14 years agoMention DTrace in INSTALL.
David Mitchell [Mon, 20 Jul 2009 12:51:56 +0000]
Mention DTrace in INSTALL.
Also explain that a debugging build is slower

14 years agoforgot the version Bump - sigh
Yves Orton [Sun, 19 Jul 2009 23:21:15 +0000]
forgot the version Bump - sigh

14 years agosync to 1.52_03 (hopefully properly this time)
Yves Orton [Sun, 19 Jul 2009 23:18:06 +0000]
sync to 1.52_03 (hopefully properly this time)

14 years ago(mostly) sync blead with ExtUtils::Install 1.52_02
David Mitchell [Sun, 19 Jul 2009 21:56:20 +0000]
(mostly) sync blead with ExtUtils::Install 1.52_02
(Installed.t still divergent)

14 years agoBe sure to find the vmsish pragma for one-liners in exit.t.
Craig A. Berry [Sun, 19 Jul 2009 20:46:31 +0000]
Be sure to find the vmsish pragma for one-liners in exit.t.

Already in upstream at:

http://github.com/schwern/test-more/commit/c842e916e4aa2e34c36afb3a930979d363121236

14 years agobring up to date with ExtUtils-Install v1.52_02
Yves Orton [Sun, 19 Jul 2009 20:40:14 +0000]
bring up to date with ExtUtils-Install v1.52_02

14 years agomro::method_changed_in(..) ignores AUTOLOAD (RT #60220)
Tony Cook [Sat, 18 Jul 2009 19:49:04 +0000]
mro::method_changed_in(..) ignores AUTOLOAD (RT #60220)

Patch modified to use a boolean rather than an integer for tracking
mro changes in S_glob_assign_ref and test fixed not to warn.

URL: http://rt.perl.org/rt3/Ticket/Display.html?id=60220

From the bug report:

-----------------------------------------------------------------
When creating a subclass dynamically, and when adding
AUTOLOAD dynamically into the parent class, then that
AUTOLOAD is not seen in the method cache, even after
a call to "mro::method_changed_in('Parent')".

It only appears in the method cache after a call
to mro::invalidate_all_method_caches().

The attached test file demonstrates the problem.

This was detected while trying to solve bug 40159 in DBIx::DataModel.
-----------------------------------------------------------------

Message-ID: <20081031132021.GA21341@mars.tony.develop-help.com>

14 years agoFix some nits in release_managers_guide.pod
Vincent Pit [Sat, 18 Jul 2009 21:31:25 +0000]
Fix some nits in release_managers_guide.pod

14 years agoadd extra note about CoreList to release_managers_guide.pod:
David Mitchell [Sat, 18 Jul 2009 21:06:38 +0000]
add extra note about CoreList to release_managers_guide.pod:

14 years agoadd info on updating CoreList and tags to release_managers_guide.pod
David Mitchell [Sat, 18 Jul 2009 20:58:43 +0000]
add info on updating CoreList and tags to release_managers_guide.pod

14 years agoProvisional update of CoreList for 5.10.1
David Mitchell [Sat, 18 Jul 2009 20:57:08 +0000]
Provisional update of CoreList for 5.10.1

(this will be done again just prior to RC1, but I wanted to try it out.
$VERSION has been bumped to what it will be for the final release.

14 years agoadd basic usage info to Porting/corelist.pl
David Mitchell [Sat, 18 Jul 2009 20:36:14 +0000]
add basic usage info to Porting/corelist.pl

14 years agoConfigure should sort inc_version_list
Andy Dougherty [Thu, 16 Jul 2009 20:52:39 +0000]
Configure should sort inc_version_list

In searching for previous versions of perl, Configure
forgets to sort the directories according to version number,
so 5.10 sorts later than 5.8.

Specifically, without this patch, you'd get something like:

    inc_version_list='5.8.9 5.8.8 5.10.0/i686-linux 5.10.0'

i.e. 5.10.0 would be searched *after* 5.8.8.

With this patch, you get something like:

    inc_version_list='5.10.0/i686-linux 5.10.0 5.8.9 5.8.8'

14 years agoSome of the TODOs in switches.t actually work on VMS.
Craig A. Berry [Sat, 18 Jul 2009 15:38:02 +0000]
Some of the TODOs in switches.t actually work on VMS.

