Jarkko Hietaniemi [Tue, 16 Oct 2001 17:07:20 +0000]
Can't expect too much accuracy from a sleep of one second.
(I think this will also fail with high load.)
p4raw-id: //depot/perl@12458
Randolf Werner [Tue, 16 Oct 2001 18:33:02 +0000]
WIN64 (Itanium) port of Perl 5.6.1
Message-ID: <
816D93CCC927D31188570008C75D1DE10A5CB1F8@dbwdfx1a.wdf.sap-ag.de>
p4raw-id: //depot/perl@12457
Jarkko Hietaniemi [Tue, 16 Oct 2001 13:59:32 +0000]
I think the Perl_sortsv needs *not* be skipped?
p4raw-id: //depot/perl@12455
Gerrit P. Haase [Tue, 16 Oct 2001 15:32:44 +0000]
[ PATCH ] maint/perl/t/op/groups.t for build on Cygwin
Message-ID: <
3BCC531C.11494.3E433F@localhost>
p4raw-id: //depot/perl@12454
Gerrit P. Haase [Tue, 16 Oct 2001 15:01:04 +0000]
[ PATCH ] maint/perl/t/io/fs.t for build on Cygwin
Message-ID: <
3BCC4BB0.30935.2145B7@localhost>
p4raw-id: //depot/perl@12453
Jarkko Hietaniemi [Tue, 16 Oct 2001 12:03:49 +0000]
A forgotten deletion.
p4raw-id: //depot/perl@12452
Jarkko Hietaniemi [Tue, 16 Oct 2001 02:27:51 +0000]
Document the problem with the swash_fetch() API that affects
more complex case conversions.
p4raw-id: //depot/perl@12450
Jarkko Hietaniemi [Tue, 16 Oct 2001 01:55:48 +0000]
Make Time::HiRes::sleep() and usleep() to return
the number of seconds (a floating point value) and
microseconds (an integer) actually slept (well,
modulo the time spent in measuring the time actually
slept...), to be a better drop-in replacement for
the builtin sleep(). Noticed by Chris Nandor.
Also make usleep() to croak if fed illegal number
of useconds (anything greater than 1_000_000).
p4raw-id: //depot/perl@12449
Jarkko Hietaniemi [Tue, 16 Oct 2001 00:57:05 +0000]
A missing SvREFCNT_dec.
p4raw-id: //depot/perl@12448
Jarkko Hietaniemi [Mon, 15 Oct 2001 18:44:19 +0000]
(retracted by #12497)
In AIX vac 5 just can't seem to get regcomp right.
p4raw-id: //depot/perl@12446
Jarkko Hietaniemi [Mon, 15 Oct 2001 16:44:04 +0000]
Bump the version number to agree with the maintperl one.
p4raw-id: //depot/perl@12445
Jarkko Hietaniemi [Mon, 15 Oct 2001 13:40:20 +0000]
Casting tweak from Paul Marquess, helps Merijn on AIX.
p4raw-id: //depot/perl@12443
Jarkko Hietaniemi [Mon, 15 Oct 2001 13:35:36 +0000]
Sync the skip() amounts with the number of tests.
p4raw-id: //depot/perl@12442
Jarkko Hietaniemi [Mon, 15 Oct 2001 13:11:03 +0000]
Update Changes.
p4raw-id: //depot/perl@12441
Jarkko Hietaniemi [Mon, 15 Oct 2001 12:58:24 +0000]
MacOS Classic catdir() rewrite from Thomas Wegner
(backward incompatibility, but a deliberate one,
the old version simply is broken in its logic),
also documentation updates, and as suggested replicated
the File::Spec::Unix documentation updates also on the
File::Spec documentation.
TODO: there seems to be duplication of documentation
between File::Spec and File::Spec::Unix. I think
the ::Unix should be left only with specific UNIXisms,
and all the generic documentation should be in ::Spec.
p4raw-id: //depot/perl@12440
Blair Zajac [Mon, 24 Sep 2001 11:48:58 +0000]
Re: [PATCH t/op/magic.t] missing tests on Win32
Message-ID: <
3BAF801A.88A4F614@orcaware.com>
(the skip() messages)
p4raw-id: //depot/perl@12439
Craig A. Berry [Sun, 14 Oct 2001 15:27:18 +0000]
make PerlIO default on VMS
Message-Id: <5.1.0.14.0.
