On 4/5/07, Justin Erenkrantz <justin@erenkrantz.com> wrote:
> On 4/4/07, Erik Huelsmann <ehuels@gmail.com> wrote:
> > Ok, so I've done all the backporting for 1.0 .. 1.3 for the issue 2751
> > fix (rare FS corruption) because there are even occasional reports on
> > users@ which relate to 1.0. I want all distros to be able to do a new
> > release if and when they still do maintenance for anything which
> > includes a post 1.0 release of Subversion.
>
> I think us doing releases for each one of those series is massive
> overkill. We can post a patch for each series and if a distributor is
> still using them, then can repackage accordingly - but expecting us to
> go through, re-create each release environment, and vote and test on
> each release is, IMHO, insane. I think the only one we really need to
> cut a release for is the current version (i.e. 1.4.x) and let
> downstream repackagers figure out if they want to cut their own
> release from older stuff. -- justin
If this is the way we distribute updates after a .0.0 release, then
why did we ever bother to do that? First of all, we bother, because we
have always *strongly* discouraged packagers from making other
releases than the officially released ones.
BTW: Why can't we just unpack the tarball, apply the patch and call it
1.x.<last+1>? I mean, it's just what you are proposing, but the
user-friendly way...
bye,
Erik.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Apr 5 08:56:33 2007