gEDA-dev: libgeda library version & stable branch
Gabriel Paubert
paubert at iram.es
Sat Jun 2 21:29:55 EDT 2007
On Sun, Jun 03, 2007 at 01:44:48AM +0100, Peter Clifton wrote:
> On Sat, 2007-06-02 at 12:42 -0400, Ales Hvezda wrote:
> [snip]
>
> > * So, I am going to go ahead and setup a git repository as the official
> > repository of gEDA/gaf.
> >
> > * But, for those diehard CVS fans out there, I will run a git-cvsserver
> > which will does CVS as the frontend and git as the backend.
> > (http://www.kernel.org/pub/software/scm/git/docs/git-cvsserver.html)
>
> Peter B and myself will be happy to help anyone wanting to learn git, or
I could help to.
> with any questions. I'd actually recommend people use the layer on top
> of git called "cogito".
I used cogito in the beginning of git (2 years ago) but the
current git "porcelains" have improved to the point that
I no more find the need for another layer that compelling.
> Once you've used this to do off-line development
> with micro-commits, or some complex merge / diff / history enquiry,
> you'll appreciate how much more powerful it can be than CVS.
>
More powerful, way faster (sometimes you wonder whether your command
really has been executed) and much easier to learn in my opinion.
But I have always had a large impedance mismatch with CVS.
Really CVS is stone age. Now git may even have surpassed BK, it
certainly has in terms of performance and capabilities (rebasing
and local branches among them). The graphical tools are not yet
that great (gitk is ok for browsing) but they are improving.
Did I mention that git is _really_ fast, except for git-blame,
(but it does way more than CVS blame) and aggressive repacking
which is only needed occasionally to reduce repository size.
Well, I am a git fan, so take this with a grain of salt.
Regards,
Gabriel
More information about the geda-dev
mailing list