14 years agofix pod2usage2 test for all case-insensitive systems
David Mitchell [Sat, 18 Jul 2009 00:10:20 +0000]
fix pod2usage2 test for all case-insensitive systems

This test, when run from core in t/, finds pod/usage.pod before
../lib/Pod/Usage.pm. Remove '.' from the search path for all core
platforms, not just cygwin.

14 years agoFix ExtUtils::Installed failure with -Duserelocatableinc
David E. Wheeler [Fri, 10 Jul 2009 17:14:39 +0000]
Fix ExtUtils::Installed failure with -Duserelocatableinc

This patch fixes an issue with ExtUtils::Installed when Perl is compiled with
userelocatableinc. It looks a though the issue is that `%Config` contains
local paths when built with -Duserelocatableinc, some, at least, with a
leading "./". The solution is to use `File::Spec->canonpath()` to clean up the
path before comparing two paths that are otherwise the same. A better solution
might be to use some sort of other utility function that checks that paths are
the same even if they're not spelled the same. I didn't notice such a function
in File::Spec, alas.

Also, I'm not sure what effects this change might have on VMS; it deserves
further testing there.

14 years agodocument why SQLite tests are exlcuded
David Mitchell [Fri, 17 Jul 2009 20:38:44 +0000]
document why SQLite tests are exlcuded

14 years agoAllocate XPV* memory using sizeof() and the offset of the first real member.
Nicholas Clark [Fri, 17 Jul 2009 16:35:57 +0000]
Allocate XPV* memory using sizeof() and the offset of the first real member.

This avoids tripping over alignment differences between X* and x*_allocated,
because x*_allocated doesn't have a double in it.

Tidy up the documentation in comments. Eliminate relative_STRUCT_OFFSET() as
it is no longer used. Eliminate use of xpv_allocated, xpviv_allocated,
xpvav_allocated, xpvcv_allocated and xpvfm_allocated, but leave them in the
headers as they are publicly visible.

14 years agoEliminate struct regexp_allocated and xpvio_allocated.
Nicholas Clark [Fri, 17 Jul 2009 15:17:58 +0000]
Eliminate struct regexp_allocated and xpvio_allocated.

Calculate memory allocation using regexp and XPVIO, and the offset of the first
real structure member. This avoids tripping over alignment differences between
X* and x*_allocated, because x*_allocated doesn't have a double in it.

14 years agohints/aix.sh remove libgdbm from wanted libs - dbm_store() is defective in the AIX...
Rainer Tammer [Fri, 17 Jul 2009 08:23:43 +0000]
hints/aix.sh remove libgdbm from wanted libs - dbm_store() is defective in the AIX Toolbox

14 years agoREADME.aix for Perl 5.10.1
Rainer Tammer [Fri, 17 Jul 2009 08:22:42 +0000]
README.aix for Perl 5.10.1

14 years agoDescribe changes to Carp and Carp::Heavy
Rafael Garcia-Suarez [Thu, 16 Jul 2009 21:46:45 +0000]
Describe changes to Carp and Carp::Heavy

14 years agoAdd a version number to B::Lint::Debug
Rafael Garcia-Suarez [Thu, 16 Jul 2009 21:46:16 +0000]
Add a version number to B::Lint::Debug

14 years agoNo whitespace before comment allowed in VMS make utilities.
Craig A. Berry [Thu, 16 Jul 2009 20:52:57 +0000]
No whitespace before comment allowed in VMS make utilities.

All extension building was broken without this.

14 years agoAdditional perlvms.pod updates related to POSIX exit handling.
Craig A. Berry [Thu, 16 Jul 2009 14:11:55 +0000]
Additional perlvms.pod updates related to POSIX exit handling.

