18:00:39 #startmeeting 18:00:39 Let the Jenkins meeting commence! 18:00:50 #chair kohsuke rtyler danielbe_ hare_brain 18:00:50 Current chairs: danielbe_ hare_brain kohsuke rtyler 18:01:08 #topic Recap last meeting actions 18:01:18 http://meetings.jenkins-ci.org/jenkins-meeting/2016/jenkins-meeting.2016-08-17-18.00.html 18:01:30 o/ 18:01:43 seems all AIs are done 18:01:46 oleg-nenashev: there's an item here about JENKINS-37315 for the LTS, did that get integrated? 18:01:48 \o 18:01:54 ah okie doke 18:02:27 oleg-nenashev: do you mind if I throw a brief GSoC recap as the next topic? 18:02:39 np 18:02:46 ogondza is getting late 18:02:53 #topic GSoC recap 18:03:13 the summer is over! 18:03:15 :( 18:03:38 #info GSoC 2016 is over 18:03:40 I believe we will be seeing alexsomai in a couple weeks at Jenkins WOrld 18:03:52 and then we have a mentor summit in the fall some time 18:03:54 yes I'll be there :) 18:04:00 * rtyler can't remember dates past Jenkins World right now 18:04:30 oleg-nenashev: do you think there are any notable updates to share here? 18:04:30 #info WWe had two students working on second part of the project. 1 student has passed the final evaluation (alexsomai) 18:04:45 #info On Thursday we had presentations by students 18:05:01 this was our first year accepted into GSoC, so there are lots of learnings that I hope we can apply next year 18:05:04 #info Slides and recordings are referenced here: https://www.meetup.com/Jenkins-online-meetup/events/233378586/ 18:05:06 * rtyler crosses fingers 18:05:31 sorry, got disconnected silently 18:05:35 #action oleg_nenashev and batmat to follow-up on the old retrospective and feedback collection AI 18:06:02 yup 18:06:19 #info alexsomay will conduct a lightning talk and demo at Jenkins World. You can find him at the Community Booth there ;) 18:06:44 any questions/thoughts before we move on? 18:06:53 #action oleg_nenashev to write a final blogpost and to prepare his lightning talk for JW 18:07:09 No other updats from me 18:07:19 Does the "retrospective" Oleg mentioned mean discussing the learnings that Tyler mentioned? 18:07:29 #action oleg_nenashev to work on sending swags to students 18:07:37 orrc: Yes 18:07:41 orrc: somehow, yes. 18:07:42 Cool 18:07:57 We already started collecting things, but yeah 18:08:06 rtyler also has AI to go to the Mentor summit and to bring a bunch of notes from there 18:08:13 Good job, GSoC mentors :) 18:08:16 such notes 18:08:26 (and mentees, if that's a word) 18:08:34 #topic LTS status check 18:08:38 ogondza: you're up buddy 18:08:50 we are ready for the release 18:09:00 of 2.7.3 yeah? 18:09:12 special thanks to Oleg for reporting regression during RC period 18:09:16 kohsuke ogondza: So the plan is to spin it today? 18:09:16 yes, .3 18:09:36 We have been actively testing RC on these weeks, no other serious regressions 18:09:49 good to hear that 18:10:30 JENKINS-37759 has been addressed in all OSS plugins 18:10:44 So we do not need to fix it in the core 18:10:58 I mean urgent fix 18:11:03 #info https://issues.jenkins-ci.org/browse/JENKINS-37759 18:11:49 oleg-nenashev: lets get in into the next one then 18:11:57 next topic? 18:12:11 rtyler: I need the release date from kohsuke 18:12:33 action kohsuke to release 2.7.3 18:12:38 is he around? 18:12:45 #action kohsuke to release 2.7.3 18:12:50 Seems no 18:13:09 ok, lets agree it will happen for today 18:13:14 heh 18:13:28 hrmpw: you're closest, make sure he pushes buttons 18:13:28 can we move on? 18:13:30 fingers crossed 18:13:47 in a related area, ogondza and I discussed some tasks earlier for decreasins the SPoF on the release process 18:13:57 that's covered under https://issues.jenkins-ci.org/browse/INFRA-910 18:13:59 FYI 18:14:04 nice 2 hear 18:14:07 #topic Pick new LTS version 18:14:15 we are working on the same for remoting 18:14:33 #info https://jenkins.io/changelog/ 18:14:57 the default choice would be 2.18 based on age 18:15:18 I like 2.19 due to better user feedback, and the diff is also reasonable 18:15:23 2.16+ can cause Jenkins not starting - do we want to postpone that (and choose something older)? 18:15:23 2.18 ... 2.20 are more like bugfix releases 18:15:38 ogondza ? 