18:16:30 +1 18:16:33 +1 18:16:56 +1 18:17:09 BTW, there was mention on the list of us setting a precident of some kind asking the feature owner to make the changes to all affected packages. 18:17:18 #agreed firstboot is fixed so closing the ticket out and maintainers can fix remaining in the normal cycle (+:9,0:0,-:0) 18:17:43 nirik, yeah. i think you correctly pointed out it wasn't a great thing in the end, and not something we'd likely repeat 18:17:49 don't know if we want to make clear there is no such guideline... 18:18:05 I think we do need to move in that direction, but that's for a higher-bandwidth discussion I suppose 18:18:21 Otherwise we are limited to saying "no" or accepting any level of breakage. 18:18:42 well, the other alternative is for us to just do it. Either via a FES like setup or just ourselves. 18:19:14 but yeah, sidetracking 18:19:15 without an army of minions, i think 'just ourselves' is not a good solution 18:19:29 i'd be for asking a Feature owner if they are planning on doing such changes during feature submission. and if they say no, then we reject the feature unless someone volunteers 18:19:35 because wth 18:19:39 thats another option too. 18:20:07 Yeah, I think maybe we'd want to codify that into the Feature announcement too 18:20:24 So that if volunteers are needed, we start looking for them a week in advance of the FESCo meeting 18:20:25 something to keep in mind for higher impact features during our discussions 18:20:32 * nirik nods 18:20:48 https://fedoraproject.org/wiki/User:Mitr/SystemwideFeatures already contains this in the "scope" section 18:21:00 at least in some minimal form. 18:21:05 yes, it does 18:21:15 But then, #963 wasn't a feature 18:21:49 mitr, also, jsut because someone lists things other developers have to do there doesn't mean people are actually agreeing to do it 18:22:11 jwb: That's where FESCo as a voted representative can and should decide 18:22:30 anyway, i think this side discussion is more in line with the previous ticket than 963 18:22:33 When getting agreement from 100 or more people isn't practical 18:23:28 shall we move on? 18:23:45 +1 18:24:22 yep 18:24:26 ok, moving on 18:24:31 new features! 18:24:35 #topic #984 F19 Feature: 3D Printing - 18:24:35 https://fedoraproject.org/wiki/Features/3D_Printing 18:24:35 .fesco 984 18:24:35 https://fedorahosted.org/fesco/ticket/984 18:24:36 http://lists.fedoraproject.org/pipermail/devel/2013-January/175770.html 18:24:39 jwb: #984 (F19 Feature: 3D Printing - https://fedoraproject.org/wiki/Features/3D_Printing) – FESCo - https://fedorahosted.org/fesco/ticket/984 18:24:50 so. no comments on devel@. only a single comment in the wiki page 18:25:03 looks pretty cool, I'm +1 18:25:09 i am as well 18:25:18 +1 18:25:21 (i'm assuming we're doing our en bloc for the next meeting) 18:25:25 +1. Doesn't look like it's likely to break anything else 18:25:27 +1 18:25:40 jwb: I think the other two on the list might be contentious anyway 18:25:44 +1 18:25:54 +1 18:25:56 +1 18:26:33 #agreed F19 Feature: 3D Printing is accepted as a Feature (+:8,0:0,-:0) 18:26:41 #topic #985 F19 Feature: Pillow - https://fedoraproject.org/wiki/Features/Pillow 18:26:44 .fesco 985 18:26:45 jwb: #985 (F19 Feature: Pillow - https://fedoraproject.org/wiki/Features/Pillow) – FESCo - https://fedorahosted.org/fesco/ticket/985 18:26:46 https://fedorahosted.org/fesco/ticket/985 18:26:49 http://lists.fedoraproject.org/pipermail/devel/2013-January/175769.html 18:26:57 +1 18:27:02 +1 18:27:15 (err +1 to both 3D printing and Pillow) 18:27:16 +1 here as well. 18:27:16 +1 18:27:18 +1 18:27:20 +1 18:27:22 How certain are we about its drop-in potential? 18:27:36 There are a LOT of python apps reliant on PIL out there 18:27:44 +1 18:27:55 sgallagh: doesn't make much difference since the module name changes 18:28:21 pjones: So it's going to require patching all of the consumers? 