14 years agoURL and other fix-ups for README.vms
Craig A. Berry [Thu, 16 Jul 2009 13:27:05 +0000]
URL and other fix-ups for README.vms

14 years agobump Encode version in Maintainers.pl
David Mitchell [Thu, 16 Jul 2009 12:46:50 +0000]
bump Encode version in Maintainers.pl

14 years agooverloading.pm will work with 5.10.1
David Mitchell [Thu, 16 Jul 2009 12:27:46 +0000]
overloading.pm will work with 5.10.1

14 years agoNote that editing pod.list then running pod/buildtoc automates these tasks.
Nicholas Clark [Thu, 16 Jul 2009 12:07:45 +0000]
Note that editing pod.list then running pod/buildtoc automates these tasks.

14 years agopatch submission(AUTHORS)
Adam Russell [Wed, 15 Jul 2009 05:16:23 +0000]
patch submission(AUTHORS)

Re-submitting this patch.
Adding myself and Osvaldo Villalon.
Best Regards,
Adam

----- Original Message -----
From: "Jonathan Stowe" <jns@integration-house.com>
To: "Adam Russell" <arussell@cs.uml.edu>
Cc: <perl5-porters@perl.org>
Sent: Friday, July 03, 2009 8:14 AM
Subject: Re: patch submission(AUTHORS)

> Hi,
> that patch appears to made against a really old
> version of the AUTHORS
> file as it seems to be reverting some changes made
> last year (my
> e-mail address for instance :-)
>
> 2009/7/3 Adam Russell <arussell@cs.uml.edu>:
>> Added myself(Adam Russell) and Osvaldo Villalon.
>
>
>
> --
> http://rabidgravy.com/ - Music
> http://gellyfish.co.uk/ - Everything else
>

>From a3b77323290dc0da9da193cb5ac3178c1f6708db Mon Sep 17 00:00:00 2001
From: Adam Russell <arussell@cs.uml.edu>
Date: Wed, 15 Jul 2009 00:59:35 -0400
Subject: [PATCH] Adding Adam Russell and Osvaldo Villalon(recent Symbian contributors).

14 years agopatch submisson(symbian/symbian_utils.cpp)
Adam Russell [Wed, 15 Jul 2009 05:27:26 +0000]
patch submisson(symbian/symbian_utils.cpp)

This patch adds some compatibility for older Symbian
SDKs.

>From 146c431f9030e275fcf9aca35d79a72eece128c2 Mon Sep 17 00:00:00 2001
From: Osvaldo Villalon <ovillalon@dextratech.com>
Date: Fri, 3 Jul 2009 01:26:33 -0400
Subject: [PATCH 1690/1692] Osvaldo Villalon's changes to symbian_utils.cpp.

14 years agoExtUtils::MakeMaker 6.55_01
David Mitchell [Thu, 16 Jul 2009 09:53:04 +0000]
ExtUtils::MakeMaker 6.55_01

14 years agoAn update to the pod is in order for the PERL_VMS_POSIX_EXIT logical
John E. Malmberg [Thu, 16 Jul 2009 02:45:57 +0000]
An update to the pod is in order for the PERL_VMS_POSIX_EXIT logical
name.  That should help you with what to put in the delta file.

I forgot to indicate in the file that the DECC$FILENAME_UNIX_REPORT
feature also activates the PERL_VMS_POSIX_EXIT mode.

14 years agoexpand Porting/release_managers_guide.pod
David Mitchell [Wed, 15 Jul 2009 22:59:21 +0000]
expand Porting/release_managers_guide.pod
and remove some similar or out-of-date stuff from Porting/pumpkin.pod

14 years agoDocument PL_opfreehook
Vincent Pit [Wed, 15 Jul 2009 20:32:36 +0000]
Document PL_opfreehook

14 years agoThe op_free() hook doesn't need to return an OP*, so introduce a new typedef for...
Vincent Pit [Wed, 15 Jul 2009 20:08:39 +0000]
The op_free() hook doesn't need to return an OP*, so introduce a new typedef for OP hooks

14 years agodocument given/when in perltrap
Moritz Lenz [Wed, 15 Jul 2009 14:53:37 +0000]
document given/when in perltrap

14 years agoadd Porting/bump-perl-version to MANIFEST
David Mitchell [Wed, 15 Jul 2009 10:51:23 +0000]
add Porting/bump-perl-version to MANIFEST

14 years agotest_harness, not just test, needs unpack_files on VMS.
Craig A. Berry [Tue, 14 Jul 2009 23:53:27 +0000]
test_harness, not just test, needs unpack_files on VMS.

