[packagers] Re: [svn] r5000 - trunk/rpms/xine-lib-moles

[Dag Wieers]

On Tue, 19 Dec 2006, Wil Cooley wrote:

> On Mon, 2006-12-18 at 15:38 +0100, Dag Wieers wrote:
> > On Mon, 18 Dec 2006, Matthias Saou wrote:
> > 
> > > packagers at lists.rpmforge.net wrote :
> > > 
> > > > Author: thias
> > > > Date: 2006-12-18 13:50:35 +0100 (Mon, 18 Dec 2006)
> > > > New Revision: 5000
> > > 
> > > Revision 5000, wow! :-D
> > 
> > Damn, you beat me to it again !
> > 
> > Here's the score:
> 
> Interesting coincidence.  My local Linux user group had a talk last
> night by Randal Schwartz (the "Learning Perl" guy) on using Git; Linus
> was also there to heckle/correct him.  It got me thinking about one of
> the problem I have contributing to RPMforge--it's a real pain to
> contribute without Subversion write access.  It's difficult, then, to be
> a casual contributor--submitting patches through mail is a pain both for
> the sender and the gatekeeper, but without doing so for a long period of
> time, it's not possible to build up the trust necessary to gain write
> access.
> 
> Git (and Mercurial, it was mentioned) were both designed to handle this
> kind of stuff.  I'm sure you've all seen how kernel development works; a
> developer clones the source tree, makes his changes and then sends a
> message with "Linus (or whichever maintainer for whatever subsystem)
> please pull my tree from git://whatever", whereupon the tree is pulled,
> changes evaluated, and life moves on.  (And there's a special mechanism
> for automatically handling patches sent by Git through e-mail.)
> 
> I realize that version control systems may be the 4th biggest subject of
> flame wars (behind editors, distributions/operating systems and
> databases) and I have no intention of starting one.  Is switching to a
> VCS that handles distributed development more easily something you have
> or would consider?  I realize you're all busy and that learning a new
> tool, especially one as complex as Git, can be challenging and
> time-consuming; I may be naive to think that Git or Mercurial would be
> appropriate.

Hi Wil,

While I do agree with your assessement and the difference of working with 
subversion versus git, I do think we have more important things to tackle 
before changing the versioning system. :)

I've been saying this now for 2 years, but I think we may be making 
progress the coming months in having our own infrastructure and allowing 
people access.

Kind regards,
--   dag wieers,  dag at wieers.com,  http://dag.wieers.com/   --
[all I want is a warm bed and a kind word and unlimited power]