18:28:28 Oh, actually, I misread and it does make a difference. 18:28:31 but a minor one. 18:28:32 I saw the words "drop-in replacement" somewhere 18:28:34 sgallagh: It started as a fork, and the original PIL maintainer seems to support it, that's as certain as we can be 18:28:45 some things appear to need to be fixed to use the other import format 18:28:52 mitr: PIL upstream maintainer or Fedora maintainer? 18:28:58 mitr: upstream 18:28:59 sgallagh: Fedora maintainer 18:29:01 er 18:29:03 gah. 18:29:11 I said everything wrong in that statement :) 18:29:39 Hmm... from our previous discussoin, perhaps we should be asking who is going to do the porting (patch the import) work. 18:29:51 might be a good idea 18:29:53 * abadger1999 will volunteer to help out with that, though. 18:29:56 * nirik nods. 18:30:08 ok. so defer until we have an answer to that? 18:30:47 sure. If we want to make sure it's addressed. 18:31:10 Well, if abadger1999 is going to own organizing that, I think that's probably sufficient to vote now. 18:31:15 18:31:41 The way the featuree is written now ("Action needed by other maintainers"), individual package owners 18:31:41 ok. so we're at +7 with me and sgallagh not voting yet 18:31:50 I'd be ok as long as abadger1999 is willing to make sure it happens. ;) 18:31:51 * abadger1999 is still +1 since I'll work on the porting. No worries if others feel differntly :-) 18:31:52 anyone want to change their vote? 18:31:59 jwb: I'm +1 now 18:32:18 (Definitely a beneficial switch, I just want it to go smoothly) 18:32:18 #agreed F19 Feature: Pillow is accepted as a Feature (+:9,0:0,-:0) 18:32:38 #info abadger1999 will help coordinate the porting effort 18:32:55 #topic #986 F19 Feature: DualstackNetworking - 18:32:56 https://fedoraproject.org/wiki/Features/DualstackNetworking 18:32:56 .fesco 986 18:32:56 https://fedorahosted.org/fesco/ticket/986 18:32:56 http://lists.fedoraproject.org/pipermail/devel/2013-January/175773.html 18:32:57 jwb: #986 (F19 Feature: DualstackNetworking - https://fedoraproject.org/wiki/Features/DualstackNetworking) – FESCo - https://fedorahosted.org/fesco/ticket/986 18:33:01 abadger1999: I'll take some of the packages to port as well if you want. 18:33:06 sgallagh: cool. 18:33:13 so networking in both openstack and cloudstack? 18:33:15 (at minimum I'll handle the ones I own like ReviewBoard) 18:33:19 jwb: should I follow up on porting in the fesco ticket or the feature page? 18:33:21 heh. 18:33:25 notting, ha 18:33:36 abadger1999, good question. you're leading it, so your choice :) 18:33:40 will do. 18:34:03 seems like we shipped RHEL 5 with this feature. 18:34:05 more seriously, i agree with the comments that the very nebulous and potentially expansive scope here is a problem 18:34:08 This feature sounds to me more like "polish up the dualstack support"... since we really already do support dualstack. 18:34:20 and "polish up" is not very clear 18:34:24 the scope is not well defined here 18:34:49 should we send it back and have them narrow it down to specific items they intend to fix? 18:34:49 * nirik also wonders if glibc maintainers are on board here. 18:35:02 jwb: I'd be for that. 18:35:09 me as well 18:35:12 Since as described this isn't anything new AFAICS 18:35:13 * nirik has been running dualstack here for quite a while. 18:35:25 there are a number of questions on the talk page that aren't answered as well 18:35:35 This sounds like it would be better handled as a tracking BZ than a Feature page. 18:35:42 yeah, lets ask them to answer those and followup on devel list discussion? 18:35:43 sgallagh, yes 18:35:45 sgallagh, +1 18:35:49 +1 18:36:08 nirik, +1 as well 18:36:12 sgallagh: unless there really is something large and specific they're talking about, yeah. 18:36:25 +1 18:36:25 so reject and suggest a tracking bug? 18:36:28 sgallagh: yes, +1. there's nothing *wrong* with the ideas behind this feature, but it's not really a defined feature as is 18:36:30 or defer? 