14 years agoadd Porting/bump-perl-version
David Mitchell [Tue, 14 Jul 2009 22:22:37 +0000]
add Porting/bump-perl-version

utility for bumping the value of the perl version in lots of files

14 years agofix some perl version numbers
David Mitchell [Tue, 14 Jul 2009 18:20:06 +0000]
fix some perl version numbers

14 years agoUpdate my address.
Abhijit Menon-Sen [Tue, 14 Jul 2009 14:44:43 +0000]
Update my address.

14 years agoExplain camel and dromedary repository servers
David Golden [Tue, 14 Jul 2009 03:16:35 +0000]
Explain camel and dromedary repository servers

Patch provided by Dennis Kaarsemaker.

14 years agoDon't preload Carp::Heavy in warnings.pm
Rafael Garcia-Suarez [Tue, 14 Jul 2009 07:41:42 +0000]
Don't preload Carp::Heavy in warnings.pm

Use only Carp, now that all the machinery is in it.

14 years agoMerge Carp::Heavy into Carp
Rafael Garcia-Suarez [Tue, 14 Jul 2009 07:16:17 +0000]
Merge Carp::Heavy into Carp

14 years agoupate pumpking list
David Golden [Mon, 13 Jul 2009 13:03:47 +0000]
upate pumpking list

14 years agoCosmetic change
Stepan Kasal [Mon, 13 Jul 2009 08:26:27 +0000]
Cosmetic change

14 years agoUpgrade to Encode 2.35
Rafael Garcia-Suarez [Mon, 13 Jul 2009 07:40:04 +0000]
Upgrade to Encode 2.35

14 years agoClean up t/op/sort.t to constrain tests which use a particular package.
Michael G. Schwern [Sat, 11 Jul 2009 07:34:31 +0000]
Clean up t/op/sort.t to constrain tests which use a particular package.

Also make use of is() instead of a home rolled conflicting Test.pm style ok().

14 years agoBetter flock detection for autodie tests
Paul Fenwick [Mon, 13 Jul 2009 01:30:28 +0000]
Better flock detection for autodie tests

G'day p5p,

Craig A. Berry discovered that autodie's tests which attempt to detect
flock() support can cause bogus test failures on older VMS systems:

https://rt.cpan.org/Ticket/Display.html?id=47812

Attached is a trivial patch from Craig that fixes this issue.  It's already
been applied to my upstream repository.

Dave - I have no idea if this would be an issue for people building Perl
5.10.1 on old VMS systems.  You may wish to consider it for inclusion into
maint-5.10 just in case.

All the very best,

Paul

--
Paul Fenwick <pjf@perltraining.com.au> | http://perltraining.com.au/
Director of Training                   | Ph:  +61 3 9354 6001
Perl Training Australia                | Fax: +61 3 9354 2681

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoh2ph: handle "#if defined SYMBOL" better
Stepan Kasal [Thu, 9 Jul 2009 13:49:57 +0000]
h2ph: handle "#if defined SYMBOL" better

expr() contains a code to handle "defined(SYM)" in #if directives.
Unfortunately, this code is not executed for "defined SYM", without
parentheses.  This patch fixes it.

14 years agoSome missing realclean items for VMS.
Craig A. Berry [Sun, 12 Jul 2009 14:27:48 +0000]
Some missing realclean items for VMS.

14 years agoWrong skip count in t/op/stash.t
Michael G Schwern [Sun, 12 Jul 2009 09:38:42 +0000]
Wrong skip count in t/op/stash.t

stash.t has the wrong number of tests to skip when dynamic loading is not
available breaking minitest.

