[vlc-devel] [vlc-commits] hotkeys: handle viewpoint changes from the vout

[Rémi Denis-Courmont]

Le keskiviikkona 26. heinäkuuta 2017, 4.19.50 EEST Jean-Baptiste Kempf a 
écrit :
> > I don´t address respectfully people who don´t respect me and other devs
> > by  pushing crap.
> 
> This project *requires* respectfully addressing other developers.
> Whatever the reason. I'm quite sure we've been discussing that many
> times.

Pushing broken patches with trivial found and trivially fixable warnings is 
not respectful of somebody who´s been spending hours of his free time fixing 
warnings in the previous few days. Not according to my values anyway.

And what about trolling another developer on Trac and on the mailing list 
(referring to François here)?

> Everybody does mistakes, including you and me. I've been fixing commits
> from almost everyone breaking the build or the runtime regularly, and
> I've never complained. And that's been the case since forever...
> 
> Accusing Thomas of pushing crap because he adds warnings to the build,

This patch is bad. Use another word if you want. The fact is that if it had 
been *properly* checked, it wouldn´t have been merged as was.

> or saying that Thomas is not respectful to others,

What he did was not respectful of me, even if it was obviously not deliberate. 
You don´t have to agree with my moral values, but neither have I of yours or 
his.

Respect is a vague notion that means different things to different people. You 
can´t enforce that. In some cultures, even criticizing somebody´s code could 
be disrespectful, which directly clashes with open-source methodology.

> when he is probably
> the nicest active contributor with Tristan, is clearly out-of-line,
> sorry.
> 
> If you want us to move to mandatory reviews and mandatory build+test
> before push like the libav project, we can.

This is confusing subjects. The patch was signed-off, so presumably it was 
reviewed. Thus I don´t see how this concerns mandating reviews. Besides, the 
point of code review is not to replicate what the compiler and test suite can 
do better than humans.

The submission guidelines already state that warnings and test suites should 
be checked AFAIR. (And again, one can construe violating the submission 
guidelines is disrespectful to the project and its members.).

-- 
雷米‧德尼-库尔蒙
https://www.remlab.net/