On May 19, 2012, at 3:15 AM, Nicholas Clark wrote:

> tl;dr: Not everything passed. I think more failed than for Craig on
> his system. But it's not much different in its failures from 5.14.0.

It's not that my system is optimized for building Perl but that building Perl 
is optimized for my system, because that's where it gets built, several times a 
week, year in and year out.

> I don't know how to get the shared library override set up to do PERL -V

$ @PTAC$DKA0:[NCLARK.I.PERL5160-RC2-MMK]perl_setup
$ define/trans=conc perl_root PTAC$DKA0:[NCLARK.I.PERL5160-RC2-MMK.]

will make the perl built in that directory behave like an installed one; you'll 
get the "perl" command pointing to the right place and you don't even need to 
add -Ilib to everything.

Specific test failures snipped.

The next thing I would check would be what CRTL features are enabled in the 
environment, which you can do with:

$ show logical DECC$*

I generally build and test with all of them turned off, which means that's what 
gets tested on a regular basis.  

Other random notes:

The tests for Archive::Extract are very fragile in their expectations about 
what versions of what external utilities are available and how they should be 
invoked.  There are two different tar utilities likely to be found on VMS 
systems, and either or neither may be defined as "tar".  One utility (vmstar) 
can handle C<tar -"tf"> but not C<tar "-tf"> (with the minus inside the 
quotes).  It seems crazy to hard-code all the possible bug-compatible options 
into Archive::Extract.

The Time::HiRes failures are related to a CRTL bug that has been fixed, but the 
HP system appears not to be up-to-date on HP patches.

I've seen the same ext/Pod-Html/t/cache.t failure with RC2 but not with blead.  
When it fails, it looks like:

$ perl [.t]cache.t
1..10
ok 1 - No cache file to start
ok 2 - No cache file to start
ok 3 - Cache created
ok 4 - podpath
ok 5 - podroot
ok 6 - Cache created
ok 7 - podpath
not ok 8 - cache contents
#   Failed test 'cache contents'
#   at [.t]cache.t line 65.
#     Structures begin differing at:
#          $got->{eol.t} = 't/eol'
#     $expected->{eol.t} = Does not exist
ok 9 - No cache file to end
ok 10 - No cache file to end
# Looks like you failed 1 test of 10.

I suspect it's sensitive to the length of the path it's in or something like 
that, though I haven't felt motivated to look into the bowels of Pod::Html 
again.

________________________________________
Craig A. Berry
mailto:craigbe...@mac.com

"... getting out of a sonnet is much more
 difficult than getting in."
                 Brad Leithauser

Reply via email to