Re: failing regression tests (please fix them!)

[Antti Kantee <pooka%cs.hut.fi@localhost>]

On Thu Jun 03 2010 at 06:06:14 +0000, David Holland wrote:
>  >     util/sh/t_set_e:all
> 
> This, however, remains in limbo because we have, collectively, been
> unable to reach a conclusion about what the correct behavior is.
> 
> However, just switching it off in response would be counterproductive.

A test pretty much by definition implies expected outcome.  If there is
no predefined outcome, a test is seriously broken and should not be run.

> Is there no way to persuade atf to report regressions/changes from the
> current behavior?

I'm not aware of any way to easily produce "quick-glance" diffs.  But I
agree it would be useful to have that feature bundled in with atf.