found 378835 0.9-8
thanks
* Vincent Danjean [Thu, 20 Jul 2006 08:41:25 +0200]:
> Adeodato Simó wrote:
> >> * do not use default python but python2.4 instead for hg :
> >> most code is python2.3 compatible (so we still compile .py files for
> >> pyhton2.3) but "hg help" need features for python2.4
> >> Closes: #378835
> > (This got fixed everywhere except on i386. I've asked for a rebuild.)
> I ckecked the build log on anibal's autobuilder. I cannot find anything
> that shows this problem. But you are right, the current package in
> unstable use python2.3 and not python2.4 (even if it is built for
> python2.4 after python2.3).
> The build was working on my laptop, so I do not know what happened.
The scheduled binNMU produced faulty packages as well. After spending a
fair while digging, I'm pretty confident it is some kind of race
condition. If you compare the i386 log to, say, amd64's, the only
relevant difference is that the i386 one lacks the following line in the
python2.4/install phase:
copying build/scripts-2.4/hg ->
/build/buildd/mercurial-0.9/debian/mercurial/usr/bin
If you inspect distutils, you see that copying files happens by default
in "update or non-force mode", which checks the timestamp of the target
file, and does not make the copy if it's newer than the source file.
Both Anibal's system and the i386 buildd suffered from the race
condition, and both your laptop and my sistem did not (I build it
successfully myself before asking for the binNMU).
> If the rebuild does not fix it, I will add a 'sed' line in the
> debian/rule to force the good interpreter.
This is one possible solution.
Cheers,
--
Adeodato Simó dato at net.com.org.es
Debian Developer adeodato at debian.org
Don't worry about what anybody else is going to do. The best way to
predict the future is to invent it.
-- Alan Kay
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
