Re: [VOTE] Release of SpamAssassin 3.1.10

[Theo Van Dinter <fe...@apache.org>]

Wow.  This has sat around for ages! :(

It doesn't look like anything got committed to fix these issues, but we should
get that done and get 3.1.10 out.  Sidney, do you still have the code you were
working on for this?  I didn't see any open bugs w/ milestone of 3.1.10, and
3.1.11 only has a single rule-related FP bug.

JM: I think the issue is less "cpan" and more "root".


Ok, so since the tarballs weren't officially released anywhere, I'm going to
remove the 3.1.10 tag, and reset the branch to dev mode.


On Tue, Jul 17, 2007 at 10:36:18AM +0100, Justin Mason wrote:
> 
> Daryl C. W. O'Shea writes:
> > Sidney Markowitz wrote:
> > > Matt Kettler wrote, On 17/7/07 2:40 PM:
> > >> I guess since 3.1.9 is already broken, and this is actually less broken,
> > >> and it's a security fix,  I'm +1 on this release.
> > >>
> > >> However, I make that +1 with the strong recomendation that we need to
> > >> get a 3.1.11 out sometime soon that actually builds properly under CPAN.
> > > 
> > > No, 3.1.10 is not a security fix -- That was a copy and paste error by
> > > Theo in the draft announcement. The only bug fixes between 3.1.9 and
> > > 3.1.10 are the two minor ones that Theo listed. I think they were fixes
> > > that Theo wanted in 3.1.9 but were missed.
> > > 
> > > I would vote for committing the bug 5510 and 5529 fixes to the 3.1
> > > branch if anyone else wants them in there, and I'm willing to do the
> > > work of tracking down the details to make sure that we take care of
> > > anything else that has to be committed along with those patches. It is
> > > annoying to not be able to install from CPAN, especially if 3.1.9 was
> > > the release that broke it.
> > 
> > Please do so Sidney.  CPAN installs need to work as root.  If the fixes 
> > aren't backported the tests should at least be disabled.
> 
> I agree they need to be backported too.   One thing though: shouldn't
> users of "sudo cpan Mail::SpamAssassin" only get SpamAssassin 3.2.x,
> not 3.1.x, since 3.2.x is the highest release number on CPAN?
> 
> --j.

-- 
Randomly Selected Tagline:
"A cat spends her life conflicted between a deep, passionate and profound
 desire for fish and an equally deep, passionate and profound desire to
 avoid getting wet.  This is the defining metaphor of my life right now."
         - Cat Okita in <20...@iguana.reptiles.org>

Re: [VOTE] Release of SpamAssassin 3.1.10

[Sidney Markowitz <si...@sidney.com>]

Theo Van Dinter wrote, On 17/12/07 9:38 AM:
> Wow.  This has sat around for ages! :(
> 
> It doesn't look like anything got committed to fix these issues, but we should
> get that done and get 3.1.10 out.  Sidney, do you still have the code you were
> working on for this?  I didn't see any open bugs w/ milestone of 3.1.10, and
> 3.1.11 only has a single rule-related FP bug.

The two bugs mentioned here, 5510 and 5529 were committed to 3.1 branch
and closed last July 23. With those commits there should be no problem
running make test as root. Since then 3.10 has had no open bugs and was
ready to go as soon as anybody wanted to cut a release. Someone created
a 3.1.11 target, but the only open bug on it is a rule change that 1) I
don't think is required; and 2) can be handled in sa-update if it is
required.

So as far as I know we have been ready for a 3.1.10 release for almost 5
months now.


 -- sidney