[FFmpeg-devel] [RFC] Development model

[Ivan Kalvachev]

On 3/17/11, Reimar Döffinger <Reimar.Doeffinger at gmx.de> wrote:
> On 17 Mar 2011, at 15:15, Michael Niedermayer <michaelni at gmx.at> wrote:
>> A quick question
>>
>> What do people prefer
>> A. Reviewer applies patch if he is sure its ok
>> B. Reviewer ok-es patch and author applies later
>
> B, it happens often enough that I know the code better than the reviewer and
> realise (obvious) fixes necessary or that I really should add those comments
> I was too lazy for.
> However that should not be taken as a prohibition for reviewers or other
> developers to apply if they are impatient for some reason.
> I don't see any real issue with review happening also on the -cvslog list,
> it's sometimes all I have time to read so that will happen anyway (as with
> the JV demuxer), so why hinder people my making up many rules?

I had my thoughts so stuck with masters mentality that for a moment I
thought it would be wrong.

I think we could say that if the author doesn't apply the patch in 1
day, then the reviewer can push it. If the author have found some
problems and want to work on them, then he should send warning.


I would like to rise another question.

I'd like to set policy that each patch should be signed. Every patch
sender who uses git can and should sign his own patches.

Here is the official meaning of the signoff.
http://www.mjmwired.net/kernel/Documentation/SubmittingPatches#297

It have mostly legal meaning (it was introduced after SCO claims) and
means that you are the original author, or that you do know the
original author.