Skip Menu |

This queue is for tickets about the ExtUtils-MakeMaker CPAN distribution.

Report information
The Basics
Id: 29955
Status: resolved
Worked: 4 hours (240 min)
Priority: 0/
Queue: ExtUtils-MakeMaker

People
Owner: Nobody in particular
Requestors: dolmen [...] cpan.org
kahirsch [...] cpan.org
Cc:
AdminCc:

Bug Information
Severity: Normal
Broken in: (no value)
Fixed in: (no value)



Subject: make test (of ExtUtils::MakeMaker) fails on HP-UX
If you use the HP-UX 'make' command, then "make test" for the module ExtUtils::MakeMaker usually fails at least once with this error: # Makefile out-of-date with respect to Makefile.PL This is because (1) Makefile.PL files are built dynamically (2) the Makefile is created within the same second as Makefile.PL (3) the make command on HP-UX conservatively assumes that files created at exactly the same time are out-of-date w.r.t. to each other. This is rarely a problem using MakeMaker for other modules since Makefile.PL is usually not created dynamically. Workaround: Add "sleep 1;" before "return 1;" in each of the setup / setup_recurs functions in t/lib/MakeMaker/Test/Setup/*.pm modules. The disadvantage is that "make test" takes a little longer. When I ran "make test", the setup()/setup_recurs() functions we called 17 times, so the test ran about 17 seconds longer. I can send patch if desired. I know that Gnu Make and NMAKE don't have this problem. They make the opposite assumption about two files with the same timestamp.
Here is a patch which adds the "sleep 1;" to the subroutines as described in my bug. Please apply if acceptable.
--- ExtUtils-MakeMaker-6.36/t/lib/MakeMaker/Test/Setup/BFD.pm 2006-09-02 15:14:32.000000000 -0400 +++ ExtUtils-MakeMaker-NEW/t/lib/MakeMaker/Test/Setup/BFD.pm 2007-10-12 15:00:57.000000000 -0400 @@ -110,6 +110,8 @@ close FILE; } + sleep 1; # force Makefile.PL to be older than Makefile + return 1; } --- ExtUtils-MakeMaker-6.36/t/lib/MakeMaker/Test/Setup/PL_FILES.pm 2006-09-02 15:14:32.000000000 -0400 +++ ExtUtils-MakeMaker-NEW/t/lib/MakeMaker/Test/Setup/PL_FILES.pm 2007-10-12 15:01:06.000000000 -0400 @@ -102,6 +102,8 @@ close FILE; } + sleep 1; # force Makefile.PL to be older than Makefile + return 1; } --- ExtUtils-MakeMaker-6.36/t/lib/MakeMaker/Test/Setup/Problem.pm 2006-09-02 15:14:32.000000000 -0400 +++ ExtUtils-MakeMaker-NEW/t/lib/MakeMaker/Test/Setup/Problem.pm 2007-10-12 15:01:09.000000000 -0400 @@ -39,6 +39,8 @@ close FILE; } + sleep 1; # force Makefile.PL to be older than Makefile + return 1; } --- ExtUtils-MakeMaker-6.36/t/lib/MakeMaker/Test/Setup/Recurs.pm 2007-06-29 02:37:47.000000000 -0400 +++ ExtUtils-MakeMaker-NEW/t/lib/MakeMaker/Test/Setup/Recurs.pm 2007-10-12 15:01:13.000000000 -0400 @@ -52,6 +52,8 @@ close FILE; } + sleep 1; # force Makefile.PL to be older than Makefile + return 1; }
Subject: Re: [rt.cpan.org #29955] make test (of ExtUtils::MakeMaker) fails on HP-UX
Date: Tue, 16 Oct 2007 09:48:25 -0400
To: bug-ExtUtils-MakeMaker [...] rt.cpan.org
From: Michael G Schwern <schwern [...] pobox.com>
Ken Hirsch via RT wrote: Show quoted text
> This is because > (1) Makefile.PL files are built dynamically > (2) the Makefile is created within the same second as Makefile.PL > (3) the make command on HP-UX conservatively assumes that files created > at exactly the same time are out-of-date w.r.t. to each other. > > This is rarely a problem using MakeMaker for other modules since > Makefile.PL is usually not created dynamically. > > Workaround: > Add "sleep 1;" before "return 1;" in each of the setup / setup_recurs > functions in t/lib/MakeMaker/Test/Setup/*.pm modules. > > The disadvantage is that "make test" takes a little longer. When I ran > "make test", the setup()/setup_recurs() functions we called 17 times, so > the test ran about 17 seconds longer.
As an alternative, what about setting the mod time of the Makefile.PL to be one second in the past? -- Hating the web since 1994.
From: KAHIRSCH [...] cpan.org
On Tue Oct 16 09:49:04 2007, schwern@pobox.com wrote: Show quoted text
> Ken Hirsch via RT wrote:
> > Workaround: > > Add "sleep 1;" before "return 1;" in each of the setup / setup_recurs > > functions in t/lib/MakeMaker/Test/Setup/*.pm modules.
> > As an alternative, what about setting the mod time of the Makefile.PL
to be Show quoted text
> one second in the past? > >
That would be faster, but I'd worry that it might actually break things in certain cases. I put in utime time(), time() -1, $file; after the 'close' in each setup or setup_recurs loop. That worked fine. I didn't do any of these things which one might want: (1) only adjust time if $^O eq 'hpux' (2) only adjust time if $file =~ /Makefile.PL/ (3) adjust time relative to calibrate_mtime() or (stat($file))[9] If you want, I can prepare a patch.
Subject: Re: [rt.cpan.org #29955] make test (of ExtUtils::MakeMaker) fails on HP-UX
Date: Thu, 18 Oct 2007 14:19:06 -0400
To: bug-ExtUtils-MakeMaker [...] rt.cpan.org
From: Michael G Schwern <schwern [...] pobox.com>
Ken Hirsch via RT wrote: Show quoted text
>> As an alternative, what about setting the mod time of the Makefile.PL
> to be
>> one second in the past? >> >>
> > That would be faster, but I'd worry that it might actually break things > in certain cases.
Can you think of those certain cases? Show quoted text
> I put in > utime time(), time() -1, $file; > after the 'close' in each setup or setup_recurs loop. That worked fine. > I didn't do any of these things which one might want: > (1) only adjust time if $^O eq 'hpux' > (2) only adjust time if $file =~ /Makefile.PL/ > (3) adjust time relative to calibrate_mtime() or (stat($file))[9] > > If you want, I can prepare a patch.
Yes, a patch would be lovely, thank you. Umm, don't worry about #1 and #2, this should be generally applicable and portable. Please do #3. It should be a matter of changing the utime statement to: # take just one time reading so it's atomic, avoid one read at 1.99999 # and the next at 2.0000 my $time = calibrate_mtime(); utime $time, $time - 1, $file; -- I have a date with some giant cartoon robots and booze.
Subject: Re: [rt.cpan.org #29955] make test (of ExtUtils::MakeMaker) fails on HP-UX
Date: Wed, 14 Nov 2007 10:58:13 -0500
To: bug-ExtUtils-MakeMaker [...] rt.cpan.org
From: Ken Hirsch <kenhirsch [...] ftml.net>
Michael G Schwern via RT wrote: Show quoted text
> <URL: http://rt.cpan.org/Ticket/Display.html?id=29955 > > > Ken Hirsch via RT wrote: >
>>> As an alternative, what about setting the mod time of the Makefile.PL >>>
>> to be >>
>>> one second in the past? >>> >>>
>> That would be faster, but I'd worry that it might actually break things >> in certain cases. >>
> > Can you think of those certain cases? > >
Just those where utime is "broken". Hopefully calibrate_mtime will take care of those. Sorry about the delay. Here's the patch. I ran it so in order to test it: gunzip <ExtUtils-MakeMaker-6.36.tar.gz | tar xof - cd ExtUtils-MakeMaker-6.36 patch -p1 <../patch perl Makefile.PL make test .... All tests successful, 7 tests and 31 subtests skipped. Files=44, Tests=589, 52 wallclock secs (27.33 cusr + 11.22 csys = 38.55 CPU)
