20:59:40 #startmeeting 20:59:40 Meeting started Mon Jul 12 20:59:40 2010 UTC. The chair is brunowolff. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:59:40 Useful Commands: #action #agreed #halp #info #idea #link #topic. 20:59:53 #meetingname spins-sig 20:59:53 The meeting name has been set to 'spins-sig' 21:00:06 #chair brunowolff,nirik 21:00:06 Current chairs: brunowolff nirik 21:00:07 * nirik is here. 21:00:37 #topic Who's here? 21:00:49 Anybody else here for Spins SIG? 21:01:25 * nirik was hoping for a bit more turnout. oh well. 21:01:41 OxF13, did you want to talk about spins at GA release now or wait for the next releng meeting? 21:02:16 On the plus side, by tomorrow we should have heard from all of the recurring spin owners. 21:02:40 #topic GA release 21:02:41 did you hear anything from huff? 21:02:59 I have a liveCD irc meeting tomorrow that he is running. 21:03:08 ah, cool. 21:03:10 * maxamillion is here 21:03:16 I'll ask him before or afterwards what is intentions are. 21:03:37 sounds good. 21:03:54 The releng meeting last Friday didn't happen, because Jesse had some outside issues that prevented him from making the meeting. 21:04:19 There may be a make up meeting or we may just need to wait until Friday. 21:04:51 ok 21:04:59 With the poor response of our Spin Owners, I estimate that that we will likely drop or delay publishing a spin if we make 21:05:35 publishing an ISO require affirmation from the spin owners that the RC ISOs work. 21:05:48 yeah, could be. 21:05:57 I wish we could get better/faster feedback... 21:06:21 We need to make sure we know what to do if that happens. This could involve the Design Team as they have wiki pages for the spins. 21:07:02 I think we will likely only get a few days for them to review and report back. But jesse will have a better idea about that. 21:07:32 yeah, we could drop the requirement that it's a spin owner... 21:07:34 Once I talk to them I am plan on emailing the rules to the mailing list and the listed spin owners. 21:07:38 ie, _someone_ tests it. 21:07:57 Well I assume delegates are allowed, but it really shouldn't be us. 21:09:02 I think at one time QA was looking at doing it, but they don't have the manpower and we haven't developed test plans. 21:09:49 I expect the spins wrangler will act as liason, but we'll see what releng (especially Jesse) wants. 21:10:02 yeah 21:10:34 Assuming noone else volunteers I do that. I should be in a quiet time for stuff other than spins then. 21:11:20 Spins SIG has failed the last two (at least) releases and Jesse has done what he could to get things out. 21:11:29 once the spins take place, it's not difficult to have a few testers for htem 21:11:33 I am hoping we look better this time around. 21:12:19 It depends on how much spin specific testing needs to be done. 21:12:51 boot, check for avcs, fire up the apps that it is centering on and validate that they actually do run, etc.. 21:13:01 * nirik nods. 21:13:05 what else do you need tested? 21:13:06 If QA wants to help with some basic sanity testing, then maybe that's good enough. Though I would prefer an owner to check at least 21:13:18 there's often not much time for that after final compose and before release, but it should be possible. 21:13:24 one version of their spin to make sure special stuff works reasonably. 21:13:54 We don't have test plans for the spin specific stuff. 21:14:10 Does QA have generic test plans for live CDs? 21:14:38 Could we have them execute those on the live images to make sure they boot and the like? 21:14:54 do all the spins have an 'install' option, or only some? 21:15:10 It looks like there will be 8 or 9 ISO spins this time around. Do they have time to test all of those? 21:15:12 fenris02: I think they all have an install option 21:15:20 or should 21:15:31 The ones that we make ISOs for should. 21:16:33 they take a bit longer to test. how many testers exist for these 8-9 spins? 21:16:42 I don't know. 21:17:12 Releng might have some ideas. But maybe after talking to releng, I'll need to talk to QA. 21:17:22 It's a pretty busy time for QA though. 21:17:33 indeed. 21:17:59 Spins SIG has no testing resources other than the spin owners. 21:18:10 And I am not too sure about them. 21:18:23 we do have test plans for all kinds of things 21:18:31 with detailed test cases 21:18:47 you are welcomed and encouraged to execute them using your spins 21:19:15 Does QA have the resources to do some sanity testing on 8 or 9 live images for GA? 21:20:23 I am not convinced that the owners are going to actually do tests of RCs at GA. 21:21:04 If QA did some sanity testing I'd worry less about letting those go through even if the Spin Owners fail to do their job. 21:21:25 If it's on the Spin Owners, then we probably need to be more hard nosed. 21:21:51 brunowolff: no 21:22:02 But that also means having a real plan to cancel (or delay) spins at the last minute, since it is likely to happen. 21:22:10 Thanks. 21:22:18 QA does not have the resources to test the spins. It's up to spin owners to find testers. 