Bug#649173: pu: package nss-pam-ldapd/0.7.16
On Wed, 2012-01-18 at 22:26 +0100, Arthur de Jong wrote:
> On Wed, 2012-01-18 at 13:42 +0000, Adam D. Barratt wrote:
> > If the two further changes included in the debdiff from
> > <1324994307.2556.15.camel@sorbet.thuis.net> are resovled in unstable (if
> > appropriate) then I'd be happy to look at including those in a stable
> > update.
[...]
> To be clear, there are three changes up for consideration:
>
> (1) is #645599 which is essentially:
> http://arthurdejong.org/viewvc/nss-pam-ldapd/nss-pam-ldapd-0.7/debian/libnss-ldapd.config?r1=1209&r2=1555
> (2) is a problem with not-initialised variables which is:
> http://arthurdejong.org/viewvc/nss-pam-ldapd/nss-pam-ldapd-0.7/nslcd/nslcd.c?r1=1209&r2=1559
> (3) is a typo in error handling code:
> http://arthurdejong.org/viewvc/nss-pam-ldapd/nss-pam-ldapd-0.7/nslcd/myldap.c?r1=1469&r2=1561
Those are the changes I was referring to above when I said I'd be happy
to look at them, yes.
[...]
> I also have a few more fixes that I would also like to fix in a 0.7
> version. One of which is the range checking stuff in my second mail to
> the bug report and a few more that are contributed by Jakub Hrozek of
> Redhat (those patches are used at Redhat). Some of these changes are
> more intrusive. Not all of these are in a released 0.8 version yet.
>
> Which of the above fixes are OK for a stable update?
Well, any which are not "in a released 0.8 version yet" certainly
aren't. The range checking patch could be okay, although given that
we're getting quite close to the end of the window for 6.0.4 I'm tempted
to say we'll look at it for a future update. Are any of the other
changes you mentioned which are fixed in 0.8 and aren't particularly
intrusive easily reviewable?
> This leaves the question of what version number to use. I could upload a
> 0.7.16 which would also include updates to the version number in the
> code and manual pages (like in the previous debdiffs) or a
> 0.7.15+squeeze1 version (which would only also change debian/changelog).
As I mentioned, the latter would be preferable as it makes the diff much
cleaner, targetted and easier to review. In retrospect, we possibly
should have raised that for the earlier updates.
Regards,
Adam
Reply to: