[haiku-development] Re: [haiku-development] Re: [haiku-development] Re: [haiku-development] Re: [haiku-development] Re: [haiku-development] Commit access for Andreas Färber

  • From: Stephan Assmus <superstippi@xxxxxx>
  • To: haiku-development@xxxxxxxxxxxxx
  • Date: Mon, 07 Jun 2010 17:42:24 +0200


On 2010-06-07 at 16:52:42 [+0200], Niels Reedijk <niels.reedijk@xxxxxxxxx> 
> On 7 June 2010 16:49, Fredrik Holmqvist <fredrik.holmqvist@xxxxxxxxx> wrote:
> > 2010/6/7 Stephan Assmus <superstippi@xxxxxx>:
> >> Haiku has always been also a ground for experimentation and learning. The
> >> quality of a random part of the codebase is always a function of the 
> >> skill
> >> and experience of whoever worked on it. If someone wants to work on a
> >> particular part of the code, especially with no one else wanting to work 
> >> on
> >> it, then why should we make it harder for him?
> >
> > If I were a dev that had to be able to fulfill our high qualities
> > before getting commit access I'd wonder if I'd be able to do it and if
> > it is worth the effort. Luckily Haiku is a very nice project to work
> > on. I'm not saying it should be lowered much, you still need to be
> > able to be a pretty decent developer and know your way around the
> > codebase.
> >
> > But maybe it's just the way of working with patches that limits
> > outside devs. Patches tend to get stale quickly, but if you could
> > manage your own tree and changesets better it would not be much of a
> > problem. Would Git or Mercury be better?
> That would not solve aging patches in any way. You would still need
> someone to manually update them when the original files change.
> Instead I hope the better patch tracking features on dev.haiku-os.org
> will increase the turnover speed of patches.

The better patch tracking on Trac is already very helpful -- when you want to 
take some time reviewing patches. As long as that is not happening more often 
and from more of the Haiku developers, I am afraid patches will still rot, 
which is a shame. That being said, much bigger projects (for example WebKit) 
have very similar problems. There have to be people willing to review and 
often a review has to be more than pointing out the coding style issues or 
getting hung up on silly details on the surface. I believe it's a problem of 
how much time people devote to Haiku, which is at least in part a problem of 
motivation and moral. More interesting stuff happening at once and more felt 
progress are the single biggest positive influences on moral (IMHO). Hence I 
believe we should give people commit access as soon as possible, since that 
will help the entire project in more than just one way.

Best regards,

Other related posts: