19:00:18 <rtyler> #startmeeting
19:00:18 <robobutler> Let the Jenkins meeting commence!
19:00:24 <rtyler> #chair hare_brain kohsuke danielbeck
19:00:24 <robobutler> Current chairs: danielbeck hare_brain kohsuke rtyler
19:00:40 <rtyler> #topic Recap last meeting's actions
19:00:52 <rtyler> oleg-nenashev: I see an item to incorporate some feedback about press contacts
19:00:56 <rtyler> I can't remember if we did that or not
19:01:01 <rtyler> Feb 3rd seems so long ago :P
19:01:26 <oleg-nenashev> you should not drink so much beer at conferences :P
19:02:25 <kohsuke> Hmm, 1.642.2 RC
19:02:43 <rtyler> #info last meeting actions http://meetings.jenkins-ci.org/jenkins-meeting/2016/jenkins-meeting.2016-02-03-19.00.html
19:02:46 <rtyler> wake up people
19:02:59 <rtyler> there's an ogondza and kohsuke item I see here
19:03:05 <kohsuke> oh good, that one is done
19:03:28 <rtyler> oleg-nenashev: and your items?
19:04:08 <rtyler> okay, I'll answer for him
19:04:13 <kohsuke> Sounds like you guys were having GSoC conversation in #jenkins-community
19:04:22 <rtyler> the GSoC application draft was just sent to the mailing lists
19:04:24 <oleg-nenashev> Jenkins CIA is tombstoned
19:04:33 <rtyler> as a member of the board I put some initial feedback into it already
19:04:42 <rtyler> we have a deadline of this friday so if you have feedback, get it in fast
19:05:06 <kohsuke> and that draft application is where?
19:05:11 <oleg-nenashev> Yeah, GSoC application is in the ready-for-review state. We have two days to review and submit it
19:05:11 <rtyler> read your email plz
19:05:19 <rtyler> we got lots to get to today
19:05:22 <rtyler> #topic LTS status check
19:05:25 <rtyler> ogondza: the floor is yours
19:05:38 <oleg-nenashev> kohsuke: FYI https://groups.google.com/forum/#!topic/jenkinsci-dev/vTxsCLk1XS0
19:05:48 <kohsuke> yep, thanks
19:05:51 <ogondza> as danielbeck suggested we are postponing release by one week
19:06:05 <ogondza> effectively extending testing window
19:06:13 <rtyler> #info we're post-poning release by one week
19:06:21 <kohsuke> my fault
19:06:33 <rtyler> anything else to cover on this topic ogondza ?
19:06:39 <danielbeck> and we found an actual problem in the RC and fixed it
19:06:43 <oleg-nenashev> We've discovered 1 critical issue yesterday, so maybe we need a new RC
19:06:49 <ogondza> kohsuke: 2016-02-24 is good for you?
19:06:54 <kohsuke> OK
19:07:02 <rtyler> are we going to cut a new RC?
19:07:12 <kohsuke> Yes, that date'll be good
19:07:25 <rtyler> #info new LTS date is planned for 2016-02-24
19:07:28 <danielbeck> oleg-nenashev Are you also referring to the GroovyHookScript one?
19:07:39 <oleg-nenashev> +1 for the new RC. Groovy Hooks are in the bad shape in RC-1
19:07:44 <oleg-nenashev> danielbeck: yes
19:07:50 <danielbeck> #info https://issues.jenkins-ci.org/browse/JENKINS-32985
19:08:20 <danielbeck> TBH this doesn't look like it's severe enough to redo the RC
19:08:26 <ogondza> Will we move the cycle by one week?
19:08:33 <ogondza> danielbeck: could not hurt
19:08:45 <kohsuke> ogondza: we talked about your producing LTS RCs going forward, should we do that starting here? Or I'm happy to produce RC too
19:08:55 <oleg-nenashev> danielbeck: Depends on the environment. Many people rely on hooks
19:09:22 <oleg-nenashev> But I have no strong opinion. Now we test the custom core build
19:09:52 <ogondza> kohsuke: yes, lets discus detail after the meeting (are there any)
19:09:52 <kohsuke> #action kohsuke to post 1.642.2 RC2
19:10:00 <kohsuke> ok, then I scratch that
19:10:07 <danielbeck> ogondza I don't think we should move the schedule
19:10:19 <danielbeck> it's quite useful to have it match the meeting schedule
19:10:20 <ogondza> #action ogondza to push the rc
19:10:43 <rtyler> do we have enough covered for this topic in the meeting to where follow-ups can happen immediately after?