14 years ago[perl #61520] Segfault in debugger with tr// and UTF8
David Mitchell [Sat, 11 Jul 2009 16:56:06 +0000]
[perl #61520] Segfault in debugger with tr// and UTF8

commit 043e41b8 (29765), which made tr// threadsafe by moving the
swash into the pad, didn't mark the pad SV as read-only, so it was getting
removed from anon sub prototypes

14 years agoSubject: [PATCH] Have add-package.pl include the Changes file
H.Merijn Brand [Sat, 11 Jul 2009 15:54:40 +0000]
Subject: [PATCH] Have add-package.pl include the Changes file
From: David Golden <dagolden@cpan.org>
Date: Fri, 10 Jul 2009 08:20:06 -0400
Message-Id: <1247228406-29001-1-git-send-email-dagolden@cpan.org>

Also adds a "-c FILE" option for an alternate change log name

I've extended it With additional support for ChangeLog and open for more

14 years agoMaintainers.pl: bump Term::ANSIColor version
David Mitchell [Fri, 10 Jul 2009 15:58:58 +0000]
Maintainers.pl: bump Term::ANSIColor version

14 years agoExtUtils::MakeMaker 6.54
David Mitchell [Fri, 10 Jul 2009 14:16:14 +0000]
ExtUtils::MakeMaker 6.54

14 years agoRemove %patchlevel hash
Rafael Garcia-Suarez [Fri, 10 Jul 2009 14:00:37 +0000]
Remove %patchlevel hash

(as the perforce repository has been decomissionned, this information
is no longer useful and can be retrieved by other means. Every release
listed here has a tag in git)

14 years agoEncode 2.34
David Mitchell [Fri, 10 Jul 2009 13:33:57 +0000]
Encode 2.34

14 years agoadd lib/ExtUtils/ParseXS/Changes to MANIFEST
H.Merijn Brand [Fri, 10 Jul 2009 10:49:24 +0000]
add lib/ExtUtils/ParseXS/Changes to MANIFEST

14 years agoFix compiler warning in doio.c
Jerry D. Hedden [Thu, 9 Jul 2009 20:29:23 +0000]
Fix compiler warning in doio.c

14 years agoAvoid test failure on Linux when STDIN is /dev/null
Gisle Aas [Thu, 9 Jul 2009 17:27:12 +0000]
Avoid test failure on Linux when STDIN is /dev/null

sysseek($fh, -1, 0) succeed on Linux when $fh is a handle to /dev/null.
In our build environment the perl test suite runs with STDIN redirected
from /dev/null and without a /dev/tty.  The hope is that the same seek
on a regular file will always fail.

14 years agoUpgrade to Term::ANSIColor 2.01
Rafael Garcia-Suarez [Thu, 9 Jul 2009 13:41:34 +0000]
Upgrade to Term::ANSIColor 2.01

14 years agoMake kill() fatal for non-numeric pids
David Golden [Wed, 8 Jul 2009 17:28:54 +0000]
Make kill() fatal for non-numeric pids

As the debate over the best way to deal with floating point
pids stalled, this is just for non-numeric, which at least
squashes the bug even if it's not the Platonic ideal for
everyone.

It also doesn't address overloaded objects that might not have
IV, NV or PV appropriately set, but the approach mirrors what is
done elsewhere in doio.c so I recommend applying this patch now and
fixing the problem of overloaded objects at some other time when
it can be done more globally, either through an improvement or
replacement of looks_like_number

Also updated POD for kill when process is 0 or negative and
fixed Test-Harness tests that used kill with a string pid.
(Test-Harness test fix also submitted upstream)

14 years agoMake INSTALL -Duserelocatableinc aware
nperez [Wed, 24 Jun 2009 01:32:31 +0000]
Make INSTALL -Duserelocatableinc aware

14 years agoMerge commit '1640e9f0f39bf8108370c2f5d99bb100fe6c9418' into blead
Chip Salzenberg [Thu, 9 Jul 2009 07:44:18 +0000]
Merge commit '1640e9f0f39bf8108370c2f5d99bb100fe6c9418' into blead

14 years agofix segfault in -Dx without threads
Chip Salzenberg [Thu, 9 Jul 2009 07:39:41 +0000]
fix segfault in -Dx without threads

14 years agoSkip File::Copy return tests on Windows and VMS
Paul Fenwick [Thu, 9 Jul 2009 04:04:35 +0000]
Skip File::Copy return tests on Windows and VMS