20011014152339.021ec150@exchi01>
p4raw-id: //depot/perl@12438
Jarkko Hietaniemi [Sun, 14 Oct 2001 23:05:57 +0000]
Retract #12436 (Abhijit already did this at #12426)
p4raw-id: //depot/perl@12437
Kay Röpke [Sat, 13 Oct 2001 20:33:55 +0000]
(retracted by #12437)
Subject: [PATCH lib/ExtUtils/MM_Unix.pm] MakeMaker tries to parse commented out $VERSION
Message-Id: <E15sTbe-0001nt-00@mrvdom04.kundenserver.de>
p4raw-id: //depot/perl@12436
Yitzchak Scott-Thoennes [Fri, 12 Oct 2001 00:12:01 +0000]
h2xs grammar nit
Message-ID: <Pine.GSU.4.21.
0110120010540.9710-100000@garcia.efn.org>
p4raw-id: //depot/perl@12435
Paul Marquess [Sun, 14 Oct 2001 11:25:08 +0000]
Fix scoping problem with FATAL warnings
Message-ID: <AIEAJICLCBDNAAOLLOKL
CEFDDCAA.Paul_Marquess@Yahoo.co.uk>
p4raw-id: //depot/perl@12434
Jarkko Hietaniemi [Sat, 13 Oct 2001 23:20:11 +0000]
Use File::Spec (needs Thomas Wegner's upcoming patch
for File::Spec::Mac::catfile)
p4raw-id: //depot/perl@12433
Nicholas Clark [Sat, 13 Oct 2001 00:07:31 +0000]
Re: default module version for h2xs (was Re: [PATCH] Re: What sort of Makefile.PL should h2xs write?)
Message-ID: <
20011013000731.D67535@plum.flirble.org>
p4raw-id: //depot/perl@12432
Jarkko Hietaniemi [Sat, 13 Oct 2001 22:35:58 +0000]
Define PASTHRU_DEFINE and PASTHRU_INC (which are used
before DEFINE and INC), and repent by taking DEFINE and
INC away from PASTHRU.
p4raw-id: //depot/perl@12431
Jarkko Hietaniemi [Sat, 13 Oct 2001 21:23:09 +0000]
More MANIFEST anomalies.
p4raw-id: //depot/perl@12430
Jarkko Hietaniemi [Sat, 13 Oct 2001 21:01:32 +0000]
Not that many.
p4raw-id: //depot/perl@12429
Rafael Garcia-Suarez [Sat, 13 Oct 2001 23:15:55 +0000]
[DOC PATCH] pod syntax fixups for File::Spec::* modules
Message-ID: <
20011013231555.A18071@rafael>
p4raw-id: //depot/perl@12428
Jarkko Hietaniemi [Sat, 13 Oct 2001 20:54:17 +0000]
Rewrite mktables from scratch.
- Cleaner.
- Faster: 15-20 seconds as opposed to several minutes.
- More dynamic: the names of the various categories
such as the linebreak ones are dynamic, not static.
- Is.pl: long names for the general category properties
are now available.
- Ranges (<... ,First>, <..., Last>) from the general
categories work now.
- No more mktables.PL because the mktables.PL is not
and never has been run to create a mktables.
- syllables.txt and Is/Syl*.pl removed: non-standard
(not part of the Unicode), and the whole concept is
being reworked (http://syllabary.sourceforge.net/),
the old way wouldn't even work with the new Syllables.txt
(it would result in 1000+ new categories)
p4raw-id: //depot/perl@12427
Abhijit Menon-Sen [Sat, 13 Oct 2001 18:05:36 +0000]
ExtUtils::MM_Unix should ignore comments when searching for
$VERSION.
p4raw-id: //depot/perl@12426
Andreas König [Sat, 13 Oct 2001 10:17:09 +0000]
Re: New module XML::Clean
Message-ID: <m37ku0t0my.fsf@anima.de>
p4raw-id: //depot/perl@12425
Artur Bergman [Sat, 13 Oct 2001 09:39:59 +0000]
Set thread stack size if needed.
p4raw-id: //depot/perl@12424
Artur Bergman [Sat, 13 Oct 2001 09:36:49 +0000]
We do want to set the attribute to joinable even on modern posix
platforms.
