Re: proposed patch: parallel-tests: also record logs of SKIPped tests.

[Ralf Wildenhues]

Hi Jim,

* Jim Meyering wrote on Sat, Mar 14, 2009 at 12:07:15PM CET:
> Ralf Wildenhues wrote:
> > This is a patch that I don't know whether I want it or not.
> > Package authors may be interested in SKIPs and why they happen, or not.
> > (Could also be solved with leaving the choice to the package author
> > and/or user, by some TEST_RECORD_SKIPS variable or so).
> >
> > Comments appreciated.
> 
> Yes, please.
> 
> Once I taught coreutils' "make check" to tell why a test is skipped,
> I discovered a few that should not have been skipped.
> Seeing a summary (and remembering the numbers) might help too.

Yes, I agree.  The thing is, though, that there may be an abundance of
SKIPs; all of which are hopefully not very interesting.  For example,
the Automake testsuite itself can easily skip a good part of its tests
when on an unusual host (that in itself is probably a bad thing, but
that's for another thread ...).

The question is whether there should be a toggle to enable or disable
logging SKIPs, or they should always be logged, and if a toggle, what
color that bikeshed^Wtoggle should have.

I take that back.  User API questions are never really bike shedding.

Thanks,
Ralf

> >     parallel-tests: also record logs of SKIPped tests.
> >
> >     * lib/am/check.am [PARALLEL_TESTS] ($(TEST_SUITE_LOG)): Record
> >     SKIPs as well.