19:11:24 <kohsuke> +1 on danielbeck that we should stick to the same cycle as this meeting
19:11:30 <kohsuke> if that's not contested, then we can move on, I think
19:11:41 <ogondza> lets move on
19:11:53 <rtyler> #topic Jenkins Certification updaet and trademark approval request
19:11:58 <rtyler> #info https://wiki.jenkins-ci.org/display/JENKINS/Jenkins+Certification
19:12:13 <rtyler> so let me first apologize on behalf of CloudBees for epicly dropping the ball on this
19:12:31 <rtyler> the last meeting CloudBees shared updates on this was mid-last year, which is not okay
19:13:05 <rtyler> in the last meeting about this there were a number of loose ends and questions which I have tried and answer in the above wiki page
19:13:50 <rtyler> this includes questions from ogondza, orrc and KostyaSha
19:14:56 <rtyler> additionally for a number of reasons we were not able to get a non-CloudBees affiliated member of the community involved the with certification advisory board, so I would like to ask for anybody interested in that to ping me at my work email (tyler@cloudbees.com)
19:15:25 <rtyler> #info Certification Advisory Board details: https://wiki.jenkins-ci.org/display/JENKINS/Jenkins+Certification#JenkinsCertification-AdvisoryBoard
19:15:53 <batmat> Nitpicking: shouldn't, like for Jenkins the app itself, it be named something like *CloudBees* Jenkins Certification?
19:16:16 <rtyler> that's not nitpicking that touches on one of the things we've dropped the ball on :)
19:16:39 <batmat> :)
19:16:43 <oleg-nenashev> And the program is already announced, right?
19:16:45 <rtyler> so before I jump to that, one of the major themes of questions originally was around financing of this operation
19:16:58 <rtyler> it's safe to safe that CloudBees is definitely not making money on this :P
19:17:01 <rtyler> safe to say*
19:17:20 <rtyler> I've tried to detail what I'm able to detail around the financing of this in the document, no Jenkins project funds are being used here
19:17:43 <rtyler> oleg-nenashev: yes, the program is already announced, which is part of the previously mentioned ball dropping :P
19:18:22 <oleg-nenashev> OK, I have to abstain from this discussion in any case
19:18:30 <rtyler> batmat: so CloudBees Jenkins implies a specific "distribution" of Jenkins that CloudBees releases
19:18:47 <rtyler> so CloudBees has developed a "CloudBees Jenkins Certification" program *and* a "Jenkins Certification" program
19:19:00 <batmat> rtyler: yeah, so more like Jenkins *CloudBees* Certification :)
19:19:01 <rtyler> the Jenkins Certification program is oriented 100% around the Jenkins open source project
19:19:44 <rtyler> it is this program we're discussing here, and what CloudBees really needs (in my personal opinion) an non-CloudBees affiliated member of the community on the certification advisory board
19:19:48 <rtyler> SO
19:19:51 <rtyler> lots of updates AMIRITE
19:20:01 <batmat> rtyler: yes, I've read that. Just saying that that wording may imply this is the officially blessed/only one for the Jenkins community project, when some other (hypothetic, granted) companies could do the same
19:20:36 <rtyler> to my knowedge, to date, no other companies have invested in creating such a program
19:20:37 <kohsuke> The intent has been to make this program belong to the  community --- I think such certification has more values to those who take them and to this community.
19:21:13 <KostyaSha> please define what is community
19:23:33 * rtyler things of non-snarky ways to answer that
19:23:36 <batmat> rtyler: sure, hence my /hypothetic/
19:24:31 <rtyler> KostyaSha: is your question not really answered in the introduction of the wiki page I linked above?
19:24:47 <rtyler> at least as far as the intended audience and purpose of the certification program
19:26:01 <rtyler> so the second part of this topic is the trademark sublicense request, which was mistakenly not requested previously
19:27:21 <rtyler> #info the two derivative marks CloudBees is seeking to use are: "Jenkins Certification" to refer to the Jenkins open source related certification program, and "Certified Jenkins Engineer" to be used to reference practitioners who successfully complete the open-source related exam
19:28:29 <rtyler> #info https://wiki.jenkins-ci.org/display/JENKINS/Governance+Document#GovernanceDocument-Trademark
19:30:33 <KostyaSha> Are questions available somewhere? Some examples
19:30:40 <autojack> so the topic on the table is approval for CloudBees to use these marks?