p4raw-id: //depot/perl@12423
Brian Ingerson [Fri, 12 Oct 2001 19:32:35 +0000]
Message-ID: <
20011012193235.B889@ttul.org>
p4raw-id: //depot/perl@12422
Jarkko Hietaniemi [Fri, 12 Oct 2001 23:22:45 +0000]
Update Changes.
p4raw-id: //depot/perl@12421
Jarkko Hietaniemi [Fri, 12 Oct 2001 22:51:17 +0000]
FAQ sync.
p4raw-id: //depot/perl@12420
Wilson, Doug [Fri, 12 Oct 2001 11:51:05 +0000]
RE: [PATCH File::Copy] Silence warnings for tied filehandles
Message-ID: <
35A280DF784CD411A06B0008C7B130AD0115DDAE@sdex04.sd.intuit.com>
p4raw-id: //depot/perl@12419
Jarkko Hietaniemi [Fri, 12 Oct 2001 22:44:24 +0000]
[PATCH lib/Term/Complete.t] Rethinking the Test
From: "chromatic" <chromatic@rmci.net>
Date: Thu, 11 Oct 2001 10:57:55 -0600
Message-ID: <
20011011170354.74354.qmail@onion.perl.org>
Subject: [PATCH Complete.pm] Re: [PATCH lib/Term/Complete.t] Rethinking the Test
From: Rafael Garcia-Suarez <rgarciasuarez@free.fr>
Date: Thu, 11 Oct 2001 22:34:21 +0200
Message-ID: <
20011011223421.A693@rafael>
plus undef $Term::Complete::stty as suggested by Rafael.
p4raw-id: //depot/perl@12418
Jarkko Hietaniemi [Fri, 12 Oct 2001 20:35:49 +0000]
UUnndduupplleexx..
p4raw-id: //depot/perl@12417
Andy Dougherty [Fri, 12 Oct 2001 16:53:03 +0000]
Re: Sparc/Linux/ithreads unhappy @12391 [PATCH]
Message-ID: <Pine.SOL.4.10.
10110121647360.11279-100000@maxwell.phys.lafayette.edu>
(Potentially only band-aid)
p4raw-id: //depot/perl@12416
Jarkko Hietaniemi [Fri, 12 Oct 2001 20:13:06 +0000]
VMS pre-7.0 (which doesn't have sigaction()) uniformly
mishandles kill(), and therefore the sigaction test
should be skipped, from Charles Lane.
p4raw-id: //depot/perl@12415
Artur Bergman [Fri, 12 Oct 2001 18:47:49 +0000]
Perhaps other OLD_POSIX_API systems might be happy and not only HP-UX 10.20
p4raw-id: //depot/perl@12414
Jarkko Hietaniemi [Fri, 12 Oct 2001 18:35:31 +0000]
Upgrade to Math::BigInt 1.44 from Tels and
further fixes from John Peacock.
p4raw-id: //depot/perl@12413
Artur Bergman [Fri, 12 Oct 2001 18:28:23 +0000]
perhaps pthread_attr_t really needs to be set for the old api
p4raw-id: //depot/perl@12412
Artur Bergman [Fri, 12 Oct 2001 18:11:45 +0000]
pthread_keycreate not pthread_key_create under the here be dragons API
p4raw-id: //depot/perl@12411
Artur Bergman [Fri, 12 Oct 2001 18:07:37 +0000]
Apperently OLD PTHREADS API is a bit retarded.
This should fix another set of compile failures in HP-UX 10.20.
p4raw-id: //depot/perl@12410
Artur Bergman [Fri, 12 Oct 2001 14:55:08 +0000]
Let us avoid being smart for now.
p4raw-id: //depot/perl@12408
Artur Bergman [Fri, 12 Oct 2001 14:38:12 +0000]
(void*) the argument to make sure we work with picky compilers
p4raw-id: //depot/perl@12407
Mikhail Zabaluev [Fri, 12 Oct 2001 12:47:47 +0000]
(replaced with #12600)
Subject: [PATCH 5.6.1 perldoc] use File::Temp
Message-Id: <
20011012124747.E13918@localhost.localdomain>
p4raw-id: //depot/perl@12406
Andreas König [Fri, 12 Oct 2001 11:12:07 +0000]
(Was: lib/Memoize/t/speed.................FAILED at test 2)
Message-Id: <m3wv21te6w.fsf_-_@anima.de>
p4raw-id: //depot/perl@12405
Artur Bergman [Fri, 12 Oct 2001 12:46:15 +0000]
A) Support OLD_PTHREADS_API
B) Change from using pthread_t to using a TLS to store the
index of the thread in an hash, this is to avoid problems when
in fact pthread_t was not a seralizable type.
