develooper Front page | perl.perl5.porters | Postings from November 2000


Thread Next
Jarkko Hietaniemi
November 29, 2000 10:06
Message ID:

Changes since the last:
[  7929] By: jhi                                   on 2000/11/29  18:50:55
        Log: Update the va_copy() metaconfig unit.
     Branch: metaconfig/U/perl
           ! need_va_copy.U
[  7928] By: jhi                                   on 2000/11/29  17:50:43
        Log: Subject: [ID 20001128.003] Not OK: perl v5.7.0 +DEVEL7891 on ppc-linux 2.2.17  
             From: Dominic Dunlop <>
             Date: Tue, 28 Nov 2000 16:41:57 +0100
             Message-Id: <p04320400b649771b9797@[]>
             Patch to catch a core dump in the Configure va_copy test.
     Branch: perl
	   ! Configure config_h.SH
[  7927] By: jhi                                   on 2000/11/29  17:31:59
        Log: Retract #7921, the patch shouldn't be needed.
     Branch: perl
	   ! sv.c
[  7926] By: jhi                                   on 2000/11/29  17:01:41
        Log: Make the Class::Struct import() wiser.
     Branch: perl
	   ! lib/Class/ t/lib/class-struct.t
[  7925] By: jhi                                   on 2000/11/29  16:36:33
        Log: All the core library users of Class::Struct seem to be
             using "use Class::Struct 'struct';" instead of the bare
             "use Class::Struct;", which isn't documented in Class::Struct.
             This can't be right.
     Branch: perl
	   ! lib/File/
[  7924] By: jhi                                   on 2000/11/29  16:32:15
        Log: A test works better is it has the right 1..$n output.
     Branch: perl
	   ! t/lib/class-struct.t
[  7923] By: jhi                                   on 2000/11/29  16:29:11
        Log: Forgot to MANIFEST the new Net::hostent test.
     Branch: perl
[  7922] By: jhi                                   on 2000/11/29  16:27:49
        Log: Integrate perlio.
     Branch: perl
	  !> iperlsys.h perlio.c
[  7921] By: jhi                                   on 2000/11/29  16:27:10
        Log: (Retracted by #7927.)
             Subject: [PATCH] Is infinity a number?
             From: Simon Cozens <>
             Date: Wed, 29 Nov 2000 14:48:20 +0000
             Message-ID: <>
     Branch: perl
	   ! sv.c
[  7920] By: jhi                                   on 2000/11/29  16:23:02
        Log: Subject: net_hostent.t (was Re: [ID 20001128.002] what's the point of example code if it is buggy?)
             From: (Robert Spier)
             Date: Wed, 29 Nov 2000 02:17:49 -0500
             Message-ID: <14884.44445.760322.278647@localhost.localdomain>
             Add a test for Net::hostent.
     Branch: perl
	   + t/lib/net-hostent.t
[  7919] By: jhi                                   on 2000/11/29  16:21:31
        Log: Make "use Class::Struct 'struct';" work again (broken by #7617);
             add a test for Class::Struct.
     Branch: perl
	   + t/lib/class-struct.t
	   ! MANIFEST lib/Class/
[  7918] By: jhi                                   on 2000/11/29  15:48:40
        Log: Subject: Re: Minor suggestion for Sys::Syslog  [PATCH]
             From: (Robert Spier)
             Date: Wed, 29 Nov 2000 02:25:39 -0500
             Message-ID: <14884.44915.86028.422238@localhost.localdomain>
             More checking in case someone has broken their services or
             protocol databases.
     Branch: perl
	   ! ext/Sys/Syslog/
[  7917] By: jhi                                   on 2000/11/29  15:46:25
        Log: Subject: [PATCH] $^O  win32 -> MSWin32
             From: (Robert Spier)
             Date: Wed, 29 Nov 2000 02:17:38 -0500
             Message-ID: <14884.44434.340627.126145@localhost.localdomain>
             plus similar nits for vms, err, VMS, and UNICOS.