21:22:35 we provide test plans, test cases, test tools, documentation, guidelines, assistance, etc. 21:22:39 manpower, sadly, is up to you 21:22:44 I appreciate that. 21:23:15 I am kind of down on Spin Owners lately, because they seem to think they just have to provide ks files and everything else 21:23:19 happens by magic. 21:23:59 yeah. ;( 21:24:28 Well, I think we have gotten some useful stuff out in the open, but we should probably move on. I'll have more after I get releng's expectations. 21:24:30 probably the simplest way to find the test *cases* right now is to look at old test *results* and follow the links back from there 21:24:54 http://fedoraproject.org/wiki/Category:Desktop_Acceptance_Test_Cases 21:25:01 is probably the most interesting set of test cases 21:25:32 #action brunowolff will dig up some test case pointers well before GA so that spin owners can be directed to them. 21:25:51 should be a place to record the results of those for each spin as well. 21:26:05 there's also the l10n tests: http://fedoraproject.org/wiki/QA:Fedora_13_l10n_Results_Template 21:26:19 for test results you'll want to use the corresponding results template 21:26:26 and fill in the blanks as documented in the wiki page 21:26:43 #link http://fedoraproject.org/wiki/QA:Fedora_13_l10n_Results_Template 21:26:54 #link http://fedoraproject.org/wiki/Category:Desktop_Acceptance_Test_Cases 21:27:17 #info probably the simplest way to find the test *cases* right now is to look at old test *results* and follow the links back from there 21:27:19 see also http://fedoraproject.org/wiki/QA:Create_Install_Test_Result_Page 21:27:31 #link http://fedoraproject.org/wiki/QA:Create_Install_Test_Result_Page 21:27:48 I know there's a results template page for the desktop cases somewhere but I can't find it 21:28:27 aha: http://fedoraproject.org/wiki/QA:Desktop_validation_results_template 21:28:39 #link http://fedoraproject.org/wiki/QA:Desktop_validation_results_template 21:28:42 should have looked at http://fedoraproject.org/wiki/Category:QA_Templates 21:29:13 anyway, hope that helps - feel free to ask in #fedora-qa or on fedora-test-list if you need help setting up results pages, finding templates, etc 21:29:15 Thanks a lot, that will save me some searching and gave me some ideas that I might not have thought of. 21:29:23 also fedora-test-list might not be a bad place to ask for test assistance 21:29:33 #info anyway, hope that helps - feel free to ask in #fedora-qa or on fedora-test-list if you need help setting up results pages, finding templates, etc 21:29:47 #info also fedora-test-list might not be a bad place to ask for test assistance 21:29:58 #topic meego 21:30:28 nirik, you said you changed things for moblin to meego, but i didn't see it changed in the nightlies yet? 21:30:33 * nirik saw stickster's note on the list about that. 21:30:38 Is it too early or is there a problem? 21:30:44 brunowolff: it was changed sunday, todays composes all failed. 21:31:09 OK, then we don't need to worry about that. 21:31:19 (Well about meego.) 21:31:31 #topic spin submissions 21:31:47 Tomorrow is the deadline for spin submissions 21:31:59 yep. 21:32:07 looking like no new ones this cycle. 21:32:14 I looked at spins in development, and only two seemed like they might request new spins. 21:32:35 One just started in the last week and I really doubt they will shoot for F14. 21:33:03 the music guys? 21:33:08 The other is Dan Walsh's which needed some changes to livebase that maybe wouldn't be a good idea. 21:33:40 We haven't had those changes seriously reviewed and unless someone does, I don't think his can be included. 21:34:10 wasn't someone going to look into it for us? 21:34:17 * nirik can't recall who tho 21:34:25 He does have a version on live journal that doesn't need our livebase, so maybe delayign this isn't a huge deal. 21:35:00 The music guys mentioned they had some nonfree stuff, so I don't think we are going to have to worry about those guys. 21:35:06 yeah. ;( 21:35:32 I don't like not doing stuff for Dan, because he is very good at jumping on stuff for other people. 21:35:52 yep. So the version on his journal doesn't require changes in livebase? 21:36:19 I don't know who is even a good person to do the review. I am not at that level now. I don't have time to get to that level this release. 21:37:04 I don't believe it does or he has an alternate livebase there. He said something about needing to tweak things if livebase changed. 21:37:27 So it's not ideal, but will let people experiment. 21:37:47 I think he probably wanted a ks only, no ISO. 21:37:53 nirik, no, but only because he used %post to remove all the things anaconda did 21:38:07 ah, ok, so a revert hack. 21:38:44 Anything other than those two would be a surprise. 21:39:07 perhaps we can ask him to clarify that and repost his patches? 21:39:30 I have set up the release page for spins with all 11 spins from F13 with 9 marked as getting ISOs and all needing Spins SIG approval. 21:40:08 It would have been better to have had some up for approval already so we didn't need to do a bunch at once, but since they are 21:40:43 all recurring it probably won't be too bad. Probably I'll report all minor changes to the ks files at the next meeting or two and we 21:40:50 can rubberstamp them. 21:41:20 MeeGo might be a bit more on our end, but the rest probably aren't doing anything major. 