Why commit access really isn't important (was: Re: [Xcb] Committing to xcb?)

[Matthias Hopf]

On May 02, 07 08:13:27 -0700, Carl Worth wrote:
> On Wed, 2 May 2007 14:58:01 +0200, Matthias Hopf wrote:
> > requesting commit permissions is at the utter core of the open source
> > development process, and *should* be answered (yes sure, no way, yes but
> > only in a branch, prove yourself worthy, we only accept patches, etc.).
> 
> As a total aside, allow me to totally disagree about the "utter core"
> aspect of commit permissions. The whole notion of commit permissions
> is vestige of centralized version-control systems.

Maybe I sounded wrong - it's not utter core aspect of giving somebody
commit access if he wants to, but at least for a project that apparently
has an "upstream" (in contrary to the kernel), there should be a policy
how commit requests are handled. I don't think there is one on the wiki
yet, except for that commit requests should be handled by Bugzilla (this
is still back from CVS days).

Note also that I certainly don't expect a 24/7 reaction ;)

If I sounded like *demanding* anything, I apologize, I just stated my
general belief in how things should be handled. YMMV, of course, your
opinion as well.

> With tools like git, it's really, really easy to ignore the whole
> question of commit permissions. So, if Jamey and Bart are being slow
> to respond here, you should just be happily creating commits. Then
> when they wake up, they'll be able to merge all the work you've done
> with no real effort.

Sure, I've already done that.
It's just that sometimes commit policies (and even sometimes the right
branch - oops, developed for an ancient version of the software) aren't
perfectly clear. I thought it was clear in this case, and was just
asking myself, whether I overlooked something.

> This is what Linus is talking about when he talks about the ease of a
> possible fork keeping the maintainer honest. See his very recent
> discourse on why "commit access" is not fundamental, (rebutting recent
> text that appeared in a book on open source development):

Yes, but I still consider the X.org project having a central repository.
At least, a de-facto central one. This is not about the tools, it's
about the politics around them. Well, partially, at least. With CVS you
didn't have much choice...

CU

Matthias

-- 
Matthias Hopf <mhopf at suse.de>      __        __   __
Maxfeldstr. 5 / 90409 Nuernberg   (_   | |  (_   |__          mat at mshopf.de
Phone +49-911-74053-715           __)  |_|  __)  |__  R & D   www.mshopf.de