     Branch: perl
	   ! t/lib/syslfs.t t/op/lfs.t
[  7916] By: jhi                                   on 2000/11/29  15:42:51
        Log: Subject: [PATCH] Tokeniser debugging
             From: Simon Cozens <>
             Date: Wed, 29 Nov 2000 14:15:45 +0000
             Message-ID: <>
     Branch: perl
	   ! perl.c perl.h pod/perlrun.pod toke.c
[  7915] By: jhi                                   on 2000/11/29  15:41:36
        Log: Subject: Re: [PATCH 5.7.0] OUT keyword for xsubpp
             From: Ilya Zakharevich <>
             Date: Wed, 29 Nov 2000 02:13:14 -0500
             Message-ID: <>
             OUT keyword nits.
             Subject: Re: [PATCH 5.7.0] OUT keyword for xsubpp
             From: Ilya Zakharevich <>
             Date: Wed, 29 Nov 2000 03:09:04 -0500
             Message-ID: <>
             OUT and IN_OUT documentation.
     Branch: perl
	   ! lib/ExtUtils/xsubpp pod/perlxs.pod
[  7914] By: jhi                                   on 2000/11/29  03:35:01
        Log: Subject: Re: [ID 20001127.002] const subs hurt under debugger
             From: John Tobey <>
             Date: Tue, 28 Nov 2000 01:13:41 -0500 (EST)
             Message-Id: <m140e1N-000FOoC@feynman.localnet>
             plus a performance speedup noticed by Sarathy.
     Branch: perl
	   ! op.c
[  7913] By: jhi                                   on 2000/11/29  03:16:59
        Log: Use "1 while unlink" so that VMS gets clean, too.
     Branch: perl
	   ! t/lib/filter-util.t
[  7912] By: jhi                                   on 2000/11/29  03:09:08
        Log: Subject: [PATCH 5.7.0] OUT keyword for xsubpp
             From: Ilya Zakharevich <>
             Date: Tue, 28 Nov 2000 03:27:09 -0500
             Message-ID: <>
     Branch: perl
	   ! lib/ExtUtils/xsubpp
[  7911] By: jhi                                   on 2000/11/29  03:06:30
        Log: This should have been part of #7872: no need to scan UTF-8
             until eternity.
     Branch: perl
	   ! utf8.c
[  7910] By: jhi                                   on 2000/11/29  02:50:04
        Log: One more perltie.pod nit from Casey R. Tweten.
     Branch: perl
	   ! pod/perltie.pod
[  7909] By: jhi                                   on 2000/11/29  02:44:23
        Log: Document the known sprintf test failures, exact standard
             reference from Dominic Dunlop, NonStop-UX testing from Tom Bates.
     Branch: perl
	   ! pod/perldelta.pod t/op/sprintf.t
[  7908] By: jhi                                   on 2000/11/29  02:36:23
        Log: Add the Encoding table format documentation.
     Branch: perl
	   + ext/Encode/Encode/EncodeFormat.pod
[  7907] By: jhi                                   on 2000/11/29  02:07:24
        Log: Subject: Re: [PATCH] Updating perltie.pod for arrays
             From: "Casey R. Tweten" <>
             Date: Tue, 28 Nov 2000 12:11:41 -0500 (EST)
             Message-ID: <>
             Subject: Re: [PATCH] Updating perltie.pod for arrays
             From: "Casey R. Tweten" <>
             Date: Tue, 28 Nov 2000 15:03:50 -0500 (EST)
             Message-ID: <>
     Branch: perl
	   ! pod/perltie.pod
[  7906] By: nick                                  on 2000/11/28  22:40:59
        Log: Initial tidy of setmode() muddle (UNIX does not have, nor need it).
     Branch: perlio
	   ! iperlsys.h perlio.c
[  7905] By: nick                                  on 2000/11/28  22:04:34
        Log: Integrate mainline
     Branch: perlio
	  !> (integrate 39 files)
[  7904] By: jhi                                   on 2000/11/28  16:44:43
        Log: Subject: [ID 20001128.002] what's the point of example code if it is buggy?
