develooper Front page | perl.perl5.porters | Postings from February 2009

a job: sync bleed and CPAN module versions

Thread Next
From:
Dave Mitchell
Date:
February 23, 2009 10:48
Subject:
a job: sync bleed and CPAN module versions
Message ID:
20090223184755.GA3168@iabyn.com
If anyone wants a job to help get 5.10.1 out the door, one thing that
would help would be to get CPAN and bleed better synced.

There are lots of dual-core modules in bleed that have been locally
hacked. So I need:

1) Someone to compare bleed with CPAN tarballs. Note that just because
a module has the same version number, it doesn't mean it has the same
contents. That's why I would be really happy to see a file-by-file diff.
If this can be scripted for future use, even better! (Note that under
Porting, there is cmpVERSION.pl and corecpan.pl, but neither of these do
quite what I'm asking). This should be somewhat easier now that Nicholas
has reorganised ext/

(Its possible that someone already does this and I'm just not aware of it).

2) Someone, armed with the info from (1), liaises with module authors to
either get a minor release pushed out to CPAN that syncs with bleed, or to
get the changes removed from bleed (which might be there by accident).
I'd prefer a minor update soonish rather than a "if you hold on a few more
weeks we'll be releasing v.2.00 which we've rewritten from scratch"

Also, if anyone knows of any particular module version X in bleed that
absolutely must or mustn't be in 5.10.1, or absolutely requires module Y
to be updated too, please let me know!

Thanks.

PS - some warts that I'm already aware of:

CPAN
    change 34702 did a last minute hack to lib/CPAN/FirstTime.pm that's in
    Bleed, 5.10.1 and 5.8.8, but not CPAN

CPANPLUS
    bleed bumped to 0.84_01 due to hack to make it work with IPC::Cmd 0.42
    (35119)

    also, lib/CPANPLUS/Internals/Constants/Report.pm has a silent change
    in bleed added by 34630


Data::Dumper
    5.10.0 was released with version 2.121_14; it would be nice if CPAN
    gained a non-devel release that could be included in 5.10.1


Errno
    bleed bumped to 1.10_01 due to haiko port (34630)

ExtUtils::CBuilder
    lib/ExtUtils/CBuilder.pm has a silent change in bleed added by 34630;
    lib/ExtUtils/CBuilder/Platform/cygwin.pm in Bleed,5.10.x is 0.24_01:
	(the ld ->cc hack)

ExtUtils::MakeMaker
    in bleed, MM_Unix.pm has been bumped to 4.68_01 to match hack in
	4adc95e616bac7eea015e9e47e439b063c1132d5.

     Change 34630 (haiku port) modified MM.pm
     and change 7b1668ed3d bumped its version (not yet matched)

     Change 34634 added MM_Haiku.pm,
     which doesn't seem to have made it into 6.48

File/Copy.pm and Copy.t
    version 2.14 in bleed has new preserve-permissions semantics;
    version 2.13 in maint doesn't.
    XXX what's in maint isn't 2.13; its 2.12 plus 34519 where nick
    (I think) accidently bumped the version to 2.13
    if 2.13 isn't in the wild, then this is probably ok

File::Path
    CPAN         = 2.07;
    Bleed,5.10.x,5.8.9 = 2.07_02
    due to last-minute security(?) hacks that went into 5.8.9.

Filter::Util::Call
    ext/Filter/Util/Call/Call.pm
    bleed has hacks taking it to 1.07_02

IO
    ext/IO/IO.pm
	CPAN 1.23;   5.10.0,maint,bleed: at 1.23_01 (no change since 5.10.0)

    ext/IO/lib/IO/Dir.pm
	33983, 33984 small local fix in bleed with _01 version bump

    IO::Handle
	    bleed bumped to 1.27_01 

    ext/IO/lib/IO/Dir.pm
	bleed bumped to 1.06_01
	    
IPC::Cmd
    14/2/09 5.10.x=0.40_1 CPAN=0.42 Bleed=0.42_01    
	    bleed includes hack to do with testing output

IPC::SysV
    3/1/09: bleed,CPAN at 2.00_02, maint at 2.00.


Math::BigInt

    22/2/09: lib/Math/BigInt/CalcEmu.pm in bleed contains a typo fix
	    ("it's" -> "its") missing from Math-BigInt-1.89.tar.gz, and
	    also not yet merged into maint. The fix was in change 33129,
	    which was merged into maint (skipping that fix) by 33610

Safe
    bleed at _01:

Test::Simple

    CPAN=0.86, Bleed,5.10.x=0.86_01


Test::Harness
    these change have been silently added to bleed, 5.10.x
    without bumping any version numbers:
	34361
	    ext/Test/Harness/t/sample-tests/delayed
	ditto 35076
	    ext/Test/Harness/t/compat/regression.t
	    this one needed otherwise it fails with Tes::Simple 0.86

	Also, the migration from Test/Harness to Test-Harness resulted in
	path fixups in lots of files.
    Apparently 3:15 due soon


Time::Piece
    bleed at 1.13_02



-- 
The Enterprise successfully ferries an alien VIP from one place to another
without serious incident.
    -- Things That Never Happen in "Star Trek" #7

Thread Next


nntp.perl.org: Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at ask@perl.org | Group listing | About