Locales

[Benjamin Kosnik]

The changed behavior between 3.3 and 3.4+ is intentional. The XPASSES on
the old branches are annoying.

>Therefore, the current behavior is definitely not accidental.

Yep. This was compromise behavior. I think there was previous commentary
on this on the mailing list.

>That said, I agree that the present situation is not optimal, but I'm also
>sure that we do *not* want hundreds of XPASSes!

Yep.

>In my opinion, very-very roughly, testcases involving not available named
>locales should not be run at all. This would require a preliminary check
>before running the testsuite. But, if we decide to go this way, we should
>make sure is obvious from the results *why* some tests are not run...

Well what we do now is check at configure time if certain functionality

threads
wchar_t

(to be specific) is going to be used, then prune the testlist
accordingly. Perhaps this is the better solution.

-benjamin