             From: Nicholas Clark <>
             Date: Tue, 28 Nov 2000 11:16:57 +0000
             Message-Id: <>
     Branch: perl
	   ! pod/perlipc.pod
[  7903] By: jhi                                   on 2000/11/28  16:40:31
        Log: Subject: [PATCH] Updating perltie.pod for arrays
             From: "Casey R. Tweten" <>
             Date: Mon, 27 Nov 2000 17:47:20 -0500 (EST)
             Message-ID: <>
     Branch: perl
	   ! pod/perltie.pod
[  7902] By: jhi                                   on 2000/11/28  16:39:37
        Log: Subject: [PATCH] Re: 5.6 bug: split /^/ implies /m modifier (from CLPM)
             From: Robin Barker <>
             Date: Mon, 27 Nov 2000 17:56:44 GMT
             Message-Id: <>
     Branch: perl
	   ! pod/perlfunc.pod t/op/split.t
[  7901] By: jhi                                   on 2000/11/28  16:28:04
        Log: setmode() is a DOSish-only thing.
     Branch: perl
	   ! perlio.c
[  7900] By: jhi                                   on 2000/11/28  16:21:46
        Log: Subject: [PATCH] perlcc.PL cleanups
             From: Simon Cozens <>
             Date: Tue, 28 Nov 2000 12:44:35 +0000
             Message-ID: <>
     Branch: perl
	   ! utils/perlcc.PL
[  7899] By: gsar                                  on 2000/11/28  06:32:55
        Log: reintegrate files missed by change#7895
     Branch: maint-5.6/perl
	  +> ext/ByteLoader/bytecode.h utils/Makefile
	   - utils/perlbc.PL
[  7897] By: gsar                                  on 2000/11/27  18:22:47
        Log: can't integrate these two files, for some reason
     Branch: maint-5.6/perl
	   - ext/ByteLoader/bytecode.h utils/Makefile
[  7895] By: gsar                                  on 2000/11/27  18:11:21
        Log: integrate changes#6763..6766,6770,6773,6775..6776,6778,6780,
             6757..6890,6892..6901 from mainline
             Bytecompiler patches from Benjamin Stuhl.
             More bytecompiler.
             Subject: [PATCH blead] B:: missing dependency
             Subject: [PATCH: 6757] updates and syslog build
             Long double Gconvert fixes from Yitzchak Scott-Thoennes
             and Spider Boardman.
             Subject: [PATCH blead] nextchar() abuse misses an optimisation
             Long double fixes from Spider Boardman.
             Make the selection of NVff et al stricter.
             cSVOPo_*v things index into the current PL_curpad
             under ithreads, which is different from the curpad
             used by the XSUB.  (In other words, the code as-is
             before this patch wouldn't work under ithreads.)
             Be portable.
             VMS MMS (make) wants null action.
             Mac and other portability updates from Chris Nandor.
             Storable support, v-version fixes.
             Subject: beta for testing available
             Portability fix from Hugo van der Sanden.
             Bad makefile.
             Subject: [ID 20000823.004] [PATCH 5.6.0+] Pod::Html is too self-contained
             Subject: [PATCH] (Mac OS X): Don't #define environ unless PERL_CORE
             Subject: [PATCH] Re: [ID 20000821.008] Negitive numbers with vec dumps core
             Replace #6705 with a minimal doc patch.
             Subject: [PATCH 5.6.0] replace change #6705
             Drop the separate perlbc, perlcc -b should be enough.
             installperl couldn't tell whether it had run tests or not.
             Subject: [PATCH] Re: installperl and t/TEST
             Add silencer flags to installperl.
             Subject: [PATCH] Making installperl silent.
             Make "make install" by default silent.  A new "install-verbose"
             target is verbose.
             More liberal parsing of version numbers.
             Subject: Re: beta for testing available
             Create directories in silence.
             Subject: [PATCH] Another silencer for MakeMaker
             DOS patches and portability/porting notes, from Tim Jenness.
             Make installman to recognize the silence flag -S.
             Actually do something with the silencer option.
             Continue silencing.
             Show the doc file, not the temp file.
             Regen perltoc.
             Subject: [PATCH] More silencing of installman.
             Better wording for the vec lvalue diagnostic.
             Subject: Re: [PATCH] Re: [ID 20000821.008] Negitive numbers with vec dumps core
             Subject: [PATCH: 6805] several more tweaks to
             Subject: [PATCH perl@6805, 5.6.0, 5.005_03] prevent rare Perl hang on VMS
             Missing parts of 
             Subject: [PATCH: 6789] some endl fixes for VMS wackiness
             Subject: [ID 20000824.029] MakeMaker manifypods fails on DJGPP systems
             (applied slightly modified)
             installperl --verbose and --silent.
