On Monday, December 21, 2015 at 2:58:34 AM UTC+3, Stephen Connolly wrote: > > If we are going to go down the road of forbidding direct committing to > master and forcing people to go through PRs (let's assume we can find a way > to let release commits go through) we'd need a better criteria for when we > can actually merge PRs. > > I see lots of PRs languishing for ages with disagreement over what to do > and no clear outcome one way or the other... > > We have 25 PRs that are older than Jun 1st 2014... > > 50 PRs that are at least 1 year old > + i closed all my >1 year old PRs when realised that 2.0 will be cosmetic.
> > 75 PRs that are 6 months or older > > Now KS, it's not CloudBees place to decide what the community wants to > have merged... the community has not defined how to address these PRs... > > If we are to move to PRs without direct commit then I want to see a > defined process whereby no PR goes more than say 1 month without the > community deciding if it is a Go / No Go on the general idea. > > I am quite sure that CloudBees would be willing to help get PRs into a > better state for merging if we knew that those PRs were the direction the > community wanted to go. Right now we seem to end up saying "ok this is what > we think, here's our contribution, do you want it?" and there is no > movement further... after a while we then remove our CloudBees hat and don > our core contributors hat and say something like "ah for jebus's sake, > nobody else has expressed an opinion either way for the past 2-3 weeks, > let's just merge it" but don't for one second think that we like doing this. > Do you really think that community has enough core devs not already hired by CB? I have problem with classloading and only 3 guys (you are one of) may have answer, all in CB. > > I would say that the community needs to show interest in PRs before we can > switch to a PR model as the route for change. > PR instead direct commit at least provides chance to get review, i think you are expecting review in PRs and not in master tree. Such review already happens, i.e. Daniel provides comments in all PRs. > > My suggestion is that when a PR has been open for a week or so, the > community should start a vote thread to decide if the change is the right > direction (Go) or the wrong direction (No Go). > I may say that if it was open for 1 week, then community had enough time for review. Obviously CB relies on ability merging their payed changes faster and somebody may kick you for such blocker proposals :-) > If No Go then close the PR providing the reason... if Go then the PR > author can be helped to get the PR to a mergeable quality and then we merge > the change and move forward. > And if it doesn't fit into current 1.x, then wait for 10 years to get ability (One of my first contributions https://github.com/jenkinsci/jenkins/pull/902#issuecomment-134412342 ) After CB hiring not all changes can be reviewed i.e. fundamental https://github.com/jenkinsci/jenkins/pull/1936#issuecomment-162191335. Of course there was no objections for merge. > > If that process gets all the open PRs down such that most PRs are open for > no more than 1 month, then and only then would I say that preventing direct > core commits might be worth pursuing... > In PRs you may have degree of go asap/wait for community, while in direct commits there is nothing than "don't care". > Just my €0.02 > > -Stephen > Stephen, btw, could you attend meetings on IRC? > > On 20 December 2015 at 17:22, Andrew Bayer <[email protected] > <javascript:>> wrote: > >> So, addressing a few aspects of this thread: >> >> - I'd strongly oppose ICLA/push permission revocation for pushing >> directly to master. That's overly harsh. >> - I do support this policy overall - I'm personally a big fan of a >> "Review then Commit" policy. >> - There is a caveat/exception, of course - release-related commits. >> >> I think this is worth proposing for the next meeting - Kostya, could you >> add it to the agenda on the wiki? There's no need to name-and-shame >> specific cases of people pushing directly to master - this is a worthwhile >> policy to advocate even if no one was actually breaking it at this point. >> >> A. >> >> >> On Sun, Dec 20, 2015 at 9:41 AM, Kanstantsin Shautsou < >> [email protected] <javascript:>> wrote: >> >>> >>> On Dec 20, 2015, at 17:32, Baptiste Mathus <[email protected] >>> <javascript:>> wrote: >>> >>> +1 with all Oleg said... >>> The subject might indeed be eligible to discussion, and I also think we >>> might want to proceed with only PRs, but the way you do it... >>> And the name you use for kk in CC is, well… >>> >>> Name was allowed, see meeting logs. >>> >>> >>> 2015-12-20 15:26 GMT+01:00 Oleg Nenashev <[email protected] >>> <javascript:>>: >>> >>>> Hi Kostya, >>>> >>>> I understand your concern, but messages of such kind can be considered >>>> as a personal offense. >>>> >>> Any question can be transformed in any way you want. >>> >>> Kohsuke is not the only person committing in such way, so it's >>>> definitely a wider problem, which requires a discussion. BTW