--- ExtUtils-MakeMaker-6.36/t/lib/MakeMaker/Test/Setup/BFD.pm 2006-09-02 15:14:32.000000000 -0400 +++ ExtUtils-MakeMaker-NEW/t/lib/MakeMaker/Test/Setup/BFD.pm 2007-10-23 15:14:04.000000000 -0400 @@ -108,6 +108,9 @@ open(FILE, ">$file") || die "Can't create $file: $!"; print FILE $text; close FILE; + + my $time = calibrate_mtime(); + utime $time, $time - 1, $file; # ensure file at least 1 second old } return 1; --- ExtUtils-MakeMaker-6.36/t/lib/MakeMaker/Test/Setup/PL_FILES.pm 2006-09-02 15:14:32.000000000 -0400 +++ ExtUtils-MakeMaker-NEW/t/lib/MakeMaker/Test/Setup/PL_FILES.pm 2007-10-23 15:14:04.000000000 -0400 @@ -100,6 +100,9 @@ open(FILE, ">$file") || die "Can't create $file: $!"; print FILE $text; close FILE; + + my $time = calibrate_mtime(); + utime $time, $time - 1, $file; # ensure file at least 1 second old } return 1; --- ExtUtils-MakeMaker-6.36/t/lib/MakeMaker/Test/Setup/Problem.pm 2006-09-02 15:14:32.000000000 -0400 +++ ExtUtils-MakeMaker-NEW/t/lib/MakeMaker/Test/Setup/Problem.pm 2007-11-14 10:20:54.000000000 -0500 @@ -7,6 +7,7 @@ use strict; use File::Path; use File::Basename; +use MakeMaker::Test::Utils; my %Files = ( 'Problem-Module/Makefile.PL' => <<'END', @@ -37,6 +38,9 @@ open(FILE, ">$file") || die "Can't create $file: $!"; print FILE $text; close FILE; + + my $time = calibrate_mtime(); + utime $time, $time - 1, $file; # ensure file at least 1 second old } return 1; --- ExtUtils-MakeMaker-6.36/t/lib/MakeMaker/Test/Setup/Recurs.pm 2007-06-29 02:37:47.000000000 -0400 +++ ExtUtils-MakeMaker-NEW/t/lib/MakeMaker/Test/Setup/Recurs.pm 2007-10-23 15:14:04.000000000 -0400 @@ -50,6 +50,9 @@ open(FILE, ">$file") || die "Can't create $file: $!"; print FILE $text; close FILE; + + my $time = calibrate_mtime(); + utime $time, $time - 1, $file; # ensure file at least 1 second old } return 1;
This is still broken with EEMM 6.50 on HP-UX B.11.23 either with /usr/bin/make and /usr/local/bin/gmake (GNU Make 3.80). Has the patch been applied?
Here is my log: Show quoted text
cpan> o conf makepl_arg
makepl_arg INSTALL_BASE=/tmp/omengue/home/CPAN INSTALLARCHLIB=/tmp/omengue/home/CPAN/lib/perl5/IA64.ARCHREV_0-thread-multi-LP64 INSTALLPRIVLIB=/tmp/omengue/home/CPAN/lib/perl5 INSTALLBIN=/tmp/omengue/home/CPAN/bin INSTALLSCRIPT=/tmp/omengue/home/CPAN/bin INSTALLMAN1DIR=/tmp/omengue/home/CPAN/man/man1 INSTALLMAN3DIR=/tmp/omengue/home/CPAN/man/man3 INSTALLHTMLDIR=/tmp/omengue/home/CPAN/html Show quoted text
cpan> install ExtUtils::MakeMaker
[...] CPAN.pm: Going to build M/MS/MSCHWERN/ExtUtils-MakeMaker-6.50.tar.gz Checking if your kit is complete... Looks good 'INSTALLHTMLDIR' is not a known MakeMaker parameter name. Using included version of ExtUtils::Manifest (1.56) as it is newer than the installed version (1.46). Using included version of ExtUtils::Command (1.16) as it is newer than the installed version (1.09). Using included version of ExtUtils::Installed (1.43) as it is newer than the installed version (0.08). Using included version of ExtUtils::Packlist (1.43) as it is newer than the installed version (0.04). Using included version of ExtUtils::Install (1.52) as it is newer than the installed version (1.33). Writing Makefile for ExtUtils::MakeMaker cp inc/ExtUtils/Manifest.pm blib/lib/ExtUtils/Manifest.pm cp lib/ExtUtils/MM_VOS.pm blib/lib/ExtUtils/MM_VOS.pm cp lib/ExtUtils/Mksymlists.pm blib/lib/ExtUtils/Mksymlists.pm cp lib/ExtUtils/MM.pm blib/lib/ExtUtils/MM.pm cp lib/ExtUtils/MM_UWIN.pm blib/lib/ExtUtils/MM_UWIN.pm cp lib/ExtUtils/testlib.pm blib/lib/ExtUtils/testlib.pm cp lib/ExtUtils/MM_Cygwin.pm blib/lib/ExtUtils/MM_Cygwin.pm cp lib/ExtUtils/MakeMaker/vmsish.pm blib/lib/ExtUtils/MakeMaker/vmsish.pm cp lib/ExtUtils/MM_DOS.pm blib/lib/ExtUtils/MM_DOS.pm cp lib/ExtUtils/MM_Win95.pm blib/lib/ExtUtils/MM_Win95.pm cp lib/ExtUtils/Liblist.pm blib/lib/ExtUtils/Liblist.pm cp lib/ExtUtils/MM_AIX.pm blib/lib/ExtUtils/MM_AIX.pm cp lib/ExtUtils/MM_Darwin.pm blib/lib/ExtUtils/MM_Darwin.pm cp lib/ExtUtils/Liblist/Kid.pm blib/lib/ExtUtils/Liblist/Kid.pm cp lib/ExtUtils/Mkbootstrap.pm blib/lib/ExtUtils/Mkbootstrap.pm cp lib/ExtUtils/MakeMaker/FAQ.pod blib/lib/ExtUtils/MakeMaker/FAQ.pod cp inc/ExtUtils/MANIFEST.SKIP blib/lib/ExtUtils/MANIFEST.SKIP cp lib/ExtUtils/MakeMaker/bytes.pm blib/lib/ExtUtils/MakeMaker/bytes.pm cp lib/ExtUtils/MM_NW5.pm blib/lib/ExtUtils/MM_NW5.pm cp inc/ExtUtils/Command.pm blib/lib/ExtUtils/Command.pm cp lib/ExtUtils/MakeMaker.pm blib/lib/ExtUtils/MakeMaker.pm cp lib/ExtUtils/MM_OS2.pm blib/lib/ExtUtils/MM_OS2.pm cp lib/ExtUtils/MM_Unix.pm blib/lib/ExtUtils/MM_Unix.pm cp inc/ExtUtils/Installed.pm blib/lib/ExtUtils/Installed.pm cp lib/ExtUtils/MM_Win32.pm blib/lib/ExtUtils/MM_Win32.pm cp inc/ExtUtils/Packlist.pm blib/lib/ExtUtils/Packlist.pm cp lib/ExtUtils/MY.pm blib/lib/ExtUtils/MY.pm cp lib/ExtUtils/MM_MacOS.pm blib/lib/ExtUtils/MM_MacOS.pm cp lib/ExtUtils/MM_VMS.pm blib/lib/ExtUtils/MM_VMS.pm cp lib/ExtUtils/MM_BeOS.pm blib/lib/ExtUtils/MM_BeOS.pm cp lib/ExtUtils/MakeMaker/Tutorial.pod blib/lib/ExtUtils/MakeMaker/Tutorial.pod cp lib/ExtUtils/MM_QNX.pm blib/lib/ExtUtils/MM_QNX.pm cp inc/ExtUtils/Install.pm blib/lib/ExtUtils/Install.pm cp lib/ExtUtils/Command/MM.pm blib/lib/ExtUtils/Command/MM.pm cp lib/ExtUtils/MakeMaker/Config.pm blib/lib/ExtUtils/MakeMaker/Config.pm cp lib/ExtUtils/MM_Any.pm blib/lib/ExtUtils/MM_Any.pm cp bin/instmodsh blib/script/instmodsh /opt/perl_64/bin/perl "-Iblib/arch" "-Iblib/lib" "-MExtUtils::MY" -e "MY->fixin(shift)" blib/script/instmodsh Manifying blib/man1/instmodsh.1 Manifying blib/man3/ExtUtils::Manifest.3 Manifying blib/man3/ExtUtils::MM_VOS.3 Manifying blib/man3/ExtUtils::Mksymlists.3 Manifying blib/man3/ExtUtils::MM.3 Manifying blib/man3/ExtUtils::MM_UWIN.3 Manifying blib/man3/ExtUtils::testlib.3 Manifying blib/man3/ExtUtils::MM_DOS.3 Manifying blib/man3/ExtUtils::MM_Cygwin.3 Manifying blib/man3/ExtUtils::MakeMaker::vmsish.3 Manifying blib/man3/ExtUtils::MM_Win95.3 Manifying blib/man3/ExtUtils::Liblist.3 Manifying blib/man3/ExtUtils::MM_AIX.3 Manifying blib/man3/ExtUtils::MM_Darwin.3 Manifying blib/man3/ExtUtils::Mkbootstrap.3 Manifying blib/man3/ExtUtils::MakeMaker::FAQ.3 Manifying blib/man3/ExtUtils::MakeMaker::bytes.3 Manifying blib/man3/ExtUtils::MM_NW5.3 Manifying blib/man3/ExtUtils::Command.3 Manifying blib/man3/ExtUtils::MakeMaker.3 Manifying blib/man3/ExtUtils::MM_OS2.3 Manifying blib/man3/ExtUtils::MM_Unix.3 Manifying blib/man3/ExtUtils::Installed.3 Manifying blib/man3/ExtUtils::MM_Win32.3 Manifying blib/man3/ExtUtils::Packlist.3 Manifying blib/man3/ExtUtils::MY.3 Manifying blib/man3/ExtUtils::MM_MacOS.3 Manifying blib/man3/ExtUtils::MM_VMS.3 Manifying blib/man3/ExtUtils::MM_BeOS.3 Manifying blib/man3/ExtUtils::MakeMaker::Tutorial.3 Manifying blib/man3/ExtUtils::MM_QNX.3 Manifying blib/man3/ExtUtils::Command::MM.3 Manifying blib/man3/ExtUtils::Install.3 Manifying blib/man3/ExtUtils::MakeMaker::Config.3 Manifying blib/man3/ExtUtils::MM_Any.3 /usr/local/bin/gmake -- OK Running make test PERL_DL_NONLAZY=1 /opt/perl_64/bin/perl "-Iblib/arch" "-Iblib/lib" "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/00compile.............ok 28/56 skipped: Test::Pod not installed t/backwards.............ok t/basic.................ok 2/81 skipped: Test::YAML::Meta >= 0.11 required t/build_man.............ok t/bytes.................ok t/cd....................ok 1/4 skipped: VMS' cd requires vmspath which is only on VMS t/config................ok t/dir_target............ok t/FIRST_MAKEFILE........ok t/fixin.................ok t/hints.................ok t/INST..................ok t/INST_PREFIX...........ok t/INSTALL_BASE..........ok t/Liblist...............ok t/make..................ok t/maketext_filter.......ok t/metafile_data.........ok t/metafile_file.........ok 3/16 skipped: various reasons t/min_perl_version......ok t/Mkbootstrap...........ok t/MM_Any................ok t/MM_BeOS...............skipped all skipped: This is not BeOS t/MM_Cygwin.............skipped all skipped: This is not cygwin t/MM_NW5................skipped all skipped: This is not NW5 t/MM_OS2................skipped all skipped: This is not OS/2 t/MM_Unix...............ok t/MM_VMS................skipped all skipped: This is not VMS t/MM_Win32..............skipped all skipped: This is not Win32 t/oneliner..............ok t/parse_version.........ok 4/32 skipped: need version.pm t/PL_FILES..............ok 1/9 t/PL_FILES..............NOK 3# Failed test in t/PL_FILES.t at line 38. # got: '256' # expected: '0' # Makefile out-of-date with respect to Makefile.PL # Cleaning current config before rebuilding Makefile... # make -f Makefile.old clean > /dev/null 2>&1 # /opt/perl_64/bin/perl Makefile.PL # Writing Makefile for PL_FILES::Module # ==> Your Makefile has been rebuilt. <== # ==> Please rerun the make command. <== # false # *** Error exit code 1 # # Stop. # Failed test 'single.out was created' t/PL_FILES..............NOK 4# in t/PL_FILES.t at line 41. # Failed test '1.out was created' t/PL_FILES..............NOK 5# in t/PL_FILES.t at line 41. # Failed test '2.out was created' t/PL_FILES..............NOK 6# in t/PL_FILES.t at line 41. # Failed test 'blib/lib/PL/Bar.pm was created' t/PL_FILES..............NOK 7# in t/PL_FILES.t at line 41. # Looks like you failed 5 tests of 9. t/PL_FILES..............dubious Test returned status 5 (wstat 1280, 0x500) DIED. FAILED tests 3-7 Failed 5/9 tests, 44.44% okay t/pm....................ok t/postamble.............ok t/prefixify.............ok 1/3 skipped: Test for DOSish prefixification t/prereq................ok t/prereq_print..........ok t/problems..............ok t/prompt................ok t/recurs................ok 1/26 t/recurs................NOK 6# Failed test 'recursive make exited normally' # in t/recurs.t at line 61. # got: '256' # expected: '0' # Makefile out-of-date with respect to Makefile.PL # Cleaning current config before rebuilding Makefile... # make -f Makefile.old clean > /dev/null 2>&1 # /opt/perl_64/bin/perl Makefile.PL # Writing Makefile for Recurs::prj2 # Writing Makefile for Recurs # ==> Your Makefile has been rebuilt. <== # ==> Please rerun the make command. <== # false # *** Error exit code 1 # # Stop. t/recurs................ok 7/26 t/recurs................NOK 14# Failed test 'recursive make exited normally' # in t/recurs.t at line 84. # got: '256' # expected: '0' t/recurs................ok 24/26# Looks like you failed 2 tests of 26. t/recurs................dubious Test