             Subject: Re: [PATCH] More silencing of installman.
             Add install-silent target.
             AIX 4.3.3 has SOCKS in libc with a differently named init routine,
             the problem reported in
             Subject: [ID 20000825.007] Building stable 5.6.0 on AIX 4.3.3 using SOCKS
             Tweak the sfio/useperlio logic, hopefully as wished in
             Subject: [ID 20000825.004] Not OK: perl v5.7.0 +SUIDMAIL +DEVEL6804 on i586-linux 2.2.12 (UNINSTALLED)
             One forgotten file from #6816.
             Subject: [PATCH @6820] installman under -w and strict (was Re: [PATCH] More silencing of installman.)
             Remove duplicately applied patch shards.
             Subject: [ID 20000825.012] [PATCH@6822] t/lib/cgi-html.t produces ugly cruft during 'make test'
             Support preserving extremely big/small angles.
             Subject: Re: [ID 20000825.019] Not OK: perl v5.7.0 +SUIDMAIL +DEVEL6820 on alpha-dec_osf 5.1 (UNINSTALLED) 
             Subject: [PATCH] installation not quite silent yet.
             Update the test count.
             Use UVxf, PTR2UV, NVff.
             Document PTR2XX and INT2PTR.
             no-install target a la make -n.
             Subject: [PATCH] make no-install (was Re: [PATCH] installation not quite silent yet.)
             grep -e isn't portable.
             Subject: [ID 20000825.027] let me ( know how I blew it
             Can't get the test to reliably work thanks to the
             inaccurateness of floating point.  "Resolves" bug ids
             20000826.003, 20000826.009, 20000826.010,
             Subject: installman buglet
             DJGPP update from Laszlo Molnar.
             Subject: LD_RUN_PATH niggles on Solaris
             Passing -R in ldflags makes now it to appear in the default
             for lddlflags, just like with -L.
             Subject: Re: LD_RUN_PATH niggles on Solaris
             Test nit.
             Use the actual thread type, not the pointer-to-struct.
             Provice virtual $Config{ccflags_nolargefiles} etc. 
             display_format used as a class method without arguments was broken,
             reported in
             Subject: Math::Complex->display_format() sets style to 'Math::Complex'
             Subject: [ID 20000828.006] dir name "0" not safe with
             Subject: [ID 20000828.009] Not OK: perl v5.7.0 +SUIDMAIL +DEVEL6855 on i586-linux 2.2.12 (UNINSTALLED)
             Subject: [PATCH@6855] _Minor_ change to pod
             opmini.o may be left around if a build is interrupted.
             Typo in #6858.
             Fix for ID 20000828.001, long doubles were not formatted
             correctly (showed up in $], which stopped installing perl).
             An attempt to fix the problem reported in
             Subject: Building perl@6856 using gcc/AIX 4.3.3 
             I can't test this properly since the gcc installation I have
             access to seems to be botched (gcc is calling the AIX cpp,
             a losing proposition...)
             Add -ld to archname on long tr...double platforms.
             Subject: hv.h Doc Patch
             Potential cruft.
             Subject: [PATCH bleedperl@6856] warnings fixes
             -S is the silent flag, -s is the strip flag.
             Subject: [PATCH] Re: [PATCH] make no-install          
             Take out the SUIDMAIL thing, that will not be
             a problem in 5.7.*.
             Subject: [PATCH bleedperl@6866] spellings
             Subject: [PATCH] Re: files not cleaned even by veryclean
             Use minimal @INC in tests, most of the time just '../lib',
             so that we simply can't pick up stuff from other Perls than
             the one we are testing.  Pointed out by
             Subject: Re: [PATCH: 6757] make new Storable tests forgiving of places where not built   
             Update to Getopt::Long 2.24, from Johan Vromans.
             Fix for thinko in #6848.
             Subject: Compiler error in ext/Thread/Thread.c (bleadperl@6866)
             Patches all over for people and the files they (hopefully) care about.
             Subject: Net::protoent does not export 'getproto'
             Missed a change in #6869.