19:30:50 <orrc> KostyaSha: there's a study guide on the CB page with some samples
19:31:10 <rtyler> orrc: I actually don't have a link handy for that
19:31:13 <rtyler> autojack: yes
19:31:13 <KostyaSha> orrc, i thought non CB part was already done and now request is only about CB
19:31:38 <KostyaSha> orrc, will be glad to see non CB cert questions to get idea what is this about
19:32:25 <autojack> +1 for the sublicense.
19:32:31 <alyssat> the study guide is https://www.cloudbees.com/sites/default/files/cje_study_guide_final.pdf
19:32:47 <alyssat> the FAQ is https://www.cloudbees.com/jenkins-certification
19:32:50 <oleg-nenashev> allyssat: It's CJE, not open-source
19:32:53 <KostyaSha> alyssat, as foss member i'm not using any CB* staff
19:33:33 <oleg-nenashev> ahh, CJE = Certified Jenkins Engineer
19:33:35 <heidig__> CJE = Certified Jenkins Engineer
19:33:45 <rtyler> we gotta stop using all these silly acronyms :P
19:33:52 <alyssat> kostyaSha: that's what currently avail.
19:33:57 <KostyaSha> So if the question about sub-licensing for "$company jenkins certificaiton" i would be glad to see firstly plain "jenkins certification". Either there will be possible 2 orthogonal licenses.
19:33:59 <heidig__> rtyler +1
19:34:29 <rtyler> KostyaSha: same here :P
19:36:01 <KostyaSha> Jenkins Engineer? is it also approved mark?
19:36:09 <kohsuke> KostyaSha: I'd rather see those efforts join hands together if somebody else wants to do it
19:36:53 <danielbeck> KostyaSha See rtyler's #info above, it's one of the requested marks
19:37:16 <KostyaSha> danielbeck, oops
19:37:22 <kohsuke> ... which is why the program proposal emphasizes the community oversights (aka "advisory board")
19:37:52 <KostyaSha> is it possible to approve everything CB wants, but without enforcing everything to Jenkins itself?
19:38:10 <rtyler> KostyaSha: what do you mean? enforcing everything to Jenkins?
19:38:15 <kohsuke> And it's hard to see how anyone else wants to develop a competing certification
19:38:22 <rtyler> heh
19:38:35 <KostyaSha> imho there is no project resources (non-CB) for making jenkins cert programs etc.
19:38:52 <rtyler> developing certification examples is a great way to take a big pile of money and turn it into a small pile of money
19:39:08 <rtyler> KostyaSha: right, and that's not great but fine from my perspective
19:39:24 <rtyler> I really would like to see some more people express interest and commit time to help review the questions via the advisory board
19:40:02 * KostyaSha can't imagine possible questions at all
19:40:17 * autojack is interested and willing to commit time to etc etc etc
19:40:33 <rtyler> KostyaSha: given your experience level, I wouldn't imagine the questions would be terribly difficult for ou
19:40:35 <autojack> I've been out of the Jenkins loop for a couple of months but this sounds like something I could do.
19:40:42 <autojack> so there's that.
19:41:04 <kohsuke> going forward I believe the main effort will be around revisiting problematic/stale questions and adding new ones
19:41:32 <orrc> some of this feels a bit odd. community members should donate their time to help develop a proprietary set of questions, to help CloudBees charge $250 for certification :)
19:41:33 <rtyler> not to derail signing autojack up for more work, but any more comments on the "Jenkins Certification" and "Certified Jenkins Engineer" marks?
19:41:51 <KostyaSha> rtyler, from my exp i want say that there is messy core (i.e. crazy matrix job with unclear flyweights logic) and plugins written by devs that may change in any time.
19:42:04 <kohsuke> FWIW I took the test and I was impressed at the quality
19:42:14 <KostyaSha> i.e. CB may keep all features/plugins/bundle and exam/cert for it, but foss...
19:42:59 <batmat> I think I could help too. Though looking at the list, there's some plugins I actually never used.
19:43:13 <rtyler> orrc: I think that's a good point we can address
19:44:12 * rtyler ponders
19:44:19 <kohsuke> orrc: as described in the costs/revenues it's very unlikely that we'll ever turn profits, but would it help if we commit to transfer any "profits" to SPI?
19:44:24 <KostyaSha> why not just make coursera course?