Both these changes are because of HP-UX 10.20
This has not been tested on win32 but should work there.
Need to add support for NetWare.
p4raw-id: //depot/perl@12404
Andreas König [Fri, 12 Oct 2001 10:20:35 +0000]
Re: New module XML::Clean
Message-Id: <m3elo9uv58.fsf@anima.de>
p4raw-id: //depot/perl@12403
Jarkko Hietaniemi [Thu, 11 Oct 2001 19:51:04 +0000]
More HP model tweaks from Merijn.
p4raw-id: //depot/perl@12402
Jarkko Hietaniemi [Thu, 11 Oct 2001 13:03:10 +0000]
Update Changes.
p4raw-id: //depot/perl@12401
Jarkko Hietaniemi [Thu, 11 Oct 2001 11:33:20 +0000]
HP-UX update from H.Merijn Brand.
p4raw-id: //depot/perl@12400
Craig A. Berry [Wed, 10 Oct 2001 17:17:00 +0000]
fix link rot in README.vms
Message-Id: <5.1.0.14.0.
20011010171057.01bd77c8@exchi01>
p4raw-id: //depot/perl@12399
chromatic [Wed, 10 Oct 2001 15:50:30 +0000]
(was Re: lib/Term/Complete.t won't complete in background)
Message-ID: <
20011010215627.13283.qmail@onion.perl.org>
p4raw-id: //depot/perl@12398
chromatic [Wed, 10 Oct 2001 15:45:42 +0000]
Add Tests for ExtUtils::Packlist
Message-ID: <
20011010215140.8913.qmail@onion.perl.org>
p4raw-id: //depot/perl@12397
Jarkko Hietaniemi [Wed, 10 Oct 2001 20:20:22 +0000]
VMS needs quotes around the -I option to keep it from
being downcased, from Charles Lane.
p4raw-id: //depot/perl@12396
Stas Bekman [Tue, 9 Oct 2001 23:57:10 +0000]
Message-Id: <
3BC31E56.60805@stason.org>
p4raw-id: //depot/perl@12395
Abhijit Menon-Sen [Wed, 10 Oct 2001 21:59:33 +0000]
Re: /usr/include/sys/e ?? (perl@12340)
Message-Id: <
20011010215933.B2444@lustre.dyn.wiw.org>
p4raw-id: //depot/perl@12394
Jarkko Hietaniemi [Wed, 10 Oct 2001 15:53:39 +0000]
VMS: reorder the elimination of 000000 in the canonpath logic,
from Charles Lane.
p4raw-id: //depot/perl@12393
Rafael Garcia-Suarez [Wed, 10 Oct 2001 17:23:44 +0000]
two s/// bugfixes
Message-Id: <
20011010172344.C22440@rafael>
p4raw-id: //depot/perl@12392
Jarkko Hietaniemi [Wed, 10 Oct 2001 12:22:02 +0000]
Fix/improve handling of the [000000] special "root"
directory, by putting cwd in canonical form when doing
abs2rel, and prepending a 000000 to directory "chunks"
if missing from base or path. Also adjust test to
reflect the (improved) output from abs2rel.
From Charles Lane.
p4raw-id: //depot/perl@12391
Jarkko Hietaniemi [Wed, 10 Oct 2001 11:48:14 +0000]
In VMS the copy-to-untaint %ENV in the initialization also loses
all of the extra VMSish attributes (such as "treat this logical
as a device name") that are necessary to make programs run.
A tainted %ENV does not hinder VMS. From Charles Lane.
p4raw-id: //depot/perl@12390
Jarkko Hietaniemi [Wed, 10 Oct 2001 11:40:29 +0000]
VMS-specific fixes to the ExtUtils::Manifest tests.
o VMS is case-insensitive, RTL downcases filenames read
by perl... so need to re-upcase filenames like MANIFEST
and MANIFEST.SKIP
o Need to convert VMS-style to Unix-style filenames when comparing
names in MANIFEST to names in filesystem.
o Need to remove mantest/MANIFEST before rmdir'ing mantest,
since VMS won't let you (easily) delete non-empty directories.
From Charles Lane.
p4raw-id: //depot/perl@12389
Jarkko Hietaniemi [Wed, 10 Oct 2001 11:28:44 +0000]
Unix and VMS agree on "*" as a "match any number of any chars"
wildcards, but differ ("?" vs. "%") for their "match any single
char" wildcard. This patch changes "?" chars to "%" before doing a
VMS glob...at no loss of functionality, since "?" isn't a valid
filename char on VMS. From Charles Lane.
p4raw-id: //depot/perl@12388
Jarkko Hietaniemi [Wed, 10 Oct 2001 11:20:13 +0000]
(accidentally empty submit; new life as #12390)
p4raw-id: //depot/perl@12387
Jarkko Hietaniemi [Wed, 10 Oct 2001 11:12:40 +0000]
Even more VMS tweakage from Charles Lane:
Tweak #1: splitdir only works on directory part of filespec,
split off volume first
Tweak #2: VMS pre-7.0's $ENV{'HOME'} is not settable
(it's a horrible kludge in the run-time library,
improved in later versions). Check the
$Config{'d_setenv'} flag before attempting to set it.
p4raw-id: //depot/perl@12386
Jarkko Hietaniemi [Tue, 9 Oct 2001 20:34:36 +0000]
VMS tweakage from Charles Lane.
. command.com doubles an output line when prompting for extensions
. Term::Cap has no business trying to run obscure Un*x utilities on VMS
. perl5db doesn't clean up after itself
p4raw-id: //depot/perl@12385
Rafael Garcia-Suarez [Tue, 9 Oct 2001 22:10:26 +0000]
, Re: [PATCH again] Re: [PATCH mg.c gv.c and others] ${^TAINT}
Message-ID: <
20011009221026.A8579@rafael>
p4raw-id: //depot/perl@12384
Rafael Garcia-Suarez [Tue, 9 Oct 2001 15:15:06 +0000]
new tests, new TODO test
Message-Id: <slrn9s653h.h2p.rgarciasuarez@rafael.kazibao.net>
p4raw-id: //depot/perl@12383
Jarkko Hietaniemi [Tue, 9 Oct 2001 14:03:30 +0000]
Update Changes.
p4raw-id: //depot/perl@12382
Jarkko Hietaniemi [Tue, 9 Oct 2001 13:52:50 +0000]
Seems that a glob(undef) triggers an ACCVIO in Perl_flex_stat
Seems that a glob(undef) triggers an ACCVIO in Perl_flex_stat,
when a NULL is passed for the filename string pointer.
From Charles Lane.
p4raw-id: //depot/perl@12381
H.Merijn Brand [Tue, 9 Oct 2001 15:59:22 +0000]
HP systems update
Message-Id: <
20011009155614.A4F5.H.M.BRAND@hccnet.nl>
p4raw-id: //depot/perl@12379
Jarkko Hietaniemi [Tue, 9 Oct 2001 12:52:21 +0000]
CPAN::Nox does now have a test.
p4raw-id: //depot/perl@12378
David Dyck [Mon, 1 Oct 2001 12:17:36 +0000]
[ID
20011001.009] /usr/bin/perl (in addition to /usr/local/bin/perl)
Message-Id: <
200110011917.MAA11879@dd.tc.fluke.com>
p4raw-id: //depot/perl@12377
Jarkko Hietaniemi [Tue, 9 Oct 2001 12:18:39 +0000]
NetWare update from Ananth Kesari.
p4raw-id: //depot/perl@12376
Charles Lane [Tue, 9 Oct 2001 08:40:57 +0000]
File::Find test on VMS
Message-Id: <
011009083426.17530f@DUPHY4.Physics.Drexel.Edu>
p4raw-id: //depot/perl@12375
chromatic [Tue, 9 Oct 2001 00:08:37 +0000]
Add Tests for
ExtUtils::Mkbootstrap
Message-Id: <
20011009061432.46666.qmail@onion.perl.org>
p4raw-id: //depot/perl@12374
Jarkko Hietaniemi [Tue, 9 Oct 2001 02:56:13 +0000]
Enable more debugging.
p4raw-id: //depot/perl@12373
Abhijit Menon-Sen [Tue, 9 Oct 2001 02:11:59 +0000]
#12370 forgot to modify is_gv_magical().