             Subject: [PATCH] Warnings in B::Deparse
             Subject: [PATCH] Glob dumping
             Disable one of the tests for now.
             Disabling the one test is a bit tricky.
             Don't forget to tidy up.
             The #6881 removed one dump line.
             Subject: Re: [ID 20000525.003] perldoc fails when Makefile.PL is in cwd
             Under usethreads the dumped variable is IN_PAD.
             Subject: Re: [PATCH] Glob dumping
             Subject: [ID 20000829.026] [PATCH 6868] File::Temp
             Subject: [ID 20000829.022] [PATCH 6868] Minor nit in installhtml
             Subject: [ID 20000829.023] [PATCH 6868] -->
             Regen Configure for #6894.
             Subject: [PATCH: 6889] updates to perlebcdic.pod
             Undo namespace pollution of #6878.
             Subject: Re: Net::protoent does not export 'getproto'
             Admit that we are leaking scalars.
             Subject: [PATCH 5.6.0] [ID 20000608.006] panic: magic_killbackrefs with blessed global weakrefs
     Branch: maint-5.6/perl
	  !> (integrate 271 files)
[  7894] By: gsar                                  on 2000/11/27  16:00:34
        Log: a couple of nits
     Branch: maint-5.6/perl
	   ! MANIFEST pp_sys.c
[  7893] By: gsar                                  on 2000/11/27  15:10:56
        Log: integrate changes#6666..6678,6680..6682,6684..6691,6699..6733,
             Subject: Re: [ID 20000816.006] [PATCH @6655], bug fix, strict and OO Interface
             Subject: [PATCH(2) @6655] Re: perldebut.pod - spelling
             Doc nits spotted by Richard Soderberg.
             move WNOHANG definition to where other such things are
             Make $Config{byteorder} more magical so that it is
             dynamically computed: nice for 'fat binaries'.
             Subject: [PATCH]: default byteorder
             Subject: [PATCH] now uses strict
             Subject: Re: [PATCH]Re: Questions about Math::BigFloat
             Get -DLEAKTEST to compile (not necessarily to work, mind)
             Subject: [ID 20000724.006] -DLEAKTEST problem
             perldebtut 1.10 from Richard Foley, plus Celsius and Fahrenheit.
             Add perlebcdic from Peter Prymmer, regen toc.
             Don't propose using modules built for 5.005 if no binary
             compatibility with 5.005 is attempted.
             Do not use prototyping here.
             Subject: [ID 20000817.016] [PATCH] Peek.xs
             Document what the backtick returns if the command fails.
             Add byteorder to the myconfig output.
             Introduce NVef, NVff, and NVgf, use the middle one.
             (helps for lib/peek + Linux + long doubles)  Reported in
             Subject: [ID 20000814.005] Not OK: perl v5.6.0 on i686-linux-64int 2.2.13       
             Use NVs in POSIX math, not doubles.
             Subject: [ID 20000817.014] POSIX & modfl
             Subject: [PATCH 5.6.0+] newSVrv() memory leak
             The byteorder code in #6671 was wrong.
             Fix the lib/complex failure of
             Subject: [ID 20000814.005] Not OK: perl v5.6.0 on i686-linux-64int 2.2.13       
             Linux long double accuracy issue: something that
             when printed with %g looks like "2" but int() of it is 1.
             Propagate new Configure vars.
             Unbuffer the output.
             Subject: [PATCH] perltrap.pod spring cleaning
             Subject: [PATCH] perlfunc.pod -- clarifying sprintf array argument issues
             Subject: [ID 20000817.018] behaviour change 5.5.3 -> 5.6.0 re "Modification of a read-only value"
             Tiny Getopt::Long patch from Johan Vromans.
             Document code point which makes if (defined %stash::) to work
             (noted by Spider Boardman).
             Subject: [PATCH perl@6698] cygwin port
             Document the NDBM_File and ODBM_File as SDBM_File
             was documented in #6417.
             The new tests were missing from #6415.
             Add [[:blank:]] as suggested in
             Subject: [ID 20000716.024] [=cc=] / [:blank:]
             (the [=cc=] has already been taken care of by #6439
             so the whole bug report can be closed)
             and make [[:space:]] to be equivalent to isspace(3)
             (as opposed to \s, which is isSPACE()).  The difference
             is that now [[:space:]] matches the mythical vertical tab,
             while \s doesn't.
