Quoting Walter Franzini ([EMAIL PROTECTED]):

> > The package properly builds on my system. However, it fails to build
> > in pbuilder (and will therefore fail to build on autobuilders).
> 
> Then it's a different (and known) problem.
> 
> I'm not sure what autobuilders are, however on
> http://buildd.debian.org/pkg.cgi?pkg=aegis I can see some succesfull
> build (ia64 and m68k).  Is it the right page?

Yes, that's it.

> 
> If I'm not wrong pbuilder need to be run as root, because it need to
> populate /dev/.
> 
> On the other hand Aegis is not supposed to be invoked by root.
> 
> > So, trying to reproduce the failure means that you *have* to build the
> > package with pbuilder.
> 
> I've looked in the past at fixing the problem of running the testsuite
> under pbuilder.


Well, what I wonder is the point of running the testsuite when
building the package. I understand that, as upstream author, you
introduce a testsuite....this is even a very good practice.

However, I don't really see any point in running that testsuite during
builds. Moreover, from my tests, this is a very time-consuming (and
verbose) process.

Please note that I'm not the maintainer of this package. I originally
went on it to fix some Debian-specific localization issues.... So I
was considering a "non-maintainer upload", which of course involves
building the package, which I always do with pbuilder.

Then I went on this issue...but I actually know nothing about aegis,
specifically.

> *If* there is interest I can look at it again, however I may need help
> From someone with a good knowledge of pbuilder and/or fakeroot.


I think it would be better to not run the testsuite.

The maintainer scripts seem to build with "make sure", which I guess
is responsible for the testsuite being used.

Can you confirm this?


Attachment: signature.asc
Description: Digital signature

Reply via email to