21:41:44 yeah, there are not many if any xfce changes this cycle. 21:41:45 I didn't get through all of the spin wiki pages thsi past week. 21:42:04 I did some prep work, and got a couple done. 21:42:39 * nirik put the xfce one in the right category I hope 21:42:49 My plan is to finish up this week and set them all to Spins Ready For Wrangler where the owner has said they plan to continue. 21:43:00 Then I'll start looking at ks changes. 21:43:22 sounds great! 21:43:26 I hope to have at least a few we can approve at the next meeting. 21:43:30 * nirik really appreciates brunowolff working on this. 21:44:08 I'll update the release page as we approve them. We shouldn't fail to produce any spins because that page is messed up this time around. 21:44:18 I appreciate your support. 21:45:04 You are even busier than I am, but if it was just me I would have given up on the SIG and we probably would be doing any additional spins. 21:45:51 Hopefully if we get some stuff better documented it will be easier for more owners to help out. 21:45:55 yeah, long term we need a better setup or something. 21:46:36 #action brunowolff will continue reviewing spin wiki pages this week 21:46:57 #action brunowolff will have at least a few spins up for approval next week. 21:47:34 #action brunowolff will eventuall write up spin leader and spin owner documents similar to the spin wrangler document. 21:47:58 The spin owner one might try to make people feel guilty. 21:48:23 #topic spin-kickstart commit messages 21:48:33 yeah, a spin owner responsibilities might be good to have. 21:48:37 I got a commit bounce yesterday. 21:48:59 oops. ;( 21:49:13 I think we either need to suggest that committers also subscribe under their project address or we should whitelist 21:49:27 I'm happy to whitelist, but I need to know what to whitelist. 21:49:28 all *@fedoraproject.org addresses for the spins list? 21:49:40 What do you think? 21:49:50 well, we could try that I guess. might result in spam from people who forge those addresses tho. 21:50:36 shall we try and see how it goes? 21:50:39 I wouldn't expect a significant amount of spam sent to the spins list from fedoraproject.org addresses. If they went to that much 21:50:56 trouble they'd subscribe to the lists and spam away. 21:51:01 perhaps. 21:51:24 Why don't we try that and see how it goes. I think that is less ongoing work. 21:51:57 Otherwise we can document subscribing under the other address and not having messages sent to it. 21:52:06 ok, can do. 21:52:21 But that will probably generate ongoing questions as new people get added. 21:52:39 humm... 21:52:46 #action nirik will whitelist *@fedoraproject.org addresses for the spins list. 21:53:13 New topic hmm? 21:53:13 brunowolff: looks like kanarip is the owner of the list, let me check with him and make me or us owners too? there are some requests pending I see. 21:53:48 ie, if we moderate instead of bounce we can add people to whitelist. 21:53:49 as needed. 21:53:54 Yeah, though I'll try to stay away from it so that I don't turn off reply-to munging and start another flame war. 21:54:40 ok, I can get it sorted. 21:54:46 I think that will be even more ongoing work. And it will be on people who are already pretty busy, 21:55:12 Is that list holding email for moderation now? 21:55:59 I'd rather handle one off requests for whitelisting than do continual moderation. 21:56:08 yes, it is. 21:56:22 I already deal with a number of lists, one more isn't that big a deal to me. 21:56:24 Is anyone actually reviewing the email? 21:56:30 I just moderate them in the morning each day. 21:56:41 no idea... kanarip is listed as the owner... 21:57:35 OK, if you are already doing that and it's not too bad, than whitelisting as needed is viable, but I'd rather you weren't the one doing that. 21:57:36 anyhow, I will sort it out. 21:57:43 OK, thanks. 21:58:14 #topic membership list 21:58:45 We're almost out of time and have covered the important stuff, but I wanted to check for objections for having removed the 21:58:55 yeah, I think it was pretty old. 21:59:10 membership list from the wiki page, since it didn't seem to reflect the membership. 21:59:45 is there any advantage to having the list public rather than the ml ? 21:59:54 I didn't add anything back, since we aren't getting good participation and I don't know that we have much use for the list 21:59:58 anyway. 22:00:25 Pretty much the wrangler and the lead (if we get a new one) are the only people that should get contacted directly. 22:00:31 yeah. 22:00:38 everything else should be 'mail the list' ;) 22:00:53 * fenris02 nods 22:01:07 I am not sure there is much correlation between the mailing list and active members. 22:01:35 As long as noone was looking for my head, I'll just leave it. 22:01:48 #topic other topics 22:02:01 Anything else anyone wants to discuss/ 22:02:08 ? 22:02:10 * nirik has nothing. 22:02:25 Ending in 30 ... 22:02:40 Thanks for coming guys, next week we should get to do some voting. 22:02:56 cool. 22:03:03 thanks for running things brunowolff ! 22:03:06 I'll take care of sending out the notes. 22:03:14 #endmeeting