19:44:26 <orrc> a lot of this is based on good faith, e.g. making sure that the questions aren't stupid / easy, so that the community can "protect the good name of Jenkins"
19:44:42 <KostyaSha> people may then get 'coursera jenkins cert' for example
19:44:42 <orrc> but then CloudBees is also incentivised to make sure it's a decent program
19:45:06 <orrc> it's hard to argue against the whole thing, but it feels slightly and unhelpfully intangibly odd :)
19:45:15 <rtyler> heh
19:45:47 <batmat> +1 orrc nicely worded :)
19:45:55 <danielbeck> kohsuke What was your alpha test score again?
19:46:07 <batmat> danielbeck: he didn't pass ;)
19:46:27 * batmat lols
19:46:27 <rtyler> hah
19:46:36 <kohsuke> danielbeck: not as good as it should have been
19:46:43 <rtyler> we're getting offtopic here
19:46:50 <rtyler> we can all give kohsuke grief about his test scores later :)
19:47:03 <danielbeck> trying to address concerns of it being too easy
19:47:16 <orrc> so I think it's ok. but the advisory board member(s?) should be a bit clearer; how big is the board? do community members have any power, e.g. to veto questions?
19:47:36 <rtyler> orrc: agreed, I'll follow up on that and put more definition into it
19:47:55 <autojack> is there any other question on the table besides approving the trademark stuff?
19:48:03 <autojack> it sounded like no, in which case as I said, +1.
19:48:05 <rtyler> autojack: no, just updates on the program
19:48:08 <autojack> ok.
19:48:36 <orrc> kohsuke: I'm surprised that it may never make a profit, but I remember having the feeling last time that it seemed reasonable that the community should benefit from its name being used like this
19:48:57 <orrc> but as the page explains, that may well be in intangible ways. which I do kinda believe in :)
19:49:10 <autojack> yeah, I agree about the intangibles.
19:49:16 <autojack> actually they are probably reasonably tangible.
19:49:36 <autojack> hey, layoffs hitting my team as we speak, can I get a discount on certification?? :P
19:49:40 <orrc> I had a lot of [citation needed] in my head reading that page
19:50:02 <rtyler> #action rtyler to put more clear definition into the size and scope of the certification advisory board
19:50:27 <rtyler> #action rtyler to add more tangible benefits for the Jenkins project to the certification page
19:50:41 <heidig__> autojack - it's $50 right now, during the beta program. Still room in some cities to take the exam
19:51:02 <orrc> rtyler: super. I'm also ok on a trademark sublicence. not sure if we've mentioned before that these are non-exclusive, but in this case it would certainly be good
19:51:30 <rtyler> I'm somewhat biased, but I agree on the terms that orrc suggest
19:51:55 <rtyler> #action rtyler to harass kohsuke on what kind of revenue or other benefits we can bring back into the Jenkins project from certification
19:52:11 <rtyler> any other thoughts on the trademark sublicensing here?
19:52:23 <orrc> especially from non-CloudBees employees
19:52:25 <rtyler> I realize I've taken a boatload of time from this meeting on the subject, but IMO it was needed
19:52:29 * rtyler pokes ogondza
19:52:46 <rtyler> hare_brain: any thoughts to share?
19:52:48 <oleg-nenashev> yes this is a must-have discussion
19:52:55 * oleg-nenashev abstains in any case
19:53:38 <hare_brain> Others have been doing a great job in asking pertinent questions. I have nothing to add.
19:53:57 <rtyler> hare_brain: non-CB thoughts on the sublicensing request?
19:53:58 <batmat> hare_brain: your opinion would help I guess :)
19:54:30 * orrc misses being able to poke the entire channel, like in Slack
19:54:38 <orrc> (and emoji reactions, of course)
19:55:00 <batmat> I'm personally globally OK, though my only slight discomfort is the one I expressed at the beginning.
19:55:06 * rtyler nods
19:55:19 <hare_brain> @rtyler: I don't have any issues with the sub-license request, aside from the intangibles others have talked about.
19:55:53 <rtyler> sooooo, yes? :P
19:56:48 <orrc> :(
19:57:17 <autojack> hey, my +1 should be enough for anything.
19:57:20 <hare_brain> No objections.
19:57:41 <batmat> hare_brain: so this is a +0 ? :)
19:57:44 <rtyler> okay; with my cloudbees hat on I'm committing to getting my action items done on this topic prior to next meeting
19:57:51 <hare_brain> LOL.
19:57:55 <hare_brain> ! -1
19:58:00 <rtyler> can we move on to my next big topic?
19:58:04 * rtyler looks at the clock
19:58:05 * batmat likes Apache's voting way
19:58:09 <autojack> please do.
19:58:13 <autojack> I'm getting hungry.
