[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Releasing 1.0.10, 1.1.5, 1.2.4 and 1.3.3

From: Erik Huelsmann <ehuels_at_gmail.com>
Date: 2007-04-06 00:19:02 CEST

On 4/6/07, Daniel Rall <dlr@collab.net> wrote:
> On Thu, 05 Apr 2007, Malcolm Rowe wrote:
> ...
> > We also still support the 1.3 line, according to the 1.4 release notes.
> For me, this is really key. We absolutely have to put out 1.4.x and
> 1.3.x patches, per our current support claims. Let's focus on doing a
> 1.4.x and 1.3.x FIRST.


> That said, Erik has backported the changes to the older release
> streams. I'm personally willing to build an environment and do the
> testing on each. I'd assume that 1.2 would be higher priority than
> 1.1, which would be higher priority than 1.0, but someone (Erik?)
> indicated that 1.1 might have a wider distribution at the moment.

Which is also true: There are still a lot of Debian Sarge and
Sarge-derivative installs around. Sarge has 1.1.4. But I talked to
Peter Samuelson and he'll be doing a release off the branch if we
don't have anything else available. So, your proposed 'newer first'
priority order seems fine now.

> After 1.4.x and 1.3.x are wrapped up, let's tackle the older releases
> in some sort of priority order -- if downstream package maintainers
> really want a patch, it's already available from our repository.
> > I have no problem with us publishing a patch for 1.[012], if it means
> > that we can avoid recreating a build environment. As Justin says, it
> > also probably helps avoid user confusion, and I suspect it's easier to
> > upload a patch+sig to CollabNet than all the zip/tar files and sigs.
> >
> > However, even if we release a patch, we still need to:
> > - backport to the 1.x line in question and produce a patch
> > - apply the patch to the 1.x tarball
> Erik has done this work.
> > - test
> > - get signatures for testing (signed against the patch)
> > - release note
> > - release
> This -- not insignificant -- work remains outstanding.

Exactly. Which made me wonder what the benefit of signing a patch was
over making a release: We can just apply the patch to the unpacked
tarball and repackage, as far as I'm concerned... (Not without the
above-mentioned testing, ofcourse!)



To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Apr 6 00:19:23 2007

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.