18:36:35 I think they have a specific plan re: AI_ADDRCONFIG, but it's not been written down in the feature 18:36:41 Seems like the systemd cleanup task. 18:36:45 eh... defer 18:36:52 so rewrite the feature proposal 18:36:57 It sounds to me more like bug fixing than a feature. 18:37:06 It could be good release notes fodder if there's a lot of core services that get fixed. 18:37:09 defer, but we should try and communicate to feature owners... 18:37:10 it should be clear when the feature will be finished, which is not 18:37:15 I think we should reject it as-is and ask them to either file bugs or reopen with more details 18:37:18 the cost of ipv6 is eternal vigilance. 18:37:31 In principle, would we accept "we'll do stuff to improve IPv6" as a feature at all? 18:37:37 ;-) 18:37:49 mitr: if there's some large, specific gain, maybe? but otherwise probably not at this point 18:38:06 mitr: as long as the stuff was listed and needed coordination to happen? 18:38:10 since it works well enough that we've been saying it works for most of a decade. 18:38:19 ok. how does this sound: agreed F19 DualstackNetworking is deferred as the feature is too vague and it is not clear that this is more than a general improvement in existing IPv6/dual stack support 18:38:19 i am concerned, for things like AI_ADDRCONFIG, that there probably should be coordination because changing/fixing them can cause weird issues with packages later if they're broken 18:38:40 jwb: sounds fine 18:38:44 jwb: +1 18:38:53 jwb: Phrasing sounds good to me. +1 18:38:53 jwb: +1 18:38:55 agreed F19 DualstackNetworking is deferred as the feature is too vague and it is not clear that this is more than a general improvement in existing IPv6/dual stack support 18:38:59 sigh 18:39:08 +1 18:39:09 jwb, +1 18:39:19 +1 18:39:26 I might add a request that they set up a BZ tracker and start filing bugs on individual packages though 18:39:45 Because it would be good to track their efforts (I doubt they'll stop implementing just because it's not a Feature) 18:39:52 pjones: There have been _fairly recent_ problems re: AI_ADDRCONFIG. "Sorting out the AI_ADDRCONFIG mess" might make sense as a feature 18:39:55 #agreed F19 DualstackNetworking is deferred as the feature is too vague and it is not clear that this is more than a general improvement in existing IPv6/dual stack support (+:8,0:0,-:0) 18:39:58 they have one. actually two. ;) 18:40:01 sgallagh: yes, good point 18:40:11 (they made one, and there's an old one dwmw2_gone made) 18:40:12 mitr: well, let them rephrase the feature that way then, if that's what they mean 18:40:20 pjones: sure 18:40:28 #info It would be good to highlight the existing tracking bugs more prominently 18:41:10 #topic Next Week's Chair 18:41:18 I'll do it 18:41:24 great. thanks mmaslano 18:41:29 #agreed mmaslano to chair next week 18:41:32 #topic Open Floor 18:41:41 anyone have any comments/questions? 18:41:48 i think we're finally go for f18 release 18:41:54 * jwb crosses fingers and stuff 18:41:56 FUDCON: anything we should be planning for besides the Feature Revamp? 18:42:18 i think that's the big one 18:42:21 (And the short "Meet Your FESCo" session) 18:42:36 * nirik nods. 18:42:43 i guess i missed that one. we're doing that? 18:43:09 jwb: I proposed it. Basically just a 30 minute baby-shaking, hand-kissing thing. 18:43:21 ok, cool 18:43:23 good luck :) 18:43:23 Do we want to just defer next week's meeting until fudcon? or is that going to make it less productive? 18:43:33 or perhaps people yelling at us telling us we are doing it wrong. ;) 18:43:39 * abadger1999 makes a note not to bring babies to meet sgallagh 18:43:41 we could ack features on meeting 18:43:44 pjones, was wondering that myself. though we will have some f19 features by next week 18:43:51 Not everyone will be at FUDCon, and the feature load is increasing 18:43:52 perhaps just limit it to that 18:43:53 pjones: Not all of FESCo will be at FUDCON. 