Are you still a committer?

Looks like you need to check it daily yourself from now on just to be sure. In January the Eclipse PMC de-committerized a lot of committers because they were apparently inactive. Well, at least for me I can say it’s true.

I became an Eclipse Platform UI committer together with Ed Burnette when we wrote the R21Presentation for Eclipse 3.0. That was pretty much a one time implementation without ongoing development except some bug fixes after the release.

But something went incredibly wrong here. Not a single part of the “de-committerization” process is open or transparent. It all happend silently in the background. If I wouldn’t have been voted as Phoenix Committer I probably would never have discovered it. Actually, I was wondering why I have to fill all the paperwork again? The answer was hard but correct – rule 4a didn’t apply anymore to me.

I would have expect some kind of notice that my commit rights are beeing revoked in XX days because of inactivity or any other sort of notification. In fact, that could even have allowed me to potentially avoid it by doing some simple refactorings or fixes to become active again.

What do you think? Is that the right way? Is it open? Is it transparent?

11 thoughts on “Are you still a committer?

  1. I don’t think this is either open or transparent. The Eclipse is usually a great example of these qualities. However, I think it still has a long way to go to fully shake the corporate culture it has had since its inception.

  2. I wonder if this happened by accident? As a PMC member, I didn’t hear anything about such a sweep, and I assume that other PMC would at least email committers warning them about pending deactivation — this might be a good way to get inactive committers contributing again. 😉

  3. So transparent it’s invisible 🙂 Maybe Mik got to it with some Mylar and because you were focussed on another task you didn’t see it.

    PS ‘a committer’, not ‘an committer’

  4. Gunnar, I don’t see any emails from you to the PMC in either the PMC’s email list or the eclipse-dev email list or in the eclipse.platform newsgroup. Are you expecting them to read your blog? Perhaps you should tell them what you think about this in the (more-or-less official) forums, i.e., the mailing lists.

    And – while my personal opinion matches yours – that this is not very transparent – it is entirely within the project charter. Perhaps you should campaign to get the charter changed to require a public vote before inactive committers are removed?

  5. Thanks Bjorn, that is what I’m thinking about. I mean it’s weird that it takes a long process to get contributors on board with voting and all that stuff but just one meeting to get them removed. But anyway, the EMO could have at least sent out a note to the guys that they have been removed as committers on project xyz. Don’t you think?

    @Alex: Thanks, I know. I meant to write “an Eclipse committer” but obviously I failed after a 20 hours nightmare trip with British Airways from Toronto via London to Frankfurt. 😉

  6. Gunnar, I would follow up with the webmaster. I seem to recall in the past that if people have changed email addresses their commit rights have been deleted automatically. I think there is something that is triggered when the mail to the email address associated with the committer account starts bouncing, but the webmasters would know for sure.

  7. Bjorn:
    > John, you’re not a PMC member on the Eclipse PMC, you’re a PMC member on the Data Tools Platform PMC.

    Gee, thanks for telling me that: it clears up a lot of confusion with the meetings I was attending and such….

    Seriously, obviously I meant “DTP PMC” in the first instance, and assumed that it would be clear to anyone reading my comment either by (a) Associating my name with DTP or (b) Not associating my name with the platform. Next, the “other PMC” include every other PMC except DTP.

    In general, I was just trying to raise the possibility that a mistake was made, which would explain why Gunnar was not notified in advance. In my experience, Eclipse projects (no wait — the Eclipse Foundation top-level projects and their (sub)projects, hosted at, having a PMC and a number of subproject leads…) are very good about making these sorts of activities around committer status public, even if the charter or other rules allow for less public action.

  8. From looking at the commits explorer, you made seven commits in 2005, and zero in 2006. That seems to fit the definition of inactive committer. Giving you advanced notice so you can do a few simple refactorings to maintain your commit rights seems disingenuous – it suggests you only wish to remain a committer in name, but don’t actually intend fulfill the committer’s obligation to actively write and maintain code. It seems fairly transparent to me – the commits explorer clearly shows who is active, and the project charter describes the consequence of inactivity…

  9. John, I don’t disagree. I was inactive, that’s what I said before and the project character is very clear about this. The example of some simple refactorings looks disingenuous and is a bad one.

    However, whenever commit rights are granted there is at least a mailing list posting and public voting, whenever project permissions are granted there is a Bugzilla entry requesting these. Everything is open.

    Why wasn’t there any of these when the rights are revoked?

Leave a Reply