| Re: [AD] Allegro 4.2.0 RC2 |
[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
On 2005-08-30, Peter Wang <tjaden@xxxxxxxxxx> wrote:
> I mean, I don't want to redownload Matthew's files from a.cc and
> reupload them to SF.net.
Ah, ok, linked.
On 2005-08-30, Peter Wang <tjaden@xxxxxxxxxx> wrote:
> I think CVS is the cause but my conclusion is different.
> Making releases has always been tedious.
In a similar way, it's the reason I'm pretty much responsible for
the web page mechanism. It's cool that it automirrors with different
policies for binaries, but it scares everybody.
If releasing things is that a burden, maybe we should find a way
to ease that process? I've already thought of that for the web,
so it's a matter of finding some time/motivation now to fix it.
> > I'm more interested to know any plans about 4.0.x maintenance
> > plans. There are people concerned it doesn't compile with latest
> > versions of gcc or some other compiler. And after all, that's
> > what x.x.1 increments are supposed to be for, fix minor issues.
>
> CVS does not make backporting changes easy.
I carefully tried to avoid the backport word. Just applying the
vital patches to make it compile out of the box would be enough. The
point is, new users will be scared of using something labelled
unstable branch. And now with 4.2.0 near, are we going to maintain
two stable branches and the WIP one?
> If there is a nice tool for it please point it out (I haven't
> looked yet).
cvsps makes it easier to group "patchsets". But even svn lacks merge
tracking, so it relies heavily on policy/management layers like trac.