[FFmpeg-devel] (trying to be) a voice of reason
Mon Mar 14 22:42:00 CET 2011
On 14 Mar 2011, at 19:51, Felipe Contreras <felipe.contreras at gmail.com> wrote:
> 2) Everyone that had svn commit access gets git commit access
> 2.1) The first time somebody uses their commit access to push a patch
> without proper review gets their rights revoked
As I said, I tried to contribute under those rules, but it required a level of effort that means I'll be unlikely to contribute if it is the only way.
If timely and through (as in, read and tried to understand the code) were even likely it would work, but I do not see it happen.
Now, if someone provided a tree where I can still commit after a few days and someone else reliably takes care of getting it in I wouldn't care much.
However I don't really see why review can't happen on the commit mailing list, I often don't have time to follow every patch discussion, so I will sometimes only review patches after they have been committed, the number of mails on the cvslog list is far more manageable.
TL;DR: If only one or two things were changed to specifically address any issue I wouldn't have had many objections besides style, but a massive change of development model that does not work for me is an issue.
So as MPlayer developer I will be using Michael's tree since he is taking care of keeping it in sync with others, and I can still do development within the time constraints I have. Without it I'd probably have to make my own FFmpeg fork, and that's both silly and more effort than I'd like (and will really suck for distros).
More information about the ffmpeg-devel