[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Weird problem when testing for warnings
From: |
Thomas Weber |
Subject: |
Weird problem when testing for warnings |
Date: |
Sun, 16 May 2010 01:00:02 +0200 |
User-agent: |
Mutt/1.5.18 (2008-05-17) |
Hi,
I didn't file a bug report about this, as it's close to impossible to
reproduce it in any meaningfull way, so I'm asking here.
When building Debian package of octave-forge packages, the following
line in a test suite
%!warning <monotonically> warning('monotonically');
fails with the following error:
***** warning <monotonically> warning('monotonically');
!!!!! expected <monotonically> but got inverse: matrix singular to machine
precision, rcond = 3.50566e-20
But this only fails when building the package non-interactively; logging
into the chroot and issuing the commands works just fine. Also,
replacing 'warning' by 'error', ie
%!error <monotonically> error('monotonically');
works fine in both cases.
So, is there any significant difference between the code paths for error
and warning, that might be triggered in a non-interactive build
environment?
Octave used is 3.2.4.
Thanks
Thomas
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- Weird problem when testing for warnings,
Thomas Weber <=