Re: Testing automake on Cygwin 1.7

[Peter Rosin]

Stefano Lattarini skrev 2011-11-26 00:11:
> On Friday 25 November 2011, Peter Rosin wrote:
>> That works, but isn't it an awful lot of code for a corner case?
>>
> Possibly, but I prefer to be conservative in tests, and not relax
> them unless necessary.  Also, since the new code is extensivley
> commented, a future developer will know why it's there, and thus
> also whether dropping it would be acceptable.
> 
>> You choice though.
>>
> I'd say we keep this.  I'll prepare a patch tomorrow.

Ok, cool.

>> I don't think we should, because it is never correct to add the
>> extra slash after $(DESTDIR)
>>
> Right.
> 
>> and adding a comment about it just
>> adds to the confusion by somehow stating that it would be
>> desirable to add that slash, but that we sadly can't. On the
>> contrary, it's just plain wrong with the slash, and you'd have
>> to add that comment all over the place in order to be consistent.
>> But again, your choice.
>>
> Let's just drop any hypotetical extra comment.  I'll prepare a patch
> in your name if you don't beat me at it.

Thanks.

>>>> So, with those two fixes, one fail left, i.e. transform2.test.
>>>>
>>> Could you let me know if my proposed change above make it correctly
>>> skipped?  If yes, I'll prepare three proper patches tomorrow -- unless
>>> you want to beat me at it ;-), in which case, please put a reference
>>> to this thread in the git commit message and in the ChangeLog entries.
>>
>> It would be better with an xfail in my opinion, if that's possible to
>> accomplish conditionally?
>>
> Sadly, not easily (we could extend XFAIL_TESTS at configure time if we
> detect the above Cygwin limitation, but that's quote involved, and IMO
> would put the logic in the wrong place, i.e., configure.ac).  Maybe in
> a next automake version we could add a new special exit status for test
> scripts to signal "expected failure", like e.g. `77' is used to signal
> "skipped test"?  Still, that's for automake 1.11.3 at most, so I see
> only two ways out for now:

A special exit status would only be another version of skip, with no
chance of unexpected success (if I read your proposal right). That's
not right either.

>   1. Add the test skip I proposed to branch-1.11 only, but *remove it*
>      after the 1.11.2 release.
>   2. Just let the test fail, and be prepared to deal with some spurious
>      reports.
> 
> I'm 60-40 in favor of 2, since it doesn't add yet more noise to our
> repository.  WDYT?

I'm also in favor of leaving it as a plain old fail. A skip is a bit too
silent, and it's not like the bug list is flooding with reports on the
subject...

Cheers,
Peter