18:02:08 #startmeeting 18:02:08 Let the Jenkins meeting commence! 18:02:35 #chair hare_brain rtyler 18:02:35 Current chairs: hare_brain kohsuke rtyler 18:02:38 #info https://wiki.jenkins-ci.org/display/JENKINS/Governance+Meeting+Agenda 18:02:53 #topic 1.554 for LTS base: go or no-go 18:03:02 ogondza: ping 18:03:07 over here 18:03:14 Hi! This is Scott From Ericsson. Very happy to be taking part in my first Jenkins Governance meeting! 18:03:24 ScottHebertErics: Yay, thanks for joining 18:03:31 Me too ;) 18:03:42 Christian Lague as well 18:03:47 #info I would like to propose to have JENKINS-22310 backported to 1.554.x seeing as it has a major impact on CLI performance. 18:03:49 JENKINS-22310:CLI slow performance investigation (Resolved) http://jenkins-ci.org/issue/22310 18:04:00 ogondza: So the context is that we want to start executing on the new LTS plan 18:04:13 and we thought that would start ticking when we pick the next LTS base version 18:04:45 In the previous meeting it looked like 1.554 was good, but it felt bit too new and you weren't there 18:04:58 So we kicked the can down the road a bit to this one 18:05:11 yeah, 1.554 looks goo to me too 18:05:33 #info look at http://jenkins-ci.org/changelog and click "Community ratings" 18:05:50 Yes, 1.554 looks even better now than before 18:06:01 All right, so let's go with that one 18:06:06 where is the time line document... 18:06:18 #info https://wiki.jenkins-ci.org/display/JENKINS/LTS+Release+Line 18:06:24 one more thing before that. 1.532.3 18:06:45 backports are ready since March 1 18:07:13 OK, so my failing as usual then 18:07:32 I'll push out RC ASAP 18:07:40 Is 532.3 accepting backports? 18:07:41 #action kohsuke to post 1.532.3 RC ASAP 18:08:13 ScottHebertErics: It was from Feb 15th 18:08:36 ScottHebertErics, no. I do not think there will be any 1.532.4 18:09:10 I'm going to write 1.554.x LTS schedules in the calendar 18:09:45 ok...what about 554.x? Can we request backports? 18:09:51 #info http://jenkins-ci.org/content/event-calendar 18:10:00 Yippie, build fixed! 18:10:01 Project jenkins_main_trunk build #3276: FIXED in 59 min: http://ci.jenkins-ci.org/job/jenkins_main_trunk/3276/ 18:10:01 Jesse Glick: When a QueueItemAuthenticator is in use, report the user under whose authentication the build is running. 18:10:23 ScottHebertErics: yeah, see "Backporting process" under https://wiki.jenkins-ci.org/display/JENKINS/LTS+Release+Line 18:11:03 So Apr 16th would be 1.554.1 LTS RC 18:11:27 Apr 30th would be 1.554.1 ship date 18:12:06 May 14 is 1.554.2 RC 18:12:13 May 28 is 1.554.2 LTS 18:13:17 Jun 11 is 1.554.3 RC 18:13:28 JENKINS-22310 is already marked as lts-candidate. What can we do to get it into 554.x? We have resources to help backport it. 18:13:30 JENKINS-22310:CLI slow performance investigation (Resolved) http://jenkins-ci.org/issue/22310 18:13:33 Jun 25 is 1.554.3 release and picking next baseline 18:14:31 perfect. so the next backporting window is as you suggested last time April 2-16th 18:14:49 +1 on Scott's question ;) 18:15:03 ScottHebertErics: I need to check when it went into the mainline release to see if it satisfies the soaking criteria 18:15:31 kohsuke: Went in before 557 18:15:37 Generally speaking, the fix needs to be out for a while before it can be backported 18:15:53 But otherwise I think it's great that you can help us backport 18:16:03 But then again, you fixed it...so it's gold :) 18:16:59 I think the fix of JENKINS-22310 went in as JENKINS-21579, which is targeted for 1.558 and that's next Monday 18:17:01 JENKINS-22310:CLI slow performance investigation (Resolved) http://jenkins-ci.org/issue/22310 18:17:03 JENKINS-21579:Very slow resource loading from UberClassLoader (Resolved) http://jenkins-ci.org/issue/21579 18:18:14 Hmm, I'm not seeing the soaking requirement explicitly mentioned in the LTS wiki page any more 18:18:28 Did we drop that? 18:18:45 jglick: ping? 18:19:03 Oh there it is "Changes to this branch will be restricted to backported cherry-picked changes from the trunk that are "battle-tested" — meaning those commits that have already been a part of a main line release for more than a week." 18:19:23 which means it's up for backporting from April 14th 18:19:36 So I guess we can plan on it unless 1.558 flops 18:19:52 It is a change in the classloader, so bad things may happen 18:19:58 (yup here) 18:20:14 anyway 18:20:26 Yes JENKINS-21579 has high regression risk. 18:20:28 JENKINS-21579:Very slow resource loading from UberClassLoader (Resolved) http://jenkins-ci.org/issue/21579 18:20:34 But is very important too. 18:20:40 #action rename current 'stable' branch to 'stable-1.532' and move stable to 1.554 18:20:42 kohsuke: Can you add an action for Ericsson to port that JENKINS-22310/JENKINS-21579? 