returned status 2 (wstat 512, 0x200) DIED. FAILED tests 6, 14 Failed 2/26 tests, 92.31% okay t/revision..............ok t/split_command.........ok t/testlib...............ok t/VERSION_FROM..........ok t/vmsish................ok t/WriteEmptyMakefile....ok t/writemakefile_args....ok 8/30 skipped: Can't test version objects t/xs....................skipped all skipped: ExtUtils::CBuilder not installed or couldn't find a compiler Failed Test Stat Wstat Total Fail Failed List of Failed ------------------------------------------------------------------------------- t/PL_FILES.t 5 1280 9 5 55.56% 3-7 t/recurs.t 2 512 26 2 7.69% 6 14 7 tests and 47 subtests skipped. Failed 2/48 test scripts, 95.83% okay. 7/661 subtests failed, 98.94% okay. gmake: *** [test_dynamic] Error 255 /usr/local/bin/gmake test -- NOT OK Running make install make test had returned bad status, won't install without force
CC: kahirsch [...] cpan.org
Subject: Re: [rt.cpan.org #29955] make test (of ExtUtils::MakeMaker) fails on HP-UX
Date: Tue, 31 Mar 2009 13:43:15 -0400
To: bug-ExtUtils-MakeMaker [...] rt.cpan.org
From: Ken Hirsch <kenhirsch [...] ftml.net>
It looks like the patch was not applied. 2009/3/31 Olivier 'dolmen' Mengué via RT <bug-ExtUtils-MakeMaker@rt.cpan.org>: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=29955 > > > This is still broken with EEMM 6.50 on HP-UX B.11.23 either with > /usr/bin/make and /usr/local/bin/gmake (GNU Make 3.80). > > Has the patch been applied? >
Sorry for the delay. The patch has been applied and will be in the next release. Sorry you had to edit it in four different places. The setup and teardown routines really should be consolidated.
RT-Send-CC: dolmen [...] cpan.org
Le Jeu. Avr. 02 08:29:54 2009, MSCHWERN a écrit : Show quoted text
> Sorry for the delay. The patch has been applied and will be in the next > release.
Thank you for the new release. However it looks like the same problem with HP-UX mtime makes Module::Build's t/compat.t fails. This test also generates a distribution. A more generic solution may be more appropriate. I'm thinking that EEMM could wait for one second if the Makefile.PL is too recent to be sure that the generated Makefile will be strictly older.
Subject: Re: [rt.cpan.org #29955] make test (of ExtUtils::MakeMaker) fails on HP-UX
Date: Mon, 04 May 2009 10:39:16 -0700
To: bug-ExtUtils-MakeMaker [...] rt.cpan.org
From: Michael G Schwern <schwern [...] pobox.com>
Olivier 'dolmen' Mengué via RT wrote: Show quoted text
> Queue: ExtUtils-MakeMaker > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=29955 > > > Le Jeu. Avr. 02 08:29:54 2009, MSCHWERN a écrit :
>> Sorry for the delay. The patch has been applied and will be in the next >> release.
> > Thank you for the new release. > > However it looks like the same problem with HP-UX mtime makes > Module::Build's t/compat.t fails. This test also generates a > distribution. A more generic solution may be more appropriate. I'm > thinking that EEMM could wait for one second if the Makefile.PL is too > recent to be sure that the generated Makefile will be strictly older.
I don't think its worth stealing a second from every module install just to fix something that only comes up in testing and then only on a relatively obscure OS. I'll patch up t/compat.t in MB. -- 124. Two drink limit does not mean first and last. -- The 213 Things Skippy Is No Longer Allowed To Do In The U.S. Army http://skippyslist.com/list/