19:58:25 <rtyler> #topic New team lead positions: Events and Release
19:58:34 <rtyler> #info https://wiki.jenkins-ci.org/display/JENKINS/Proposal+-+Project+sub-teams
19:58:38 <rtyler> #info https://wiki.jenkins-ci.org/display/JENKINS/Team+Leads
19:58:57 <rtyler> first let's talk about the Events team lead, responsibilities have been defined here: https://wiki.jenkins-ci.org/display/JENKINS/Team+Leads#TeamLeads-Events
19:59:33 <rtyler> my goal is to create this position and open the call for candidates to help run and coordinate the ever increasing Events responsibilities we have on our plates
19:59:36 <autojack> still references the now-deprecated CIA.
19:59:43 <rtyler> yeah, gotta remove that
19:59:53 <rtyler> since we deadened CIA, I thought this would be a good time to bring this up
20:00:12 <rtyler> autojack: removed
20:00:14 <autojack> k
20:00:37 <autojack> I'm reading this and thinking, with a smile, don't we already have alyssat? :D
20:00:48 <rtyler> and similar to Security Officer, some amount of additional access would be delegated to the Events officer to help perform these duties
20:01:18 <autojack> it sounds like a solid idea.
20:01:30 <rtyler> are there any objections to this position?
20:01:40 <oleg-nenashev> +1
20:01:41 <autojack> I think it would have to be someone who is both dedicated and unusually well-organized.
20:01:58 <batmat> +1
20:02:34 <rtyler> if there are no objections, I'll jump to the other one
20:02:40 <kohsuke> I think this is an important role that would help facilitate communication with people outside
20:02:44 <autojack> I almost think it would work best with >1 person in the role.
20:02:45 <rtyler> #action rtyler to send call for candidates out for Events team lead
20:02:56 <rtyler> autojack: are you volunteering for this too? :)
20:03:02 <rtyler> so the Release team lead
20:03:11 <rtyler> #info https://wiki.jenkins-ci.org/display/JENKINS/Team+Leads#TeamLeads-Release
20:03:26 <alyssat> I'm volunteering :o)
20:03:39 <batmat> alyssat: for the Release team lead?
20:03:39 <batmat> :)
20:03:44 <oleg-nenashev> alyssat lol
20:03:48 <rtyler> with my infra hat on, my primary goal here is have somebody trusted to manage releases and keys, typically a job that only kohsuke has done
20:04:02 <alyssat> batman: of course not..events lead :o)
20:04:23 <kohsuke> yes, for me, this is primarily to open up the access to the release keys in a controlled way
20:04:23 <rtyler> with regards to LTS, as is mentioned in the responsibilities above, this person would either be responsible for performing relaeses directly, or coordinating with a documented release manager for a release line(s)
20:05:13 <oleg-nenashev> rtyler kohsuke: Does it involve release automation on CI?
20:05:14 <KostyaSha> maybe project will finally has project keys instead kohsuke's keys
20:05:22 <rtyler> KostyaSha: *yesssssssssss*
20:05:25 <autojack> +1 to this idea as well. I'm on board with anything that reduces the project's SPOF.
20:05:26 <danielbeck> KostyaSha Work in progress by abayer
20:05:27 <rtyler> oleg-nenashev: yes
20:05:38 <KostyaSha> in 2016 /o/ in automation project (party)
20:05:42 <rtyler> \o/
20:05:51 <rtyler> are there any objections to me opening a call for candidates for this position?
20:06:16 <oleg-nenashev> rtyler: Is there an overlap with LTS team?
20:06:24 <oleg-nenashev> Maybe a single lead is enough
20:06:30 * batmat thinks alyssat will call me batmat forever :)
20:06:40 <ogondza> as I understand this some/one of the release team members will be the LTS guy
20:06:41 * batmat batman*
20:06:49 <rtyler> ogondza: correct
20:06:52 <KostyaSha> oleg-nenashev, they should help each other :)
20:07:12 <alyssat> batmat: gotta love auto correction :o)
20:07:18 <kohsuke> I think we can record agreement and move on
20:07:20 <autojack> no objections.
