On Mon, 03 Jun 2013 10:31:10 +0300, Niko Tyni wrote: > Failed builds on mips, mipsel, s390, sparc: > > Test Summary Report > ------------------- > t/api/err_headers_out.t (Wstat: 0 Tests: 6 Failed: 3) > Failed tests: 2-3, 5
Ouch. Also reproducible on amd64 (as expected, with the newer libhttp-message-perl. > Looking at the test code in that file: > > # HTTP::Headers 6.00 makes the next 2 tests fail. When the response comes > # in the header name is stored as "x-err_headers_out". But when it is to > # be read below it is referred as "x-err-headers-out" and hence not found. > local $HTTP::Headers::TRANSLATE_UNDERSCORE= > $HTTP::Headers::TRANSLATE_UNDERSCORE; > undef $HTTP::Headers::TRANSLATE_UNDERSCORE > if defined HTTP::Headers->VERSION and HTTP::Headers->VERSION==6.00; > > libhttp-message-perl 6.03-1 has HTTP::Headers VERSION 6.00 > but the recently uploaded 6.06-1 has 6.05. > > Changing that test to read 'HTTP::Headers->VERSION>=6.00' fixes > the problem (although the equality comparison with a float makes me > slightly nervous.) Note that there are two instances of this code block > in the file. I've committed a patch to git that uses version->parse(). Does this sound right? > I haven't verified this, but I assume this also applies to 2.0.8 with > Apache 2.2 and upstream should be informed. I couldn't find a CPAN > RT ticket about this in the HTTP-Message queue. Right, I guess there are some details by now that we should tell upstream about :) I'd prefer if someone who is already susbcribed to (mod-)perl-dev could do this ... Cheers, gregor -- .''`. Homepage: http://info.comodo.priv.at/ - OpenPGP key 0xBB3A68018649AA06 : :' : Debian GNU/Linux user, admin, and developer - http://www.debian.org/ `. `' Member of VIBE!AT & SPI, fellow of the Free Software Foundation Europe `- NP: Andrew Lloyd Webber & Tim Rice
signature.asc
Description: Digital signature