Many thanks to Craig Berry for tracking this down.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoModify the return value of Gv_AMupdate to indicate a compilation error
Rafael Garcia-Suarez [Wed, 8 Jul 2009 22:03:47 +0000]
Modify the return value of Gv_AMupdate to indicate a compilation error

This way we'll restore most of the performance on object desctruction
lost by the previous commit

14 years agoAdd a parameter "destructing" to Gv_AMupdate()
Rafael Garcia-Suarez [Wed, 8 Jul 2009 14:56:50 +0000]
Add a parameter "destructing" to Gv_AMupdate()

This boolean parameter indicates if the function has been called
to update the overload magic table while looking up the DESTROY
method. In this case, it's probably best to avoid croaking if
those tables could not be updated (for example due to a method
that could not be loaded.)

14 years agoUpdate ExtUtils::ParseXS to 2.20_01
David Golden [Wed, 8 Jul 2009 17:13:56 +0000]
Update ExtUtils::ParseXS to 2.20_01

2.20_01 - Wed Jul  8 12:12:47 EDT 2009

 - Fix XSUsage prototypes for testing [Jan Dubois]

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoAdd a pluggable hook in op_free()
Vincent Pit [Wed, 8 Jul 2009 14:49:36 +0000]
Add a pluggable hook in op_free()

14 years agoMention in perldelta that "use 5.011" enables strictures
Rafael Garcia-Suarez [Wed, 8 Jul 2009 09:58:41 +0000]
Mention in perldelta that "use 5.011" enables strictures

14 years agoIgnore libperl.so.
Florian Ragwitz [Wed, 8 Jul 2009 09:23:26 +0000]
Ignore libperl.so.

14 years agoFix compiler warning when cuserid is absent and sizeof(int) != sizeof(pointer)
Tim Jenness [Tue, 7 Jul 2009 21:49:11 +0000]
Fix compiler warning when cuserid is absent and sizeof(int) != sizeof(pointer)

The attached patch fixes a compilation warning from POSIX.xs when cuserid
is missing and ints and pointers have different size. eg on Mac OSX when
using 64-bit mode.

gcc-4.2 -c   -fno-common -DPERL_DARWIN -no-cpp-precomp -arch x86_64 -DDEBUGGING -fno-strict-aliasing -pipe -fstack-protector
-I/usr/local/include -I/opt/local/include -O3 -g   -DVERSION=\"1.17\"
-DXS_VERSION=\"1.17\"  "-I../.."   POSIX.c
POSIX.c: In function 'XS_POSIX_cuserid':
POSIX.c:4096: warning: cast to pointer from integer of different size
Running Mkbootstrap for POSIX ()
chmod 644 POSIX.bs
rm -f ../../lib/auto/POSIX/POSIX.bundle

Also relevant for 5.10.1.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoUpdating Module::Build to 0.34
David Golden [Wed, 8 Jul 2009 02:15:59 +0000]
Updating Module::Build to 0.34

0.34 - Tue Jul  7 16:56:47 EDT 2009

 No changes from 0.33_06

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoMake File::Copy::rmscopy return the documented values.
Craig A. Berry [Wed, 8 Jul 2009 03:29:10 +0000]
Make File::Copy::rmscopy return the documented values.

These were always supposed to be 0 for failure and 1 for success,
not booleans.  On VMS, rmscopy implements syscopy.

14 years agoMerge branch 'strict-by-default' into blead
Rafael Garcia-Suarez [Tue, 7 Jul 2009 20:32:39 +0000]
Merge branch 'strict-by-default' into blead

14 years agosome fixes to makerel:
David Mitchell [Tue, 7 Jul 2009 17:48:36 +0000]
some fixes to makerel:
 * add cmd-line switches
 * make ExtUtils::Manifest silent
 * remove some autogenerated files from the fixup list

14 years agoupdate to CPANPLUS 0.88
David Mitchell [Tue, 7 Jul 2009 16:06:43 +0000]
update to CPANPLUS 0.88

14 years agorebump Hash::Util::FieldHash from 1.03_01 to 1.04
David Mitchell [Tue, 7 Jul 2009 14:40:58 +0000]
rebump Hash::Util::FieldHash from 1.03_01 to 1.04
(it's not dual-lived so there's no need for a devel version number)

14 years agoFast enabling of strictures when version 5.11.0 is required
Rafael Garcia-Suarez [Tue, 7 Jul 2009 09:04:13 +0000]
Fast enabling of strictures when version 5.11.0 is required

We don't load strict.pm, we just manipulate the hint bits.
Plus more tests.