18:20:44 JENKINS-22310:CLI slow performance investigation (Resolved) http://jenkins-ci.org/issue/22310 18:21:09 #action ScottHebertErics to backport JENKINS-22310 and JENKINS-21579 18:21:20 ogondza: I assume you are OK with it 18:21:34 Is there any way to make an "Invoke Batch Task" step block subsequent steps? We have a step that runs a deploy, and then we want to run QA on the deployed code. But the deploy gets invoked, and then before finishing the build goes to the next step, so QA tests always fail. 18:21:49 FarrisG: let's hold the question while the project meeting is in progress 18:22:04 not at all 18:22:14 oh that's not good 18:22:36 oh, I do not _mind_ at all 18:22:42 phew 18:22:43 my apologies 18:22:54 #agreed next LTS would be 1.554.x and schedule is on the calendar now 18:23:14 shall we move on to the next topic? 18:23:33 #topic Patron of Jenkins program approval 18:23:38 I hope hare_brain is here for this one 18:24:05 We have discussed this a few meetings back 18:24:22 #info and the record is here http://meetings.jenkins-ci.org/jenkins/2014/jenkins.2014-01-22-19.01.html 18:24:44 The basic idea is to add another donation thank you program 18:25:10 ... which is also like an advertisement on our Wiki 18:25:22 I am here 18:25:35 Yay 18:25:35 The feedback in the original conversation was that people generally seemed OK with it 18:25:50 and so lisawells and I went back and worked on details 18:26:06 That's captured in https://wiki.jenkins-ci.org/display/JENKINS/Patron+of+Jenkins+program 18:26:25 And for the last 2 weeks or so I have test deployed it on our Wiki 18:26:43 So some of you might have seen a text banner for security advisory and JUC 18:27:15 I hadn’t. I got distracted by Mr. Clean. ;) 18:27:36 It only shows every once in 7 page hits, so it was rare 18:27:52 hare_brain: glad to hear you look at blogs! 18:28:18 Ah, I got a blank banner space just now. 18:28:28 So nothing programmed against it at the moment? 18:28:31 And the backend does seem to work now, and I haven't heard any complaint so that was good sign 18:28:46 hare_brain: I guess now it's Q2 so we lost the messages 18:28:51 My test broadcast was only for Q1 18:28:57 BTW data comes from here: https://github.com/jenkinsci/patron/blob/gh-pages/messages.xml 18:29:23 And you can see ones in ... 18:29:57 http://jenkinsci.github.io/patron/tester.html 18:30:12 or https://jenkins-ci.org/patron/2014Q1-0_0.html 18:30:23 hmm not sure why image animates on the 2nd one 18:30:44 It gave me a chuckle. 18:31:09 Between the muted tone of the message and relative infrequency of this, hopefully this satisfies the concern that it's not too distracting 18:31:39 The other concern was the diversity 18:31:55 The only tweak I would suggest is fine print in the box to let people know this is a patron message. 18:32:06 ah 18:32:21 That's a good point 18:33:00 I'll add something 18:33:37 #action kohsuke to add a fineprint in the top right corner to say this is a message from a patron 18:33:41 It doesn’t have to be too serious. “And now, a message from our patrons” gives it a bit of character. 18:33:47 Or something along those lines 18:33:53 Yeah 18:34:17 So back to diversity 18:34:43 lisawells talked to a number of contacts she had and arranged 3 willing patrons filling all 4 slots for a launch 18:34:59 you can see their messages at https://github.com/jenkinsci/patron/blob/gh-pages/messages.xml#L22 18:35:42 it's currently put as Q3 but if this meeting approves the messages, they'd be happy to donate the amount and we'd put these messages live 18:35:50 I don’t recall the details around the original concern. 