[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: [Fwd: bacula_1.36.3-1_i386.changes ACCEPTED]



On Thu, May 12, 2005 at 01:17:20PM +0200, José Luis Tallón wrote:
> Steve Langasek wrote:

> > On Thu, May 12, 2005 at 02:24:53AM +0200, José Luis Tallón wrote:

> >> This is to request the new version of Bacula to be included in
> >> Sarge. This version fixes several important bugs, most
> >> importantly an incompatibility with MySQL-4.1(now it is linked to
> >> libmsqlclient12 instead), but also a reproducible crash in
> >> "normal operational conditions"; I believe that this fix is very
> >> important for our users to be able to normally use this piece of
> >> software.

> > $ debdiff b/bacula/bacula_1.36.2-2.dsc
> > ../../../queue/accepted/bacula_1.36.3-1.dsc | wc -l 1351472 $

> > Even discounting that the majority of these changes appear to be to
> > documentation, leaving only about 8,000 lines of non-doc changes,
> > this is really too much to contemplate reviewing during the freeze.

> The only other option i have is to backport all of the changes to
> 1.36.2...

Well, no; the option would be to backport the *release-critical* fixes to
1.36.2, and upload just those fixes to testing-proposed-updates.  (Getting
rid of the libmysqlclient10 dep would be "pseudo-RC" for these purposes.)

> Apart from the fact that upstream applied some of my patches (and then
> i had to retouch them -- whitespace related fixes mostly), the rest of
> the changes are related to adding "fseeko" support and fixing some
> bugs since the release of 1.36.2
> 
> Yes, there are many changes in the package, due to the fact that the
> html documentation was removed and replaced by LaTeX sources... which
> need latex2html(non-free) to build. I have fixed it by removing the
> requirement on latex2html and generating just PDF documentation for
> now. This is *much* better than the old one.

> The rest of the differences...

> I have just reviewed the output from debdiff ... there are not that
> many changes from what i see...

They may not seem like "that many changes" to you, but they're way too much
for the release team to reasonably review.  The release team's time is a
finite resource, and it needs to be focused on addressing release-critical
issues.

-- 
Steve Langasek
postmodern programmer

Attachment: signature.asc
Description: Digital signature


Reply to: