Nicholas Clark [Wed, 3 Oct 2007 20:34:13 +0000]
Test for a subtle pre-5.10 bug. Before 5.10 the overloading flag was
stored on the reference rather than the referent. Despite the fact that
objects can only be accessed via references (even internally), the
referent actually knows that it's blessed, not the references. So
taking a new, unrelated, reference to it gives an object. However, the
overloading-or-not flag was on the reference prior to 5.10, and taking
a new reference didn't (use to) copy it (prior to 5.8.9).
So test that the bug can't return - overloading should work on a
reference to something already blessed into a package with overloading.
p4raw-id: //depot/perl@32016
Rafael Garcia-Suarez [Wed, 3 Oct 2007 16:21:59 +0000]
Reverse change #31978
p4raw-link: @31978 on //depot/perl:
d804f4346b490171e547d5cc512063e53da10708
p4raw-id: //depot/perl@32015
Marcus Holland-Moritz [Wed, 3 Oct 2007 11:29:58 +0000]
One const too much. (Unfortunately, the Copy macro casts
the constness away, so the compiler cannot detect this.)
p4raw-id: //depot/perl@32014
Rafael Garcia-Suarez [Wed, 3 Oct 2007 11:21:59 +0000]
Adapt definition of Sighandler_t to go with change #32012
p4raw-link: @32012 on //depot/perl:
b6455c53c26be8a62e12a3f2a24a3a5086dd2c7b
p4raw-id: //depot/perl@32013
Slaven Rezic [Sat, 29 Sep 2007 05:38:08 +0000]
[perl #45513] Test failures on amd64-freebsd 6.2
From: "slaven@rezic.de via RT" <perlbug-followup@perl.org>
Message-ID: <rt-3.6.HEAD-5916-
1191069483-1553.45513-15-0@perl.org>
p4raw-id: //depot/perl@32012
ben@linuxgazette.net [Tue, 2 Oct 2007 21:33:49 +0000]
[perl #46019] Documentation for 'srand' has a problem in code snippet
From: ben@linuxgazette.net (via RT) <perlbug-followup@perl.org>
Message-ID: <rt-3.6.HEAD-5916-
1191386029-1116.46019-75-0@perl.org>
p4raw-id: //depot/perl@32011
Steve Peters [Wed, 3 Oct 2007 04:07:08 +0000]
Update to a file missed in change #32008.
p4raw-link: @32008 on //depot/perl:
547d3dfd0918e1335aa86e352904fbb22cccfb6d
p4raw-id: //depot/perl@32010
Steve Peters [Tue, 2 Oct 2007 13:56:07 +0000]
Upgrade to Getopt-Long-2.37
p4raw-id: //depot/perl@32009
Steve Peters [Tue, 2 Oct 2007 13:39:06 +0000]
Upgrade to CPAN-1.9203
p4raw-id: //depot/perl@32008
Rafael Garcia-Suarez [Tue, 2 Oct 2007 12:44:18 +0000]
Need to quote the version number of perl in perlivp, now
that it finishes by 0
p4raw-id: //depot/perl@32007
Jerry D. Hedden [Mon, 1 Oct 2007 19:32:19 +0000]
/cygdrive is configurable
From: "Jerry D. Hedden" <jdhedden@cpan.org>
Message-ID: <
1ff86f510710011632n175427fdx39e173372862526e@mail.gmail.com>
p4raw-id: //depot/perl@32006
Shawn [Mon, 1 Oct 2007 12:45:32 +0000]
use base problem
Message-ID: <
1191267932.150186.12570@d55g2000hsg.googlegroups.com>
(This patch is actually not needed for 5.10, but useful for earlier perls)
p4raw-id: //depot/perl@32005
Jim Cromie [Mon, 1 Oct 2007 18:59:43 +0000]
Re: [patch] clean out remnants of DEBUGGING_OPS (was opcode op_type:9)
Message-ID: <
470197FF.5040709@gmail.com>
p4raw-id: //depot/perl@32004
John Peacock [Tue, 2 Oct 2007 05:28:31 +0000]
was Re: Freeze ?
Message-ID: <
47020F3F.9070604@havurah-software.org>
p4raw-id: //depot/perl@32003
Nicholas Clark [Tue, 2 Oct 2007 07:21:12 +0000]
Audrey spotted a spurious "Unicode".
p4raw-id: //depot/perl@32002
Nicholas Clark [Mon, 1 Oct 2007 12:13:36 +0000]
Nothing is using IoSUBPROCESS() so eliminate xio_subprocess.
Make xio_flags a U8, which is what it should have been all along
(rather than a char, which might be signed). B.xs had its wrapper as U8
p4raw-id: //depot/perl@32001
Rafael Garcia-Suarez [Mon, 1 Oct 2007 12:00:30 +0000]
Patch by jlv <jloverso@mathworks.com> to stop NEXT::AUTOLOAD
from emitting warnings under the debugger.
See: http://rt.cpan.org/Public/Bug/Display.html?id=25053
p4raw-id: //depot/perl@32000
John E. Malmberg [Sun, 30 Sep 2007 09:13:09 +0000]
[patch@31998] Fix M:B: tilde.t tests on VMS
From: "John E. Malmberg" <wb8tyw@qsl.net>
Message-id: <
46FFAEF5.1060702@qsl.net>
p4raw-id: //depot/perl@31999
John E. Malmberg [Sat, 29 Sep 2007 14:50:37 +0000]
Re: [patch@31988] Revised Module::Build fixes for VMS.
From: "John E. Malmberg" <wb8tyw@qsl.net>
Message-id: <
46FEAC8D.1080506@qsl.net>
Missed one case in #31995.
p4raw-id: //depot/perl@31998
David Landgren [Sun, 30 Sep 2007 15:19:49 +0000]
bring File::Path up to 2.01 (was: Re: Freeze ?)
Message-ID: <
46FFA275.6010304@landgren.net>
p4raw-id: //depot/perl@31997
John E. Malmberg [Fri, 28 Sep 2007 08:55:27 +0000]
[patch@31988] Revised Module::Build fixes for VMS.
From: "John E. Malmberg" <wb8tyw@qsl.net>
Message-id: <
46FD07CF.5040504@qsl.net>
p4raw-id: //depot/perl@31995
Nicholas Clark [Fri, 28 Sep 2007 18:04:55 +0000]
Make all of B work on 5.8.x
p4raw-id: //depot/perl@31994
Rafael Garcia-Suarez [Fri, 28 Sep 2007 13:07:35 +0000]
Document hint symbols in B::Concise
p4raw-id: //depot/perl@31993
Nicholas Clark [Fri, 28 Sep 2007 12:40:39 +0000]
Hopefully really fix test failures with -C
p4raw-id: //depot/perl@31992
Rafael Garcia-Suarez [Fri, 28 Sep 2007 08:20:32 +0000]
Remove unneeded cast
p4raw-id: //depot/perl@31991
Nicholas Clark [Fri, 28 Sep 2007 07:32:28 +0000]
Change 31987 forgot to re-run embed.pl
p4raw-id: //depot/perl@31990
Jim Cromie [Thu, 27 Sep 2007 19:29:43 +0000]
[patch] s/thinngy/thingy/
Message-ID: <
46FC5907.4010205@gmail.com>
p4raw-id: //depot/perl@31989
Rafael Garcia-Suarez [Thu, 27 Sep 2007 09:55:35 +0000]
I forgot to patch for the threaded case
p4raw-id: //depot/perl@31988
Nicholas Clark [Thu, 27 Sep 2007 09:40:58 +0000]
Move the bool v_string_ok into the U16-sized gap in the interpreter
structure. Add a "spare" variable so that we can add the first post-
5.10.0 bool variable without growing the interpreter structure.
p4raw-id: //depot/perl@31987
Rafael Garcia-Suarez [Thu, 27 Sep 2007 09:38:53 +0000]
Hopefully fix test failures with -C
p4raw-id: //depot/perl@31986
Steve Hay [Thu, 27 Sep 2007 08:56:05 +0000]
Fix test introduced in #31973 that failed on Win32
(Win32 doesn't like the runperl() prog having newlines in it)
p4raw-id: //depot/perl@31985
Craig A. Berry [Thu, 27 Sep 2007 00:57:00 +0000]
Make large file support the default on VMS systems that have it. This in turn
makes _USE_STD_STAT and symbolic link support the defaults where available.
p4raw-id: //depot/perl@31984
Rafael Garcia-Suarez [Wed, 26 Sep 2007 16:57:34 +0000]
API spelling patch, by Jerry D. Hedden
p4raw-id: //depot/perl@31983
Rafael Garcia-Suarez [Wed, 26 Sep 2007 16:46:36 +0000]
Maybe the default order of directories in @INC should be modified.
p4raw-id: //depot/perl@31982
Nicholas Clark [Wed, 26 Sep 2007 16:32:45 +0000]
Change 31977 introduced a bug (failing to set the return value of
mro::get_linear_isa() when passed a second parameter). However, there
were no tests for this. Fix both.
p4raw-id: //depot/perl@31981
Jerry D. Hedden [Tue, 25 Sep 2007 11:55:53 +0000]
Check rename() under Cygwin
From: "Jerry D. Hedden" <jdhedden@cpan.org>
Message-ID: <
1ff86f510709250855o2c9115cdi6ce830a9ccaec696@mail.gmail.com>
p4raw-id: //depot/perl@31980
Jerry D. Hedden [Tue, 25 Sep 2007 10:45:47 +0000]
No more ld2 and perlld under Cygwin
From: "Jerry D. Hedden" <jdhedden@cpan.org>
Message-ID: <
1ff86f510709250745s1d17ae54y6403711e00b306a0@mail.gmail.com>
p4raw-id: //depot/perl@31979
Robin Barker [Wed, 26 Sep 2007 13:25:49 +0000]
RE: [PATCH] use 5.010 is ugly; use 5.10.0 warns
From: "Robin Barker" <Robin.Barker@npl.co.uk>
Message-ID: <
2C2E01334A940D4792B3E115F95B7226C9D1C3@exchsvr1.npl.ad.local>
p4raw-id: //depot/perl@31978
Nicholas Clark [Wed, 26 Sep 2007 10:21:50 +0000]
Change the way of determining the MRO algorithm used from an enum
defined in the header, to a pointer to a structure. This allows the
flexibility to easily add more MROs in the future, and to provide an
API to do so. Dispatch in mro.c is now via the structure pointed to,
rather than switch statements on the value of the enum.
p4raw-id: //depot/perl@31977
Rafael Garcia-Suarez [Wed, 26 Sep 2007 09:07:59 +0000]
Option spelling nit for AIX
p4raw-id: //depot/perl@31976
Rafael Garcia-Suarez [Wed, 26 Sep 2007 09:00:08 +0000]
On VAX, feeding a bare "NaN" to nawk causes nawk to core dump since
it tries to interpret it as a number, which causes an FP exception.
(Patch from NetBSD)
p4raw-id: //depot/perl@31975
Rafael Garcia-Suarez [Wed, 26 Sep 2007 08:57:41 +0000]
Assume NetBSD has touch
(patch from NetBSD package)
p4raw-id: //depot/perl@31974
Jim Cromie [Tue, 25 Sep 2007 17:34:09 +0000]
Re: [patch] improve B::Concise -src support
Message-ID: <
46F99AF1.1080108@gmail.com>
p4raw-id: //depot/perl@31973
Rafael Garcia-Suarez [Wed, 26 Sep 2007 08:20:14 +0000]
Update to README.cygwin by Jerry D. Hedden
p4raw-id: //depot/perl@31972
Dave Mitchell [Tue, 25 Sep 2007 18:59:02 +0000]
extend PL_veto_cleanup to all platforms
The conditional skip of PERL_SYS_TERM introduced by change #29827
now happens on all platforms, not just UNIX.
p4raw-link: @29827 on //depot/perl:
c301d6064f299b8a77670348b81d25d2d94d6a2f
p4raw-id: //depot/perl@31971
Dave Mitchell [Tue, 25 Sep 2007 18:00:32 +0000]
make PERL_SYS_INIT/INIT3/TERM into functions
p4raw-id: //depot/perl@31970
Rafael Garcia-Suarez [Tue, 25 Sep 2007 14:28:20 +0000]
Hey, but "or" is low-prec. So this statement never worked.
p4raw-id: //depot/perl@31969
Rafael Garcia-Suarez [Tue, 25 Sep 2007 14:24:53 +0000]
DragonflyBSD is a BSD too
p4raw-id: //depot/perl@31968
Slaven Rezic [Wed, 19 Sep 2007 21:56:00 +0000]
Re: [perl #45513] Test failures on amd64-freebsd 6.2
Message-ID: <87abri1lbj.fsf@biokovo-amd64.herceg.de>
p4raw-id: //depot/perl@31967
Rafael Garcia-Suarez [Tue, 25 Sep 2007 12:20:25 +0000]
Upgrade to Test::Simple 0.72
p4raw-id: //depot/perl@31966
Rick Delaney [Sat, 22 Sep 2007 16:30:40 +0000]
Re: [PATCH 5.10] Improve diagnostic for reloads of bad modules
Message-ID: <
20070922203040.GK11931@bort.ca>
p4raw-id: //depot/perl@31965
Rafael Garcia-Suarez [Tue, 25 Sep 2007 09:45:46 +0000]
Make OptreeCheck.pm more forgiving about parent pad indexes
p4raw-id: //depot/perl@31964
Sébastien Aperghis-Tramoni [Tue, 25 Sep 2007 11:55:48 +0000]
Trans.: CPAN Upload: S/SA/SAPER/constant-1.11.tar.gz
Message-ID: <
1190714148.46f8db2431f0c@imp.free.fr>
p4raw-id: //depot/perl@31963
Abe Timmerman [Sun, 23 Sep 2007 01:02:22 +0000]
Re: Smoke [5.10.0] 31938 FAIL(F) netbsd 3.1 (i386/1 cpu)
Message-ID: <
46F59EFE.7050303@ztreet.demon.nl>
p4raw-id: //depot/perl@31962
Tels [Sat, 22 Sep 2007 14:27:29 +0000]
Re: [perl #45605] Regexp failure with utf8-flagged string and byte-flagged pattern
Message-Id: <
200709221427.30425@bloodgate.com>
p4raw-id: //depot/perl@31961
Sébastien Aperghis-Tramoni [Tue, 25 Sep 2007 02:11:46 +0000]
[patch] Dual-lifing constant.pm
Message-Id: <
454B4D1C-48EE-4CA3-834A-A0F872C9E5FC@free.fr>
p4raw-id: //depot/perl@31960
Slaven Rezic [Mon, 24 Sep 2007 23:24:37 +0000]
stop lib/perl5db.t failure if Term::ReadLine::Perl is installed
Message-ID: <871wcnwycq.fsf@biokovo-amd64.herceg.de>
p4raw-id: //depot/perl@31959
Jerry D. Hedden [Mon, 24 Sep 2007 20:58:00 +0000]
README.cygwin update
From: "Jerry D. Hedden" <jdhedden@cpan.org>
Message-ID: <
1ff86f510709241758o4027c237w7387e7ab86680d55@mail.gmail.com>
p4raw-id: //depot/perl@31958
Jerry D. Hedden [Mon, 24 Sep 2007 20:39:44 +0000]
$^E is not Win32::GetLastError under Cygwin
From: "Jerry D. Hedden" <jdhedden@cpan.org>
Message-ID: <
1ff86f510709241739q62b4f2eaycf13cb9956f8fab3@mail.gmail.com>
p4raw-id: //depot/perl@31957
Jerry D. Hedden [Mon, 24 Sep 2007 20:23:49 +0000]
Re: Change 31935: Upgrade to Win32-0.32
From: "Jerry D. Hedden" <jdhedden@cpan.org>
Message-ID: <
1ff86f510709241723o58cc1167m2c44221b77959994@mail.gmail.com>
(Check Win32::GetLastError() rather than $^E for Cygwin, on which $^E
is the last posix error (the same as $!), not the last Win32 error)
p4raw-id: //depot/perl@31956
Jim Cromie [Mon, 24 Sep 2007 00:28:28 +0000]
[patch] improve B::Concise -src support
Message-ID: <
46F7590C.1070101@gmail.com>
p4raw-id: //depot/perl@31955
Dave Mitchell [Mon, 24 Sep 2007 14:18:03 +0000]
sixth attempt to silence S_emulate_eaccess compiler warning
p4raw-id: //depot/perl@31954
Rafael Garcia-Suarez [Mon, 24 Sep 2007 13:19:10 +0000]
Upgrade to Compress::Zlib 2.007
p4raw-id: //depot/perl@31953
Rafael Garcia-Suarez [Mon, 24 Sep 2007 12:50:02 +0000]
Upgrade to threads::shared 1.14
Corrects warnings reported by Devel::PPPort.