20:07:30 <oleg-nenashev> i'm +1 on the role in any case
20:07:32 <batmat> +1
20:07:34 <rtyler> alright, batmat I need to shuffle one more time-sensitive topic up ahead of yours
20:07:45 <rtyler> #topic Google Summer of Code 2016 application process
20:07:55 <oleg-nenashev> #info https://groups.google.com/forum/#!topic/jenkinsci-dev/vTxsCLk1XS0
20:08:04 <oleg-nenashev> #info https://docs.google.com/document/d/1dv9UXyT_FPukJXrBYauIKvxrZeEU6BT7v90LSQX36MU/edit#
20:08:07 <rtyler> this should mostly be just an update, so we can get through it real quick
20:08:16 <oleg-nenashev> #info https://wiki.jenkins-ci.org/display/JENKINS/Google+Summer+Of+Code+2016
20:08:24 <rtyler> #info the project application deadline is this friday (Feb 19th)
20:08:33 <oleg-nenashev> See the ML for details
20:08:56 <oleg-nenashev> As a summary, we're almost ready to apply
20:09:01 <KostyaSha> danielbeck, https://groups.google.com/d/msg/jenkinsci-dev/6NXjjQq5VgQ/ahPIPIBd3OAJ :D
20:09:09 <oleg-nenashev> So we just need reviews of the application docs
20:09:25 <rtyler> please please please get your comments in on the application document oleg-nenashev prepared within the next 24 hours ish
20:09:33 <kohsuke> IIRC, we need to have mentors ready
20:09:38 <rtyler> kohsuke: we do
20:09:50 <kohsuke> Do we have a mental tally of who?
20:09:58 <oleg-nenashev> We need MOAR mentors, but we have a startup pack
20:10:11 <oleg-nenashev> kohsuke: see Wiki
20:10:22 <rtyler> oleg-nenashev: is drulli not participating/
20:10:31 <kohsuke> Wow
20:10:41 <kohsuke> I am impressed
20:10:42 <oleg-nenashev> kohsuke: there are also 3 almost-ready mentors, but I'm waiting for their final agreement
20:10:53 <rtyler> cool, can we move onto the last topic oleg-nenashev ?
20:11:14 <oleg-nenashev> rtyler drulli finaly decided to skip this year. Continuous baby delivery
20:11:18 <rtyler> hah
20:11:27 <rtyler> that's a waterfall project if I've ever seen one
20:11:38 <oleg-nenashev> Yes, we can go forward if there is no comments/proposals
20:11:39 <orrc> was there any feedback or lessons learned from the 2009 rejection?
20:11:49 <orrc> (I guess that's off-topic)
20:11:51 <rtyler> orrc: not that I'm aware of
20:11:58 <rtyler> #topic clarify that plugins available through the update center are required to have their source code canonical repository hosted under the Jenkinsci GitHub organization
20:12:03 <kohsuke> this program proposal is much more rich than the last round
20:12:15 <oleg-nenashev> orrc: I've read the ML regarding this topic && review competitor programs
20:12:17 <kohsuke> and it was also tied up between Google/Sun disputes on some IP
20:12:28 <rtyler> batmat: you're on :)
20:12:29 <orrc> ouch, ok
20:12:38 <oleg-nenashev> Now we're in much better shape, I hope
20:12:49 <kohsuke> Yes, I think so. Fingers crossed
20:13:43 <kohsuke> Going back to the topic at hand...
20:13:49 <rtyler> batmat: one of the things I had thought of last week was perhaps defaulting the parent pom to release into the experimental update center by default, which would satisfy nigel's concerns, and then make the jenkinsci org requirement for inclusion in latest/stable update centers
20:14:13 <rtyler> now, I had this idea with very little idea of how this would actually work
20:14:16 * rtyler doesn't maven good
20:14:38 <batmat> rtyler: actually, experimental is by version pattern only
20:14:43 <rtyler> ORLY
20:14:49 * rtyler facepalms
20:14:52 <orrc> yup
20:14:54 <KostyaSha> why facepalm?
20:15:07 <danielbeck> makes the suggestion not work
20:15:11 <orrc> and there's no way of knowing, when generating the UC, where the plugin is hosted
20:15:16 <batmat> rtyler: but the idea is great, if it was feasible this way :)
20:15:17 <orrc> (though it is a good idea)
20:15:17 <kohsuke> With a bit of automation we could certainly do what you are suggesting
20:15:21 <rtyler> KostyaSha: the update center generation is a /little/ ugly with various hacks like that in it :P
20:15:27 <KostyaSha> jenkins plugins distribution is facepalm ;)
20:15:43 <rtyler> I don't want kohsuke to throw anything at me, so..no comment :)
20:16:04 <batmat> rtyler: one clean(er) solution would to redefine the distributionManagement tag so that there's many repos
20:16:29 <KostyaSha> batmat, that would add mess to already existing plugins/configs
20:16:42 <batmat> anyway, I fear we may be slightly offtopic.
