On Thursday 07 February 2002 21:59, Roman Neuhauser wrote:
> > Date: Thu, 7 Feb 2002 20:35:27 +0100
> > From: Mads Martin J?rgensen <[EMAIL PROTECTED]>
> > Lets keep this like the kernel. One guy gets all the patches, and
> > decides what goes in or not.
Decisions are not a reason for or against cvs, it's only the method that's
different.
> > If you write good code for the project, does it then matter if you
> > send the patches to him or put them directly in CVS anyway?
>
> yes, it does. I, as someone who has a potential patch might want to
> have access to the latest-and-greatest code. I don't have that
> today. I might want to submit improvements to the BB documentation.
> I can't do it today. It wouldn't clash with the code 90% of time,
> but I'd still have to bother Sean with any patches I might have:
> because there's no repository of the code.
Especially for maintaining patches, cvs is important. Updating a patch to new
code is just
$ cvs update
$ cvs diff
You simply can't do it that easily without cvs. This, and having access to
file history are the only two reasons that Sean might want to switch to
actively using cvs. Everything else he can do just as easily himself.
However, I can't see the reason not to use cvs. I don't think it imposes
major changes in the way people work.
Bo.
--
Bo Thorsen | Praestevejen 4
Free software developer | 5290 Marslev
SuSE Labs | Denmark