p4raw-id: //depot/perl@31952
Tels [Sat, 22 Sep 2007 11:33:34 +0000]
Re: BigInt bug with non-integer accuracy/precision
Message-Id: <
200709221133.35110@bloodgate.com>
p4raw-id: //depot/perl@31951
H.Merijn Brand [Mon, 24 Sep 2007 08:34:14 +0000]
Announce plans for -Duse32bit* options for Configure, and document
that there is no guarantee on 32bit builds when no args passed
p4raw-id: //depot/perl@31950
Craig A. Berry [Mon, 24 Sep 2007 02:51:23 +0000]
Allow IEEE math to be deselected on OpenVMS I64 (but it remains
the default).
p4raw-id: //depot/perl@31949
Dave Mitchell [Mon, 24 Sep 2007 01:20:00 +0000]
silence a spurious compiler warning in regcomp.c
p4raw-id: //depot/perl@31948
Dave Mitchell [Mon, 24 Sep 2007 00:42:07 +0000]
fifth attempt to silence S_emulate_eaccess compiler warning
p4raw-id: //depot/perl@31947
Craig A. Berry [Sun, 23 Sep 2007 23:44:59 +0000]
And some more VMS-specific additions to PL_bincompat_options.
p4raw-id: //depot/perl@31946
Nicholas Clark [Sun, 23 Sep 2007 21:10:27 +0000]
Add USE_IEEE to PL_bincompat_options, as use_ieee in the configuration
on VMS affects binary compatibility.
p4raw-id: //depot/perl@31945
Craig A. Berry [Sun, 23 Sep 2007 15:09:51 +0000]
Record IEEE usage in config.h on VMS.
p4raw-id: //depot/perl@31944
H.Merijn Brand [Sun, 23 Sep 2007 13:57:00 +0000]
Revert #31942, as it was inconsistent between threaded and
non-threaded builds
p4raw-id: //depot/perl@31943
Abe Timmerman [Sun, 23 Sep 2007 01:11:50 +0000]
netbsd's setlocale () doesn't support LC_COLLATE and LC_NUMERIC
Subject: Re: Smoke [5.10.0] 31938 FAIL(F) netbsd 3.1 (i386/1 cpu)
Message-ID: <
46F5A136.8080000@ztreet.demon.nl>
p4raw-id: //depot/perl@31942
Nicholas Clark [Sun, 23 Sep 2007 07:20:49 +0000]
Under -DDEBUGGING (on gcc), assert that SvRV(sv) is being called on
a valid SvTYPE().
p4raw-id: //depot/perl@31941
Nicholas Clark [Sat, 22 Sep 2007 15:46:44 +0000]
Fix bug 45607 - for the corner case *{"BONK"} = \&{"BONK"} the order
of op evaluation means that what had been a reference to a constant
can turn into a typeglob before the sassign gets to run.
p4raw-id: //depot/perl@31940
Marcus Holland-Moritz [Sat, 22 Sep 2007 08:01:59 +0000]
Upgrade to Devel::PPPort 3.12
p4raw-id: //depot/perl@31939
Nicholas Clark [Fri, 21 Sep 2007 09:16:37 +0000]
hv_stores() on a literal string is now fractionally more efficient than
hv_store().
p4raw-id: //depot/perl@31938
Steve Hay [Fri, 21 Sep 2007 08:25:47 +0000]
Fix Win32 breakage caused by #31926
p4raw-id: //depot/perl@31937
Rafael Garcia-Suarez [Fri, 21 Sep 2007 07:54:47 +0000]
Silence a load of "value computed is not used" warnings
p4raw-id: //depot/perl@31936
Steve Hay [Fri, 21 Sep 2007 07:39:02 +0000]
Upgrade to Win32-0.32
p4raw-id: //depot/perl@31935
John Peacock [Thu, 20 Sep 2007 21:15:51 +0000]
version-0.73 (was Re: Change 31920: Don't use ~0 as a version
Message-ID: <
46F31B47.6030601@cpan.org>
p4raw-id: //depot/perl@31934
Nicholas Clark [Thu, 20 Sep 2007 19:34:51 +0000]
Get mad compiling as C++. (At least for me)
p4raw-id: //depot/perl@31933
Nicholas Clark [Thu, 20 Sep 2007 16:52:06 +0000]
Remove now-unnecessary =cuts where POD is now adjacent to POD.
p4raw-id: //depot/perl@31932
Nicholas Clark [Thu, 20 Sep 2007 16:44:24 +0000]
Add a new function Perl_hv_common_key_len(), which contains the
manipulations to convert negative lengths to positive length + UTF-8
flag. hv_delete(), hv_exists(), hv_fetch(), hv_store() and
hv_store_flags() all become mathoms. The macros hv_fetchs() and
hv_stores() call hv_common() directly.
p4raw-id: //depot/perl@31931
Nicholas Clark [Thu, 20 Sep 2007 15:11:01 +0000]
Move the SV dereference of Perl_hv_fetch()/Perl_hv_store()/
Perl_hv_store_flags() into Perl_hv_common(). All already pass in
HV_FETCH_JUST_SV, which only these three pass in.
p4raw-id: //depot/perl@31930
Steve Peters [Thu, 20 Sep 2007 14:00:07 +0000]
ext/XS/APITest/APITest.xs was suffering casting issues with g++
as well after the addition of Perl_hv_common().
p4raw-id: //depot/perl@31929
Jos I. Boumans [Thu, 20 Sep 2007 15:46:24 +0000]
Archive::Extract 0.24 (was Re: Archive::Extract test failures on Solaris)
From: "Jos I. Boumans" <kane@dwim.org>
Message-Id: <
174BF3BE-B640-4CCA-AB47-BE527382B236@dwim.org>
p4raw-id: //depot/perl@31928
Steve Peters [Thu, 20 Sep 2007 13:32:18 +0000]
Casts needed in mathoms.c to make it compile with g++ after the
making Perl_hv_common() return void *.
p4raw-id: //depot/perl@31927
Nicholas Clark [Thu, 20 Sep 2007 12:59:25 +0000]
Switch Perl_hv_common() to returning void * rather than HE *.
p4raw-id: //depot/perl@31926
Nicholas Clark [Thu, 20 Sep 2007 12:20:59 +0000]
Change 31919 forgot to add Perl_hv_common to global.sym
p4raw-id: //depot/perl@31925
Nicholas Clark [Thu, 20 Sep 2007 12:13:16 +0000]
Convert hv_delete_ent(), hv_exists_ent(), hv_fetch_ent() and
hv_store_ent() to macros, and consign the function bodies to history.
Er, mathoms.c
p4raw-id: //depot/perl@31924
Nicholas Clark [Thu, 20 Sep 2007 10:21:30 +0000]
Use Perl_hv_common() to test disabling the key conversion in hash
lookups.
p4raw-id: //depot/perl@31923
Nicholas Clark [Thu, 20 Sep 2007 10:20:47 +0000]
assert that what is passed into the hash functions is really an HV.
(MRO code is calling hash functions during global destruction, hence
the check on SVTYPEMASK.)
p4raw-id: //depot/perl@31922
Nicholas Clark [Thu, 20 Sep 2007 09:42:50 +0000]
SVs know their length, so avoid 2 calls to strlen().
As we have an SV, call hv_fetch_ent() rather than hv_fetch().
p4raw-id: //depot/perl@31921
Rafael Garcia-Suarez [Thu, 20 Sep 2007 08:59:34 +0000]
Don't use ~0 as a version number -- it could overflow.
p4raw-id: //depot/perl@31920
Nicholas Clark [Thu, 20 Sep 2007 08:39:28 +0000]
Make hv_fetch_common() non-static, and change its name to hv_common(),
as it now also performs hv_delete()/hv_delete_ent()
p4raw-id: //depot/perl@31919
Nicholas Clark [Wed, 19 Sep 2007 22:55:27 +0000]
Change 31849 forgot to regenerate the documentation - removing a
documented function from the published API moves the POD file in which
its documentation resides.
p4raw-id: //depot/perl@31918
Nicholas Clark [Wed, 19 Sep 2007 22:52:53 +0000]
Inline and abolish S_hv_magic_uvar_xkey().
p4raw-id: //depot/perl@31917
Nicholas Clark [Wed, 19 Sep 2007 22:04:46 +0000]
Send all delete()/delete_ent() calls via S_hv_fetch_common().
This puts all the key normalisation code in one place.
p4raw-id: //depot/perl@31916
projects / p5sagit/p5-mst-13.2.git / log