20:16:58 <kohsuke> I forgot exactly how I did it, but we can know what repository contains what plugin and their versions
20:17:08 <orrc> so long as there's no check made when uploading to the repository, then that isn't really enforcable
20:17:09 <rtyler> batmat: bring us back ontopic :)
20:17:24 <batmat> As stated by abayer kohsuke and many others, the fact that plugin source code has always been assumed to be under the hudson/jenkins infra has always been assumed
20:17:34 <batmat> the thing is, legacy.
20:17:48 <abayer> It's been assumed but not required, yeah.
20:18:05 <KostyaSha> or i may say that it optional
20:18:10 <KostyaSha> either host either not
20:18:26 <batmat> having had to look up a plugin source code after a user's question on #jenkins some time ago, I really feel this should be required at some point
20:18:42 <batmat> since there's apart from that really not a lot of enforcement
20:18:50 <abayer> Based on the defaults, I'd say hosting outside of jenkinsci org has been optional, rather than hosting inside of jenkinsci org being optional - you *can* opt out, but you have to specifically do so.
20:18:51 <batmat> so, IMO would sound quite reasonnable
20:18:52 <rtyler> the adopt-a-plugin thing makes me feel this should be a requirement
20:19:02 <rtyler> so folks can pick up on plugins that have been left behind
20:19:06 <batmat> rtyler: exactly
20:19:40 <KostyaSha> rtyler, in such cases they are forked
20:19:54 <KostyaSha> so there is no problem
20:20:07 <rtyler> KostyaSha: that assumes that the source code was ever available
20:20:10 <danielbeck> KostyaSha from non public source code?
20:20:30 <kohsuke> one of the nigel's concern was that we might set additional bar to the entry into the jenkinsci org.
20:20:39 <KostyaSha> danielbeck, let's check, topic is about hosting under jenkinsci not about public/unpublic?
20:20:44 <batmat> with that bunch of plugins we have, unmaintained plugins arise almost every other day. And having to dig around to find the associated source code seems really a waste of time.
20:20:45 <rtyler> so there are some open source projects, nibalizer from openstack was in here earlier, that host their plugins in their own open source infrastructure
20:20:47 <kohsuke> Maybe re-affirming that such bar shouldn't be arbitrary might help
20:21:11 <abayer> Re-affirming wouldn't hurt regardless. =)
20:21:30 <rtyler> KostyaSha: we don't have private repos under jenkinsci fwiw
20:22:16 <rtyler> I'm curious if we have proprietary plugins in our update center, and if so, do we wish to continue distributing them
20:22:19 <KostyaSha> rtyler, danielbeck mean sources: 1) under jenkinsci (public) 2) under user/company accounts(public) 3) unavailable at all
20:22:21 <batmat> We basically generally fork repo under a week, discussions/roundtrips included
20:22:24 <danielbeck> KostyaSha anywhere else is essentially non-public as the author can decide to delete the repo.
20:22:42 <rtyler> this proposal would definitely affect those developers
20:22:50 <KostyaSha> danielbeck, sources will exist in jars
20:22:53 <orrc> rtyler: there may be one or two floating around; we've certainly removed a couple in the past year
20:23:10 <batmat> so saying this is a limiting requirement is IMO a bit excessive. It's IMO actually excessively simple to host a new plugin under the jenkinsci org...
20:23:36 <orrc> KostyaSha: someone published a plugin, which *explicitly* excluded the sources
20:23:44 <batmat> rtyler: about proprietary plugins, I'm pretty sure this was already clarified as forbidden
20:23:50 <rtyler> danielbeck: do you have guidelines floating around for HOSTING requests? I don't want forking to be blocked by a single persons subjective opinion on the merit of a fork
20:23:55 <rtyler> batmat: okie doke
20:24:06 <KostyaSha> orrc, then you can request sources according to license?
20:25:01 <batmat> #info "The Jenkins project does not host closed-source plugins" https://wiki.jenkins-ci.org/display/JENKINS/Hosting+Plugins
20:25:04 <KostyaSha> of course closed source is bad :)
20:25:10 <rtyler> batmat: excellent
20:25:12 <rtyler> thanks
20:25:20 <abayer> I could buy into the idea of an "external origin" update center that you'd end up in if your repo wasn't under jenkinsci.
20:25:33 <abayer> As long as we integrated that "external origin" UC into the default UI somehow.