18:36:50 let me see if I can find that in the full log quickly 18:37:32 IIRC abayer wanted to make sure it wasn't dominated by one patron 18:38:28 They are in http://meetings.jenkins-ci.org/jenkins/2013/jenkins.2013-12-11-19.01.log.html 18:38:37 Yeah, scanning that too. 18:38:47 around 19:21:09 18:39:29 rtyler: you here? 18:40:15 Oh, an even earlier discussion on this. 18:40:35 hello, I am getting strange exceptions while checking out a git project http://pastebin.com/NS4dEGPL I've tried ~15 times, and I am hitting the same issue at various times during the checkout. I have a fast internet pipe (> 20Mbits both ways, even though it doesn't help). Not using abort / build time out anymore. Has anyone seen that before ? 18:40:39 I obviously have some conflict of interest here, but I think the current proposal adequately addresses the concern 18:40:54 lacostej: can you hold on your question until after the meeting? 18:41:03 oopps sorry. 18:41:05 Even with CloudBees' 2 slots it still only shows up once in 14 page hits 18:41:05 So from the messages.xml, it looks like Lisa has CB taking up two slots. 18:41:06 hi all. anyone here using https://wiki.jenkins-ci.org/display/JENKINS/GitHub+pull+request+builder+plugin ?? im having the issue that its not detecting any new PR's or changes 18:41:23 hare_brain: yes 18:41:33 electrical: can you hold on your question until after the meeting? 18:41:45 What happens if not all slots are filled? Just rotate through what we have, or does CloudBees become a default slot filler? I would be concerned about the latter. 18:42:01 kohsuke: meeting? 18:42:02 No, just empty box will show up, as you saw today 18:42:11 electrical: governance meeting is going on 18:42:25 Slide-O-Mix: ahh okay. yeah of course :-) 18:42:40 So it seems like we should have a default Jenkins community filler for empty slots. 18:42:43 Or Chuck Norris. 18:42:57 Yeah, I could that 18:43:30 #action kohsuke to enhance the backend so that idle slots will be filled by the community messages 18:43:35 With those comments, I’m good with it. 18:44:05 Any other thoughts anyone else? 18:44:14 * kohsuke is trying to cautious to avoid any backlash later 18:44:18 I was wondering about whether to allow one patron to buy up all the slots, or if there’s a maximum number of slots for a patron, but I haven’t formed an opinion one way or another about it yet. 18:44:51 lisawells: I think we will refrain from sponsoring more than 2 given the past concern 18:44:55 But definitely not something to hold this up for. 18:45:10 Good of her to say that. I was thinking about other patrons. :) 18:45:17 I think it would make sense to say you can only sponsor 2 slots per quarter - then no one sponsor 'dominates' 18:45:55 Let’s give it a whirl. 18:46:13 We are asking prospective patrons to get in contact with us before they pull the trigger 18:46:40 hare_brain: I assume you mean you don't think we have to codify it just yet 18:47:20 #action kohsuke to get abayer's +1 offline 18:47:44 Oh, and do the current list of messages look good? I'd like to have them approved as well 18:47:57 I'm hoping to launch this today 18:48:06 since it's the beginning of a new quarter 18:48:19 #info the proposed messages are in https://github.com/jenkinsci/patron/blob/gh-pages/messages.xml#L22 18:48:25 Yes, that’s what I meant. 18:48:43 any concerns/+1 from folks? 18:49:38 +1 from me for messages 18:49:45 The messages are fine. 18:49:50 #agreed the program and current proposed patron messages are approved with the recorded TODOs 18:49:55 ScottHebertErics: thanks for your +1 18:49:59 I assume we’re not going to be generating so much traffic that we tank those servers. 