[Duplicity-talk] Adding test suite to distribution

[Ben Escoto]

On Tue, Jan 14, 2003 at 12:48:47AM -0600, Rob Browning wrote:
> > ).  Right now including the tests and the sample data they use would
> > make the tarball several times bigger, so they aren't included.
> 
> Hmm.  Well, is there any way we could generate the test data during
> "make check", i.e. via dd and /dev/urandom or just copy some of the
> source tree data or similar?  That might help reduce the size.
        ...
> Of course I'd still also be favorably inclined toward something like a
> runtime "duplicity --test" option.  Aside from allowing users to test
> the tool just before a backup run, I could also run --test in the
> duplicity post-install.  That wouldn't catch a case where a new python
> install breaks an existing duplicity install, but it would help catch
> any system anomolies at the time a new duplicity package is installed.

Yes, I agree.  I came to this conclusion a while ago also
(unfortunately it was still after most of the tests had been written)
so the newer tests just contain code to create the files they work on
instead of a separate tarball.  Another advantage is that when a test
is revised a text diff to the relevant code is much more informative
than a binary diff to the tarball.

So 'duplicity --test' sounds good, but I don't think I will get to
this immediately.


-- 
Ben Escoto

pgp4SuO01wg32.pgp
Description: PGP signature