18:44:03 so? I wasn't suggesting to not have it on irc 18:44:19 though I guess that'd make it a solid mixed bag 18:44:34 I think we can keep the meeting time and do just features 18:44:42 yeah 18:44:49 fine with me 18:45:14 So, regular time but features-only? 18:45:16 sure. I think it would be good to knock them out if we can 18:45:51 open floor? 18:45:57 #agreed Next week's FESCo meeting will be new F19 Features only 18:45:59 Viking-Ice, yes 18:46:00 Viking-Ice: Yes, do you have something to discuss? 18:46:33 fesco input on https://bugzilla.redhat.com/show_bug.cgi?id=810040 18:47:03 * nirik already provided his input. Other interested fesco folks feel free to. 18:47:18 ...really. 18:47:19 we're already spinning an RC4 for this, right? 18:47:23 we only have the go/no-go in 13 minutes. 18:47:39 jwb: yes but it would require retesting/slip I think. 18:47:50 why? 18:47:50 jwb: yes. rc4 is being built which includes this. rc3 does not. they are identical otherwise. we can argue about what to do / which one to ship at go/no-go. in 13 minutes. 18:48:08 wait, rc3 just showed up like a few hours ago, right? 18:48:15 jwb: because it has a package that affects the login process. at least desktop/gdm using things would need retested. 18:48:24 jwb: the only thing that changed were the release notes, though 18:48:30 between RC2 and RC3 18:48:34 oh 18:48:36 it's there since f17 and it's blocker now? 18:48:46 anyhow, we can discuss in other meeting, imho 18:49:05 mmaslano, we missed it last release cycle 18:49:06 Viking-Ice: What FESCo input are you expecting? 18:49:20 Viking-Ice: I think you should try to be less caustic on bugs. 18:49:43 pjones: i was the most caustic one, tbf. 18:49:46 * jreznik already prepared a batch of features for next meeting, will file tickets tmrw 18:49:56 adamw: but you didn't ask for my input. 18:50:08 jreznik: did you see the one feature owner that sent their own announce/ticket? 18:50:30 nirik: which one? 18:50:44 https://fedorahosted.org/fesco/ticket/988 18:50:56 and there's a devel-announce post from them in moderation 18:51:26 * jreznik is checking 18:51:36 anyhow, I have nothing more for open floor. ;) 18:51:36 * sgallagh needs to talk to that guy and figure out where that's going to fit with OpenLMI/CIM 18:51:50 nirik, we should just close the ticket and ask him to follow the process 18:51:57 pjones, so you assume I'm like that on every bug 18:52:18 nirik: well, someone took really fast track skipping wrangler at all :) 18:52:32 jreznik: yeah. 18:52:42 jreznik: if you could talk with them that would be great. 18:52:48 Viking-Ice, pjones: i'm not interested in turning the FESCo meeting into a discourse on bug ettiqutte. take it elsewhere 18:52:48 yep, definitely 18:52:54 thanks for heads up 18:53:51 t8m: I'll check it tmrw and will talk to him - if it's going to look good, no need to close the ticket... 18:54:11 Viking-Ice, i think the general consensus is that FESCo members can offer input in the bug or at the go/no-go meeting in a few minutes. 18:54:20 jreznik, probably - hopefully nobody else will follow with doing it the same way 18:54:22 i doubt you'll get much else during this meeting 18:54:48 jwb, yup 18:54:53 or via the bug. 18:55:07 i said that ;) 18:55:10 well, would be great to see fesco guys around at go/no-go to help to decide which rc we want to go with 18:55:18 * nirik can't read. 18:55:20 so thanks :) 18:55:29 jreznik, which channel is that one in again? 18:55:49 * sgallagh will be there 18:55:56 fedora-meeting-2 18:55:58 thx 18:56:02 to avoid conflicts 18:56:06 ok, anything else for FESCo Open Floor? 18:56:13 and I expected it's going to be looong one :) 18:56:58 if there's nothing else in 2 min, i'll close this out 18:58:33 #endmeeting