18:50:07 :-) 18:50:22 OK, moving on to the next topic of Slide-O-Mix 18:50:27 #topic JIRA versions backend app 18:50:31 howdy 18:50:46 Slide-O-Mix: sorry the other topics ate too much time. The floor is all yours now 18:50:56 I had proposed on the mailing list a while back to add specific versions to JIRA 18:51:17 There were several ideas thrown around and we ended up talking about parsing the Maven repo information and using that as a backend job 18:51:22 I created https://github.com/jenkinsci/backend-jira-versions for that purpose 18:51:41 cool 18:51:51 I'd like to get some feedback on the implementation and once its +1'd get it into the backend setup 18:52:25 In addition, I created a pull request (https://github.com/jenkinsci/backend-ircbot/pull/10) which adds the ability to create versions via the Jenkins IRC bot 18:52:37 So IIUC, the version field will show the combined versions of all the plugins 18:52:48 which would be ... like a few 1000s? 18:53:01 You mean if you looked at "versions" in JIRA? 18:53:08 #info the thread that Slide-O-Mix is referring to is https://groups.google.com/d/msg/jenkinsci-dev/ncvbwPj_R78/FnE38TGToJ0J 18:53:08 yes 18:53:26 Yes, there would be many there, JIRA doesn't seem to have a way to assign them to specific components, they would be global 18:53:39 is it possible to get all the additional details about the build from run and job here? https://github.com/jenkinsci/notification-plugin/blob/master/src/main/java/com/tikal/hudson/plugins/notification/Phase.java#L51 18:53:47 The issue I am trying to fix is when someone chooses "current" for the version when submitting an issue, current is not helpful 18:53:47 I literally just read an email about field contexts in JIRA. 18:54:03 Have we considered the possibility of just using numbers like "1.0" without the plugin name? 18:54:22 kohsuke: that's a possibility, though if you wanted to report on bugs per version, it wouldn't give you the info you wanted 18:54:34 plus, you have lots of different versioning schemes from what I've seen 18:55:04 Yes, but still presumably a lot less than adding plugin name as a prefix 18:55:08 True 18:55:17 But I guess what you want is to be able to query by a specific version of a specific plugin 18:55:49 Well, if an issue is for multiple plugins with different versions, what do you do in that case? or if plugins have the same version, which versions match to which plugins for the issue? 18:55:50 or that perhaps can be achieved by combining version with component 18:56:45 Oh I see, you can have many versions in one issue 18:56:50 OK, that makes sense 18:56:52 right 18:57:02 anyone out there in lala land use the Swarm plugin? 18:57:06 And it's an auto-completion field so presumably having lots of lots versions won't be too bad 18:57:17 Maybe we can filter the maven repo based on date and only start importing versions after a certain date 18:57:41 Everyone in that thread seems to be happy with this, so shall we just go ahead and deploy this? 18:57:55 I suppose the worst case we can delete all those versions 18:58:38 Do you still feel the need to add this to ircbot as well? 18:58:40 Deleting versions might break saved queries. Not sure if anyone would care. 18:58:43 I am trying to run the swarm client JAR and getting this error: http://pastie.org/8989430 can anyone help me out? 18:59:02 hare_brain: I was thinking about deleting versions as a way of completely rolling back the deployment of this new version scheme 18:59:08 Oh OK 18:59:18 kohsuke: the only reason I would like to have the IRC bot stuff is so that plugin developers can add releases before they are actually released, that way they can mark something as fixed in a specific version 18:59:22 I wasn't thinking about that to discard old versions after a while, which would cause the problem you mention 18:59:47 I see, that makes sense 18:59:49 The "Fixed in version" field would actually mean soemthing 19:00:04 +1 from me 19:00:20 +1 from me, too 19:00:24 cool 19:00:36 So I'll need to deploy this somewhere, right? 19:00:42 ci.jenkins-ci.org job to run this periodically 19:00:47 yes, that's correct 19:00:55 #action kohsuke to deploy https://github.com/jenkinsci/backend-jira-versions 19:01:02 It does support reading the JIRA username/password from the .jenkins.org file like the IRC bot 19:01:13 Perfect 19:01:20 I wasn't sure of the requirements for the infra stuff, so if that needs to change, let me know 19:01:20 Slide-O-Mix: thank you very much for doing this 19:01:26 I think it's really useful 19:01:31 kohsuke: great, glad to help 19:01:39 I'll make sure that you have write access to the relevant stuff 19:01:53 so that you can tweak them as needed 19:02:03 thank you 19:02:07 #topic next meeting 19:02:26 #info the next meeting would be April 16th same time 19:02:46 #endmeeting