[FFmpeg-devel] [VOTE] Equality and leader team

[Nicolas George]

L'octidi 18 pluvi?se, an CCXIX, Arpi a ?crit?:
> if they commit to code they know very well, they dont need to send
> patches to list first, i think.if they are unusure about change,
> or they have patches for code they dont know in deep then they
> will send patch first, even if no rules enforcing them to do so.
> 
> this is why we trust these people enough for giving them write access.
> at least i did so in mplayer cvs, and it worked.
> 
> other developers/maintainers still can review tehse commits and
> revert if they donlt agree, but i dont think it will happen often.

Even if you are the one who knows this part of the code best by a wide
margin, sending a patch before applying is IMHO a good thing because:

- It let knows the others that you are working on something.

- Even people who do not know the code very well can make useful
  suggestions ("Why don't you use <some macro> for <this complex formula>?")
  or spot bugs.

- If someone is working on something else in the same area, it is an
  occasion to discuss your efforts and share the work.

On the other hand, the policy requiring patches and review should be supple
enough to allow something like that:

	"I guess I'm the only one who understand that part of the code, so
	I'll approve myself and commit in a few days if there are no
	remarks."

And with Git, delaying the commit a little is much less a hassle than with
SVN.

Regards,

-- 
  Nicolas George
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20110206/42508c14/attachment.pgp>