[haiku-development] Re: Commit-access for Matt Madia

  • From: Colin Günther <coling@xxxxxx>
  • To: haiku-development@xxxxxxxxxxxxx
  • Date: Sat, 27 Feb 2010 03:21:49 +0100

Matt Madia schrieb:
To be honest: while it would be a convenience to have commit access --
mainly to reduce the workload of committers who review & apply my
smaller patches, I do see and agree with Ingo's preference for my
meatier jam patches.

As for the attribution aspect, with every patch everyone does make an
effort to credit the patch author in the commit message.

--mmadia

I know that once granted commit access, one can do everything with the repo. But as Haiku is about trust, too, I wanne add that I trust Matt that he will only touch parts he feels comfortable with. And this would be mainly scripting, which is where I based my proposal upon.

Also I can understand Ingos preference of reviewing Matts jam patches, I'd rather see Ingos improvement suggestions on the commit mailing list than - somewhat harder to take notice of - in trac.

So I'd like to know what is the difference between reviewing the patches in Trac in contrast to review them through the means of the commit list? I'm asking because I have full commit rights, feeling quite uncomfortable with jam and am glad that Matt made the rules working where I previously failed.

For me the commit list is a comfortable way of improving my understanding of the unwritten Haiku rules or to clarify the written ones by example. As I'm reading every commit to stay up to date, nonetheless, comments on this list are flowing naturally in my knowledge pool. But maybe I'm just missing another easily available stream of information to get this knowledge from trac, too. This is no offence to trac, also it may read as such.

I remember my first anxious steps in adding some Notify functions to the condition variable subsystem of Haiku's kernel. Ingo helped me a lot by encouraging me that it would be right to add those functions, as there didn't exist an equivalent to them. And of course I messed some things up, by forgetting to comment upon the correct usage of these new functions. But Ingo cleaned up behind me (r34460) and thus gave me a great example on how to do it correctly, which I added to my knowledge pool quickly :) This is what I would call "leading by example" in an efficient way.

To conclude this prosaic Mail: Once Matt feels equipped for commit access, it should be granted without the need for another proposal (or vote when i don't forget to add the prefix again).

Thanks for reading.
-Colin

Other related posts: