Re: [gentoo-dev] Re: [gentoo-commits] gentoo-x86 commit in media-libs/mlt: ChangeLog mlt-0.5.4-r1.ebuild

[Thilo Bangert]

Richard Freeman <rich0@g.o> said:
> On 08/14/2010 10:29 AM, Markos Chandras wrote:
> > So do I. Fixing your package and you don't even bother to send a
> > *ready to go* patch upstream seems like a bit rude to me as well.
> > Perhaps, we do have a complete different point of view in this one.
> > Recent example is Chí-Thanh Christopher Nguyễn who thanked me for
> > fixing his package, asked me to attach the patch so *he* can send it
> > upstream. I thought that was the *default* policy. Anyway. I should
> > talk to each maintainer separately when I fix his package. Seems to
> > me is the best approach
>
> My two cents.  In my opinion, whether a commit is good or not depends
> on whether it left Gentoo as a whole in better or worse shape than
> before it was made.
>
> Here it sounds like we had QA problems before the commit, and no QA
> problems after the commit.  Maybe the maintainer has some work to do
> now, but he had it to do anyway, and the maintainers have less work to
> do now than they did before the patches were made.
>
> Now, if he had broken something due to a sloppy commit I'd be more
> concerned.
>
> Many hands make for lighter work.  The best way to have many hands is
> to make individual tasks easier.  1+1+1+1+1 is going to happen faster
> than 3+2, since nobody ever gets around to doing 3.
>
> If we give devs an ultimatum like "fix it all or don't fix anything"
> guess which one they'll pick?

exactly. maybe the maintainer has to do some catch up work, but thats ok. 
the aim is to improve the tree and not for QA to do the work of the 
maintainer.

perhaps there is a lesson here though: if the bug isnt closed as soon as 
the patch has hit the tree, but its subject changed to 'push QA patch 
upstream', then it is clear what is left to do.

>
> Rich