18:15:46 rtyler: he's presenting right now but he does always have 2 laptops with him 18:15:52 should be able to multitask 18:15:52 heh 18:15:59 hrmpw: as long as it happens today! 18:16:04 2.19 looks pretty good 18:16:29 the fix for JENKINS-37332 shouhld be backported IMHO if we choose anything under 19 18:16:32 +1 18:16:32 ogondza Well, at some point we need to base LTS on 2.16+ for the plugin loading 18:16:42 ogondza And it's actually improved robustness 18:16:51 ogondza Assuming you refer to JENKINS-21486 18:17:16 yes. I said we are talking about postponing what will happen anyway 18:17:55 No strong opinion, but maybe sooner is better than later 18:17:58 oleg-nenashev recently mentioned something about a kind of "upgrade guide" listing what to look out for when upgrading to newer LTS, if we did that, we should be able to enable users to do this successfully 18:18:22 i.e. not just a bold line in the changelog, but a section in a short manual 18:18:36 #action oleg_nenashev to finally create a PR for the upgrade guide 18:18:47 neat 18:18:51 Needs some polishing 18:19:05 I am fine with 2.19 18:19:24 So IMO 2.16+ isn't much of a concern, and we need to get better at telling users about possibly breaking changes anyway 18:19:27 We will need to backport the most of 2.20, but I'm fine as well 18:19:49 danielbe_: agreed 18:19:59 2.20 has some potentially important/breaking changes like JENKINS-37559 18:20:06 So I would not vote for it 18:20:10 oleg-nenashev: there are only a couple fixes in 2.20 that look important, unless i'm missing the severity of some of these 18:20:34 oleg-nenashev Solaris is the only one, the rest looks harmless, no? 18:20:56 Yes, all other PRs are pretty small 18:21:17 Maybe the new permission is the second biggest risk 18:21:28 So IMO 2.20 would also work, but I think we would prefer 2.19 + backports of most of it? 18:21:37 ugh new permission :< 18:21:42 it's a feature :-) 18:22:04 right 18:22:19 I'm not happy how some plugins handle new permissions 18:22:30 But "sooner or later" again 18:22:30 oleg-nenashev: in what way? 18:22:57 let's table this for now 18:23:24 ok, so do you guys prefer 2.20? 18:23:54 content-wise, slight preference, age-wise, no. I'd rather see a 2.19 + most of the stuff backported 18:23:55 2.19 or 2.20 both seem reasonable to me 18:24:04 I can support that danielbe_ 18:24:23 I do not feel strongly and there always ware voices to pick the latest and go ... 18:24:34 danielbe_: the question is what is the effective difference 18:24:49 taking it as it is, we know people use that before 18:25:01 Yeah, prolly 2.20 would not be a good signal for an LTS I guess 18:25:15 three days old release :P 18:25:17 2.20 is a bit too recent for me 18:25:22 New permission + Solaris + JSON Signature API, which is restricted anyway 18:25:26 Yeah, right, soaked enough 18:25:33 2.19 is better from my PoV 18:26:02 so it seems like 2.19 is the consensus 18:26:03 also, easier to go 2.19 + not backport something than 2.20 + revert. 18:26:14 Sure 18:26:16 2.19 + not backport everything 18:26:20 ? 18:26:34 yes 18:26:35 Well, yes, or it's a 2.20 :) ? 18:26:40 some of it, not all of it 18:26:47 of course 18:27:03 (I don't trust the Solaris change without more feedback TBH, we have too few users on that) 18:27:23 * batmat thinks that guy ^ is right 18:27:57 alright 18:28:02 #agreed 2.19 will be the next baseline for the LTS 18:28:11 I also do not trust Windows junctions :) 18:28:27 yay 18:28:33 #topic next meeting 18:28:44 interesting for once 18:28:49 We have Jenkins World 18:28:50 the calendar would have us on September 14th, which is the first day of presentations at Jenkins World 18:29:02 I propose we simply skip a cycle, and resume four weeks from now 18:29:11 Impacts RC 18:29:19 But we can discuss it offline 18:29:23 which would be Sept 28 18:29:25 it is rc check - we can do that on ML 18:29:31 We can probably do a short meeting limited to RC if needed, or ML 18:29:55 for that, I trust our release officer to coordinate to the extent it is required :) 18:30:21 would Sept 28th work for everybody? 18:30:25 +1 18:30:32 ALL HAIL FOR THE RELEASE OFFICER 18:30:33 +1 18:30:39 sure, i am not the one who is traveling 18:31:22 #agreed next meeting will be September 28th 18:31:30 alright folks, that's a wrap for today 18:31:30 yey 18:31:32 #endmeeting