currently >>>> there is no policy prohibiting such approach, so the direct commits are >>>> generally valid even if they smell bad. >>>> >>> Never saw anybody else, could you share more examples? >>> >>> >>>> I'm +1 on prohibiting direct pushes to the master branches for >>>> everybody and in all repos. And Jenkins core core is not an exception. >>>> It makes the current release and changelogging approach a bit >>>> problematic, but it's another story. >>>> >>>> if you signed ICLA and do some questionable changes into master (here i >>>>> see 2 violations) person should get core access removal, right? >>>> >>>> >>>> Nope. There is no such policy in Jenkins project. If you have any >>>> concerns about particular contributors, raise the topic to the governance >>>> meeting. It's the *ONLY* way for discussing such topics. >>>> >>> That what core committers said to me when i asked about ICLA and perms. >>> Would be glad to see documented way without double standards. >>> >>> >>>> >>>> воскресенье, 20 декабря 2015 г., 17:03:40 UTC+3 пользователь >>>> Kanstantsin Shautsou написал: >>>>> >>>>> Situation: people doing reviews, blocking PRs for weeks,months,years >>>>> while some people do direct commits to core master without any reviews. >>>>> This ends to situations when master gets broken state that reflects on >>>>> PR builds verification, i.e. >>>>> https://github.com/jenkinsci/jenkins/commit/d86a88ab042cc55530d91e745af9e0886e8eeb79 >>>>> Unreviewed changes adds chaos. While people reviewing and close to get >>>>> rid of unconfigurable settings in >>>>> https://github.com/jenkinsci/jenkins/pull/1914 one person is doing >>>>> direct master changes >>>>> https://github.com/jenkinsci/jenkins/commit/653fbdb65024b1b528e21f682172885f7111bba9 >>>>> >>>>> <https://www.google.com/url?q=https%3A%2F%2Fgithub.com%2Fjenkinsci%2Fjenkins%2Fcommit%2F653fbdb65024b1b528e21f682172885f7111bba9&sa=D&sntz=1&usg=AFQjCNGfejtJii5ClN4CxzyDP_BzWpWFag> >>>>> >>>>> Proposal: stop doing such unreviewed changes and forbid direct master >>>>> commits (either at all, either only for mentioned person). >>>>> >>>>> PS. AFAIR/AFAIK if you signed ICLA and do some questionable changes >>>>> into master (here i see 2 violations) person should get core access >>>>> removal, right? >>>>> >>>>> >>>> -- >>>> You received this message because you are subscribed to the Google >>>> Groups "Jenkins Developers" group. >>>> To unsubscribe from this group and stop receiving emails from it, send >>>> an email to [email protected] <javascript:>. >>>> To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/jenkinsci-dev/630f6633-d494-4726-8c21-d63890ce040a%40googlegroups.com >>>> >>>> <https://groups.google.com/d/msgid/jenkinsci-dev/630f6633-d494-4726-8c21-d63890ce040a%40googlegroups.com?utm_medium=email&utm_source=footer> >>>> . >>>> >>>> For more options, visit https://groups.google.com/d/optout. >>>> >>> >>> >>> >>> -- >>> Baptiste <Batmat> MATHUS - http://batmat.net >>> Sauvez un arbre, >>> Mangez un castor ! >>> >>> -- >>> You received this message because you are subscribed to a topic in the >>> Google Groups "Jenkins Developers" group. >>> To unsubscribe from this topic, visit >>> https://groups.google.com/d/topic/jenkinsci-dev/d64UIypbzh0/unsubscribe. >>> To unsubscribe from this group and all its topics, send an email to >>> [email protected] <javascript:>. >>> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/jenkinsci-dev/CANWgJS61D%2BXzO0Xd8jV2icNU27mJSTGZM5R0OM8rNme12ZnEFg%40mail.gmail.com >>> >>> <https://groups.google.com/d/msgid/jenkinsci-dev/CANWgJS61D%2BXzO0Xd8jV2icNU27mJSTGZM5R0OM8rNme12ZnEFg%40mail.gmail.com?utm_medium=email&utm_source=footer> >>> . >>> For more options, visit https://groups.google.com/d/optout. >>> >>> >>> -- >>> You received this message because you are subscribed to the Google >>> Groups "Jenkins Developers" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected] <javascript:>. >>> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/jenkinsci-dev/D11D841C-4F09-4893-95C4-F5B146F25C88%40gmail.com >>> >>> <https://groups.google.com/d/msgid/jenkinsci-dev/D11D841C-4F09-4893-95C4-F5B146F25C88%40gmail.com?utm_medium=email&utm_source=footer> >>> . >>> >>> For more options, visit https://groups.google.com/d/optout. >>> >> >> -- >> You received this message because you are subscribed to the Google Groups >> "Jenkins Developers" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected] <javascript:>. >> To view this discussion on the web visit >> https://groups.google.com/d/msgid/jenkinsci-dev/CAPbPdOb9uQr5_ebS82FeMJoBp32xRKU541rkGoMrw2YTN-apaw%40mail.gmail.com >> >> <https://groups.google.com/d/msgid/jenkinsci-dev/CAPbPdOb9uQr5_ebS82FeMJoBp32xRKU541rkGoMrw2YTN-apaw%40mail.gmail.com?utm_medium=email&utm_source=footer> >> . >> >> For more options, visit https://groups.google.com/d/optout. >> > > -- You received this message because you are subscribed to the Google Groups "Jenkins Developers" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/jenkinsci-dev/75c80dd2-6e06-4676-8a1e-bfb5bee688f7%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