             Don't eat leading os from index entries.
             Subject: Re: [ID 20000810.006] Pod::Man Ate My 'O'!
             Subject: [PATCH 5.6.0+] fix for Win32::DomainName
             Typo in pp_complement().
             Subject: [PATCH perl-current] Deparse
             Add warnif(), check warnings further up the stack,
             all the warnings functions now can take an optional object reference.
             Subject: [PATCH bleedperl@6691] warnings pragma update
             Fix a core dump in lib/selfloader under -DDEBUGGING.
             Subject: PATCH @6698 for [ID 20000817.007] Not OK: perl v5.7.0 +SUIDMAIL +DEVEL6676 on alpha-dec_osf 4.0f (UNINSTALLED) 
             Subject: [PATCH 5.6.0+] fix for Win32::GetFullPathName and Win32::GetShortPathName
             Subject: [PATCH: 6698] tidy up the temp files left by peek tests on VMS
             Subject: [PATCH: 6698] was Re: [PATCH: 6640] VMS Makefile.SH update (fwd)
             Put back the long double avoidance code to POSIX.xs
             because VMS seems to need it still.
             Introduce a 'veryclean' target that is like 'distclean'
             but also removes *~ and *.orig.
             Subject: [ID 20000817.023] endianness description in perlfunc.pod
             Subject: [PATCH perl@6698] File::Temp fix-ups for OpenVMS
             Let's try #6717 again.
             UTF8 concat fixes.
             Subject: [PATCH @6713] Re: [ID 20000815.006] latest patched perl core dumps
             pp_open() could pass an uninitialized filename down to do_open9().
             Subject: Re: [ID 20000819.002] Not OK: perl v5.7.0 +SUIDMAIL +DEVEL6707 on i686-linux 2.2.5-16 (UNINSTALLED)
             Update to CGI 2.72, from Lincoln Stein.
             Subject: [PATCH] Silence MakeMaker (Was: installman)
             Use temporary directory instead of current directory.
             Subject: Re: [ID 20000816.011] Test failure in lib/ftmp-security.t
             Document odd vs even subreleases and -Dusedevel.
             The veryclean target needs to clobber.
             Use File::Spec->tmpdir().
             Document the number of exponent digits.
             Mention perlebcdic and perlposix-bc.
             s/this one/the 5.6.0 release/
             The #6724 is here.
             The correct cleaning order is an art.
             small tweaks for change#6705: avoid C++ style comments in C code;
             use Perl's malloc API rather than the low level system one
             Array context keeps slithering in.
             Subject: Re: 5.7.0 getting really close, new snapshot: perldelta, Storable
             Subject: [PATCH] os2.c fix for use64bitint
             Update to Pod::LaTeX 0.53.
             Subject: [PATCH] lib/Pod/ updates
             Document the endianness of Alpha more precisely.
             Subject: RE: [PATCH perl@6736] t/pragma/warn/9enabled assumes stdout buffered
             Rename the macro argument because some preprocessors
             can't tell the difference and expand arguments also inside
             double quoted strings.
             free TLS slot properly on Windows
             use Cwd 'chdir' didn't set $ENV{PWD} correctly on Windows
             Unicos/mk requires elaborate paranoia.
             Tweak the floating point output routine preferences.
             Also under djgpp the timestamps are funky.
             Apply some PodParser 1.18 patches; the Pod/
             patches cannot be applied since #6712 conflicts.
             Use PodParser 1.18 new test.
             A pod nit.
             Subject: [PATCH] pod/perlre.pod (was Re: [ID 20000821.007] $&, $1, etc. disappear when sub returns)
             Be verydeepclean.
     Branch: maint-5.6/perl
	  +> pod/perlebcdic.pod
	  !> (integrate 106 files)
[  7892] By: jhi                                   on 2000/11/27  15:02:32
        Log: Update Changes.
     Branch: perl
	   ! Changes patchlevel.h

$jhi++; #
        # There is this special biologist word we use for 'stable'.
        # It is 'dead'. -- Jack Cohen

Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About