14 years agouse strict by default if "use 5.011" is in effect!
Steffen Mueller [Tue, 7 Jul 2009 08:27:50 +0000]
use strict by default if "use 5.011" is in effect!

14 years agolimit what add-package.pl might try to delete
David Golden [Tue, 7 Jul 2009 08:10:33 +0000]
limit what add-package.pl might try to delete

Old way looked in perl source for files matching exclude
list to delete.  New way looks only in source distribution.

This means that files already in perl source will need to be
manually removed, but avoids the risk that a bad exclusion
regex blows away unrelated things in the source tree.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoupgrade to ExtUtils::MakeMaker 6.53_03
David Mitchell [Mon, 6 Jul 2009 21:16:57 +0000]
upgrade to ExtUtils::MakeMaker 6.53_03

14 years agoRegex fails when string is too long
hv@crypt.org [Mon, 6 Jul 2009 14:45:12 +0000]
Regex fails when string is too long

This looks to be a simple oversight. All tests pass here.

Hugo

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoDon't dump autodie from core (was Re: Coring Variable::Magic / autodie fights with...
Paul Fenwick [Mon, 6 Jul 2009 14:17:12 +0000]
Don't dump autodie from core (was Re: Coring Variable::Magic / autodie fights with string eval in Perl 5.10.x)

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

G'day Dave,

Dave Mitchell wrote:

> Okay I think you've convinced me. So basically you'll release a 2.06 which
> is 2.05 + bug documentation; 2.06 goes in 5.10.1; then the bug fix goes in
> 2.07?

Patch attached, including documentation fixes, and TODO tests.

I did slip a small code change in, which fixes a regression that was
introduced (but not spotted) around autodie 1.997ish.  Many thanks to rafl
and particularly to Vincent for their help in debugging and documenting.

All the best,

Paul

- --
Paul Fenwick <pjf@perltraining.com.au> | http://perltraining.com.au/
Director of Training                   | Ph:  +61 3 9354 6001
Perl Training Australia                | Fax: +61 3 9354 2681
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (MingW32)

iD8DBQFKUgdnx5N6j7FHnlURAhcbAJ4zQ5WujSXK5uNpEfdOp0RMi8FxVQCghU02
Z3ZBYlXddRMOmm9SVC7UPTE=
=qf29
-----END PGP SIGNATURE-----

>From 689bc09292ba5173542d208dca85b6f514c91aba Mon Sep 17 00:00:00 2001
From: Paul Fenwick <pjf@perltraining.com.au>
Date: Tue, 7 Jul 2009 00:11:42 +1000
Subject: [PATCH] Merge autodie 2.06 with core.

NOTE: This commit is different from that on the CPAN by three lines
of POD.  A documented bug was fixed, but documentation about this bug
was accidentally not removed from the CPAN release.

Significant changes in this release are:

        * BUG: Explicitly documented that autodie does NOT play
          nicely with string evals, especially under Perl 5.10.x.
          Please avoid using string evals while autodie is in scope.

        * TEST: Check for autodie leaking out of scope in the
          presence of string evals.  (string-eval-leak.t)
          Thanks to Florian Ragwitz and Vincent Pit for identifying
          this.

        * BUGFIX: autodie once again correctly works when used
          inside a string eval.  (This was accidently broken
          somewhere around 1.997-1.998).

