[Spice-devel] To warn or not to warn

[Alon Levy]

> > That said, you both agree about keeping that warning and going out
> > of our
> > way to please xorg, so I'm fine with any patches you agree on ;)
> 
> Sure; two of us agree.  However, Søren has raised concerns, and has
> not yet
> further commented.  And Søren clearly deserves a substantial voice in
> what happens to this code.
> 
> But I feel that I have asked politely, and have waited a polite
> amount of time,
> so it's reasonable for me to agitate for resolution.
> 
> I get the sense that this happens with this project, and with Søren
> in
> particular; he seems to be particularly slammed for time, and that
> there
> seems to be a 'oh well, we waited long enough, let's just commit it
> anyway'
> attitude.
> 
> I do have commit access, so I can just commit.  But I'd appreciate
> some explicit guidance - is there some time period that is agreed to
> be reasonable to wait for further response?  Have I met that yet?
> (I've mentally assigned '1 month' as being said reasonable window,
> and it's been a month since I first raised this issue).
> 
> Do we just prefer to go ahead and commit and revert if and
> when we later disagree?  That approach seems like a recipe
> for disaster to me...

There was already one time when code was commited and reverted later because of misunderstanding on the way the code should be organized, last time it was coding standard (which to be honest Soren has made clear twice already that he follows pixman's coding style for xf86-video-qxl), so I agree it is a bad way to go forward. Perhaps you can address Soren's concerns one to one (I admit I don't remember them exactly without looking at the back log, but you'd need to address Soren directly to proceed without invoking the long-enough rule anyway).

> 
> Cheers,
> 
> Jeremy
>