20:25:45 <oleg-nenashev> Or just a metadata with a switch in advanced options
20:25:51 <KostyaSha> abayer, could you provide tooling for this? Pinnin repos, priorities, etc?
20:25:52 <oleg-nenashev> It should be more simple
20:26:04 <abayer> KostyaSha: Pinning repos?
20:26:07 <KostyaSha> i.e. i would like to use GH releases
20:26:11 <batmat> abayer: IIRC jglick explained a fortnight ago that actually internals already support in standard multiple UC
20:26:15 <batmat> just this isn't exposed
20:26:21 <orrc> abayer: Jenkins supports multiple UCs. just expose this in the UI, and let others run their own
20:26:27 <KostyaSha> abayer, see OpenSuse dependency management (imho the best dep solving algos)
20:26:39 <abayer> I'd say that a plugin ID would only show up in one of "external" or "primary" update centers - newest version wins.
20:26:45 <abayer> Just spitballing here. =)
20:26:55 <orrc> I wouldn't like to have UCs not controlled by Jenkins / hosted by others in the default Jenkins install
20:26:56 <KostyaSha> abayer, that's standard mistake :)
20:26:56 <rtyler> zypper in github-plugin
20:26:58 <orrc> abayer: yeah
20:26:59 <rtyler> \o/
20:27:06 <jglick> batmat: a little trickier than that since you do actually need some code to manage signing certificates
20:27:08 <batmat> KostyaSha: the one who proposes every combinatorial to support should commit to work on it
20:27:31 <kohsuke> For me, forcing people into this one community is a part of the explicit design to encourage co-maintainerships and etc
20:27:34 <KostyaSha> jglick, right, certs is issue. And later ratings and other metadata staff
20:27:44 <abayer> We'd still "support" and generate the external update center, it'd still only contain plugins in our Maven repo, it'd just be plugins with source outside of jenkinsci org.
20:27:48 <kohsuke> I don't want to end up with lots of micro update centers running everywhere
20:27:59 <abayer> a la experimental.
20:28:20 <abayer> (except with a better user experience that doesn't require manually adding it to get it in the in-Jenkins plugin manager)
20:28:43 <orrc> yeah, seems reasonable
20:28:46 <KostyaSha> kohsuke, UC should die, you may create repo sources in plugin manager. That how all linux distros leaving for 20 years
20:28:53 <orrc> ci.jenkins-ci.org groans at the additional work ;)
20:28:54 <abayer> OH GOD no.
20:28:57 <rtyler> as we're running up on 90 minutes, could I ask you abayer and batmat to put together a concrete proposal on what we would move forward with here based on this feedback?
20:29:02 <abayer> Sure thing.
20:29:10 <KostyaSha> kohsuke, i.e. central + gh source for some plugin + gh source for other plugin
20:29:12 <kohsuke> I need to read nigel's concern more carefully, but I think our current proposal is sensible
20:29:32 <batmat> +1. Gonna be difficult to extract some seemingly beginning of consensus here, but anyway we can try
20:30:02 <rtyler> #action batmat and abayer to try to extract a coherent proposal based on today's IRC discussions and prior mailing list thread comments
20:30:03 <abayer> fwiw, I'd even say that we only generate the "split" update centers for Jenkins releases after we've added proper integration of the split update centers in the UI, so that older versions still see what they see now.
20:30:08 <kohsuke> nigel has been a great member of the community with lots of popular plugins under the belt, so I want him happy
20:30:21 <kohsuke> that's the challenge, I guess.
20:30:26 <KostyaSha> lol
20:30:55 <rtyler> can we button up this topic until we get a big beautiful proposal written out? :)
20:30:59 <abayer> batmat: Let's start emailing on this.
20:31:06 <batmat> +1 KostyaSha
20:31:09 <batmat> oups
20:31:13 <batmat> +1 kohsuke
20:31:15 <kohsuke> rtyler: yeah let's wrap up
20:31:20 <rtyler> #topic next meeting
20:31:23 <batmat> abayer: right
20:31:42 <rtyler> by my calendar, March 2nd is our next scheduled meeting
20:32:14 <orrc> oh right, leap year
20:32:17 <rtyler> I don't think there are any major national holidays scheduled that day
20:32:27 <rtyler> same time same place?
20:32:34 <kohsuke> as always
20:32:42 <batmat> a pleasure
20:32:51 <rtyler> thanks everybody for participating
20:32:57 <kohsuke> #info next meeting is March 2nd, same time
20:33:02 <rtyler> #endmeeting