14 years agore-add Test::Simple exclude accidentally removed during 0.92 upgrade
David Mitchell [Mon, 6 Jul 2009 14:47:43 +0000]
re-add Test::Simple exclude accidentally removed during 0.92 upgrade

14 years agoclarify usage of Porting/Maintainers
David Mitchell [Mon, 6 Jul 2009 14:43:17 +0000]
clarify usage of Porting/Maintainers

14 years agoadd core boilerplate to three new Test::Simple tests
David Mitchell [Mon, 6 Jul 2009 14:39:34 +0000]
add core boilerplate to three new Test::Simple tests

14 years agoupdate to Test::Simple 0.92
David Mitchell [Mon, 6 Jul 2009 13:59:06 +0000]
update to Test::Simple 0.92

14 years agoMaintainers.pl: Explanation as to why autodie tests are excluded.
Paul Fenwick [Mon, 6 Jul 2009 10:27:49 +0000]
Maintainers.pl: Explanation as to why autodie tests are excluded.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agobump autodie veresion in Maintainers.pl
David Mitchell [Mon, 6 Jul 2009 10:33:33 +0000]
bump autodie veresion in Maintainers.pl

14 years agosync Module::Build with CPAN
David Mitchell [Mon, 6 Jul 2009 09:43:53 +0000]
sync Module::Build with CPAN

14 years agosync ExtUtils::CBuilder with CPAM
David Mitchell [Mon, 6 Jul 2009 09:41:10 +0000]
sync ExtUtils::CBuilder with CPAM

14 years agoAdded missing 2009 to perl.h copyright header
Offer Kaye [Mon, 6 Jul 2009 09:02:52 +0000]
Added missing 2009 to perl.h copyright header

14 years agoUpdate Module::Build to 0.33_06
David Golden [Mon, 6 Jul 2009 03:37:32 +0000]
Update Module::Build to 0.33_06

0.33_06 - Sun Jul  5 10:11:40 EDT 2009

 Bug-fixes:
 - Bundled version code will use pure Perl on 5.10.0 to work around
   a corner case involving eval and locale [John Peacock]
 - Reversed VMS patch from 0.33_03 [Craig Berry]
 - PL_files in Build.PL that are in the bin/scripts directory should not be
   installed as if they are scripts (fixed for case-tolerant systems).
   [David Golden, reported by Craig Berry]

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoUpdated ExtUtils::CBuilder to 0.2602
David Golden [Sat, 4 Jul 2009 15:23:22 +0000]
Updated ExtUtils::CBuilder to 0.2602

0.2602 - Sat Jul  4 10:57:12 EDT 2009

 Bugs fixed:
 - Fixed 00-have-compiler.t if $^X is a relative path [David Wheeler]

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoUpdate CPANPLUS::Dist::Build to CPAN version 0.36
Chris 'BinGOs' Williams [Sat, 4 Jul 2009 14:06:19 +0000]
Update CPANPLUS::Dist::Build to CPAN version 0.36

Hi,

Attached is a patch to update CPANPLUS::Dist::Build in core to
CPAN version 0.36.

Highlights from the Change Log:

==============================================================
0.36 Sat Jul  4 14:51:42 BST 2009
  - Make install() honour the buildflags directive,
    absence spotted by Matt Trout
==============================================================

Many thanks,

--
Chris Williams
aka BinGOs
PGP ID 0x4658671F
http://www.gumbynet.org.uk
==========================

From 43b0b0b2f93858517fa5d21566bcd807d4bf434a Mon Sep 17 00:00:00 2001
From: Chris Williams <chris@bingosnet.co.uk>
Date: Sat, 4 Jul 2009 15:02:09 +0100
Subject: [PATCH] Update CPANPLUS::Dist::Build to CPAN version 0.36

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>

14 years agoMerge autodie 2.05 into core.
Paul Fenwick [Sat, 4 Jul 2009 06:41:39 +0000]
Merge autodie 2.05 into core.

2.05  Sat Jul  4 16:33:01 AUSEST 2009

        * BUGFIX: format_default() in autodie::exception no longer
          returns a string with file and line attached.  This would
          cause the file and line information to appear twice when
          format handlers would choose to fall back to the defaults.
          The file and line information is now always added by
          stringify().  (RT #47520, thanks to Michael Schwern)

        * BUGFIX: Exceptions thrown by 2-argument open() are more likely
          to specify the mode as 'for reading' when no explicit
          mode was given.  (RT #47520, thanks to Michael Schwern)

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>