Re: Perl 5.12.2-RC1 tests

Martin,

Thanks for the update. Unfortunately, in my case at least, directory 
protections are not the issue as the user account that executes the 
configure, build and test steps is also the owner of the directory tree 
containing the perl source archive being built. So, I will have to do 
more digging to try and figure out why you only get 6 test failures on 
OpenVMS Integrity V8.3-1H1 and I get 17 test failures with no other 
obvious differences in the build environment besides the fact that I am 
building perl using OpenVMS Alpha V8.3. I have not yet looked in detail 
at installed patches yet. Before I resort to analyzing patches,  I am 
going to dig into the tests to see where things are breaking and try and 
work backward from there and only if I come up empty there will I look 
at patches.

Eric

Vorländer wrote:
> Okay, it works out quite good now:
>
> Failed 6 tests out of 1631, 99.63% okay.
>          ../cpan/CPANPLUS/t/03_CPANPLUS-Internals-Source.t
>          ../cpan/CPANPLUS/t/15_CPANPLUS-Shell.t
>          ../cpan/Module-Build/t/tilde.t
>          ../cpan/Module-Build/t/use_tap_harness.t
>          ../cpan/Time-HiRes/t/HiRes.t
>          ../lib/vmsish.t
>
> The Module::Build tests don't work with TEST.COM - Message
> "Can't locate MBTest.pm in @INC (@INC contains: t/lib ../lib .)
> at [-.cpan.module-build.t]use_tap_harness.t line 12."
> and "... at [-.cpan.module-build.t]tilde.t line 7."
>
> The CPANPLUS tests have a problem with a temp directory:
> "cannot opendir for SYS$SYSROOT:[SYSMGR.Dhv0a3vjPb]:
> directory not found at /SYS$SYSDEVICE/BUILD/perl-5_12_2-RC1/lib/File/Temp.pm line 902"
> which leads to "FAILED--extra output after trailing 1..n"
> Apart from that, everything's fine.
>
> I have to look deeper into the Time::HiRes failure. When
> testing it through TEST.COM, it says
> ...
> ok 37^M
> # No effectual d_hires_stat
> ok 38 # skipped
> t/[-.CPAN.TIME-HIRES.T]HIRES....FAILED--expected 40 tests, saw 38
> Failed 1 test out of 1, 0.00% okay.
>
> vmsish.t fails at test 22: "not ok 22 - (time) UTC: 1283584044 VMS: 1283584044"
> even though it should have set the offset to 3600.
> No idea there...
>
> cu,
> Martin
>