On Thu, Aug 08, 2013 at 11:01:27AM +0100, Nicholas Clark wrote: > On Tue, Aug 06, 2013 at 05:26:27PM +0100, Nicholas Clark wrote: > > On Tue, Aug 06, 2013 at 05:06:04PM +0300, Niko Tyni wrote: > > > > It's not perfect, but would it be enough to just skip this test if > > > /usr/bin/man isn't present? See the attached patch. > > > > I think that that's the best solution. We found that it was quite hard to > > actually write a reliable test, let alone a portable test. I think that it > > doesn't matter if the test is skipped frequently, as long as it still runs > > on some platforms. > > > > > Also, something that caught my eye: LC_MESSAGES is spelled wrongly > > > in perl5db.t. This is most probably just cosmetic, but attaching > > > a patch anyway. > > > > D'oh! > > > > It possibly does matter. :-) > > It's certainly worth fixing. And I suspect worth backporting, in case it > > matters. > > I've pushed both to smoke-me/nicholas/perl5db-man > I don't think that there are going to be any problems, but Tony and I thought > that at least twice before... Hi, this went into 5.18.1, but it's still not in blead. Is that intentional? (I can file a perlbug too if you like but it seems a bit late for that.) -- Niko Tyni ntyni@debian.orgThread Previous | Thread Next