Josh Boyer wrote:
On Thu, 08 Nov 2007 14:08:35 -0500
Simo Sorce <ssorce@xxxxxxxxxx> wrote:
On Thu, 2007-11-08 at 19:19 +0100, Ralf Corsepius wrote:
On Thu, 2007-11-08 at 18:19 +0100, Matej Cepl wrote:
On 2007-11-08, 14:33 GMT, Lubomir Kundrak wrote:
any plans to replace it by git, mercurial, svn or other
more modern version control system?
- CVS server outage :)
Other services never have outages?
They don't -- have you ever heard about distributed VC? If the
repository I am working against has outage, then I don't care.
And how do you pull/push if the remote server is down?
You commit locally and keep working.
When the repo is up you push down everything in one go.
If you need patches from others they can share their repo or send you
mails that keep the whole meta-data.
DVCS really have *no* problems with outages. Only centralized VCS has.
It does. Our buildsys requires a single repository to pull (or
checkout) from to do builds. So until your changes are present in
there, you still suffer from the "outage" problem if you want to
actually build packages for others to consume in the official repos.
This, as Red Hat's GIT-RE team is learning, is because the notion of a
distributed build system hasn't really been spec'd out. I have some
notion of how to do it, and I've almost figured out a workflow with
buildbot.
josh
--
fedora-devel-list mailing list
fedora-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/fedora-devel-list