p4raw-id: //depot/perl@12372
Jarkko Hietaniemi [Tue, 9 Oct 2001 01:48:17 +0000]
Make the toupper/lower/title API for Unicode not right
but at least less wrong: prepare for the mapping being
more than just one-character-to-one-character.
p4raw-id: //depot/perl@12371
Ronald J. Kimball [Mon, 8 Oct 2001 22:18:23 +0000]
Re: [PATCH mg.c gv.c and others] ${^TAINT}
Message-ID: <
20011008221823.A413700@linguist.thayer.dartmouth.edu>
p4raw-id: //depot/perl@12370
Jarkko Hietaniemi [Mon, 8 Oct 2001 23:31:21 +0000]
FAQ sync.
p4raw-id: //depot/perl@12369
chromatic [Mon, 8 Oct 2001 14:05:32 +0000]
Fix Debian Failures on Test 22
Message-ID: <
20011008201126.73702.qmail@onion.perl.org>
p4raw-id: //depot/perl@12368
Michael G. Schwern [Mon, 8 Oct 2001 17:00:14 +0000]
Re: [PATCH mg.c gv.c and others] ${^TAINT}
Message-ID: <
20011008170014.L17083@blackrider>
p4raw-id: //depot/perl@12367
Charles Lane [Mon, 8 Oct 2001 16:01:33 +0000]
Time::Hires for VMS pre-7.0
Message-Id: <
011008155856.1604b5@DUPHY4.Physics.Drexel.Edu>
p4raw-id: //depot/perl@12366
Craig A. Berry [Mon, 8 Oct 2001 15:40:43 +0000]
Autosplit patch for VMS
Message-Id: <5.1.0.14.0.
20011008150808.02302618@exchi01>
p4raw-id: //depot/perl@12365
Jarkko Hietaniemi [Mon, 8 Oct 2001 20:36:58 +0000]
Don't assume that everybody is running OS X 10.1 yet;
from Kay Röpke.
p4raw-id: //depot/perl@12364
Jarkko Hietaniemi [Mon, 8 Oct 2001 14:26:35 +0000]
Update Changes.
p4raw-id: //depot/perl@12363
Jarkko Hietaniemi [Mon, 8 Oct 2001 14:05:50 +0000]
Retract #12358 for now.
p4raw-id: //depot/perl@12362
Jarkko Hietaniemi [Mon, 8 Oct 2001 13:57:10 +0000]
Can't figure out what this is testing.
(If it is trying to catch SIGKILL, it should stop
trying because that is by definition impossible.)
p4raw-id: //depot/perl@12361
Nikola Knezevic [Sun, 30 Sep 2001 21:15:13 +0000]
why some tests fail on Win95...
Message-ID: <
5033457805.20010930211513@tesla.rcub.bg.ac.yu>
p4raw-id: //depot/perl@12360
Jarkko Hietaniemi [Mon, 8 Oct 2001 12:37:17 +0000]
Clarify the ldlibpth part and condense the overall text.
p4raw-id: //depot/perl@12359
chromatic [Sun, 7 Oct 2001 16:59:26 +0000]
Re: [PATCH] & what's to be done for 5.8.0?
Message-ID: <
20011007230518.90917.qmail@onion.perl.org>
p4raw-id: //depot/perl@12358
Jarkko Hietaniemi [Mon, 8 Oct 2001 11:54:50 +0000]
Update the list because fo #12265.
p4raw-id: //depot/perl@12357
Nicholas Clark [Sun, 7 Oct 2001 20:53:33 +0000]
& what's to be done for 5.8.0?
Message-ID: <
20011007205333.S38756@plum.flirble.org>
p4raw-id: //depot/perl@12356
Jarkko Hietaniemi [Mon, 8 Oct 2001 11:51:29 +0000]
Integrate patch #12350 from maintperl; typo in
IO::Seekable doc.
p4raw-id: //depot/perl@12355
p4raw-integrated: from //depot/maint-5.6/perl@12354 'ignore'
ext/IO/lib/IO/Seekable.pm (@8175..)
Jarkko Hietaniemi [Mon, 8 Oct 2001 11:41:16 +0000]
Detypo.
p4raw-id: //depot/perl@12354
Artur Bergman [Sun, 7 Oct 2001 17:16:14 +0000]
Change 12349 didn't really get it right.
p4raw-id: //depot/perl@12353