view release on metacpan or  search on metacpan

Todo  view on Meta::CPAN

	But in megainstall.20070103T0336.out I see that 'NI-S/Tk-804.027.tar.gz'
	has 'make_test NO'. Looking at its YAML I see it is 'CALLED_FOR:
	Tk::HList'. So is Tk::HList lost in SVN?

	But this is again a case where browseentry failed and NOTHING got
	installed. Bug is reported, do not pay attention to more than one bug:)

	* Class::MethodMaker

lib/Class/MethodMaker.xs: In function 'XS_Class__MethodMaker_set_sub_name':
lib/Class/MethodMaker.xs:12: error: invalid lvalue in assignment
lib/Class/MethodMaker.xs:13: error: invalid lvalue in assignment

Module id = Class::MethodMaker
    DESCRIPTION  Create generic class methods
    CPAN_USERID  FLUFFY (Martyn J. Pearce <fluffy@cpan.org>)
    CPAN_VERSION 2.08
    CPAN_FILE    F/FL/FLUFFY/Class-MethodMaker-2.08.tar.gz
    UPLOAD_DATE  2005-11-26
    DSLIP_STATUS MdpOp (mature,developer,perl,object-oriented,Standard-Perl)
    MANPAGE      Class::MethodMaker - Create generic methods for OO Perl
    INST_FILE    (not installed)

	Given that https://rt.cpan.org/Ticket/Display.html?id=16241 is 12 months
	old, it seems abandoned.

	http://rt.cpan.org/Ticket/Display.html?id=22413
	http://aspn.activestate.com/ASPN/Mail/Message/perl5-porters/3342382

	The stupid thing is that there is a signature.t test and when we use
	distropref/patch to fix MethodMaker.xs, then the signature test breaks.
	Patching SIGNATURE does not help either because the file is signed. The
	solution is to patch 0-signature.t too and let it skip all tests.

	FINI.

	* Bug: The FTPstats file has been cleared once again. 1167639627.93624.
	Only 20 minutes ago. This must have happened when I killed the sluggish
	process. Now I made the writing of the FTPstats file atomic. FIXED in
	SVK rev. 1672

	* It seems that blead@20002 has difficulties with YAML or YAML::Syck.
	The binary search that I have running crawls and strace shows a lot of
	mremap activities and reading of FTPstats (maybe too often?) and long
	phases of no system calls at all. I kill the process and start again.
	Speed now seems normal. New zlib modules get installed. Some fail. There
	are many of them, all with 2.002 as version number--slightly confusing.
	Now CPAN::Injects demands prereqs. Compress::Bzip2 fails with "No rule
	to make target 'Makefile'". File-Find-Rule demands Number::Compare, gets
	it, finishes its own build and the tests and gets installed. Now for
	Data::Compare. It seems to get sluggish. CPAN::Checksums can't get
	installed because of missing Compress::Bzip2. Now somebody wants
	IPC::Run3 and we slowly seem to get it. I believe it is CPAN.pm which is
	losing too much time with some nonsense.

	* Somebody drops a daemon from a test running forever:

k        18692  0.0  0.0   2956  1328 pts/4    S+   Jan01   0:00 sh -c /home/src/perl/repoperls/installed-perls/perl/prOxRQY/perl-5.8.0@29651/bin/perl -Iblib/lib -Iblib/arch -I/home/src/perl/repoperls/installed-perls/perl/prOxRQY/perl-5.8.0@29651/lib...

	lsof -p 18692
lsof: WARNING: can't stat() ext3 file system /dev/.static/dev
      Output information may be incomplete.
COMMAND   PID USER   FD   TYPE DEVICE    SIZE     NODE NAME
sh      18692    k  cwd    DIR    9,1       0 11390681 /home/k/.cpan/build/Net_SSLeay.pm-1.30-KYd1Uh (deleted)

	cpan[4]> m Net::SSLeay
Module id = Net::SSLeay
    DESCRIPTION  Secure Socket Layer (based on OpenSSL)
    CPAN_USERID  FLORA (Florian Ragwitz <rafl@debian.org>)
    CPAN_VERSION 1.30
    CPAN_FILE    F/FL/FLORA/Net_SSLeay.pm-1.30.tar.gz
    UPLOAD_DATE  2005-12-26
    DSLIP_STATUS Rmcfr (released,mailing-list,C,functions,restricted_distribution)
    MANPAGE      Net::SSLeay - Perl extension for using OpenSSL
    INST_FILE    /home/src/perl/repoperls/installed-perls/perl/pSJpiak/perl-5.8.0@29656/lib/site_perl/5.9.5/i686-linux-64int/Net/SSLeay.pm
    INST_VERSION 1.30


	The queue in RT is called "Net_SSLeay.pm"

	The report is http://rt.cpan.org/Ticket/Display.html?id=24180

	FINI.
	
2006-12-31  Andreas J. Koenig  <andreas.koenig.7os6VVqR@franz.ak.mind.de>

	* new PAUSE and own key!

	* Todo: autosave/autocommit DONE

	* Why do I always see

Prepending /home/k/.cpan/build/Graphics-ColorNames-2.0_04-fxepKY/blib/arch /home/k/.cpan/build/Graphics-ColorNames-2.0_04-fxepKY/blib/lib /home/k/.cpan/build/HTML-Tidy-1.07_01-wuI8Hl/blib/arch /home/k/.cpan/build/HTML-Tidy-1.07_01-wuI8Hl/blib/lib /ho...
	
	Graphics::ColorNames   is a goto and a Module::Build
	HTML::Tidy             is a goto
	YAML::Syck             nothing special visible, maybe an artefact
	                       because it is involved in writing and reading
	                       status information?

	There must a bug in the goto implementation.

	Yes, the bug is visible when doing a manual install of
	Graphics::ColorNames in that the test is run twice. A part of the reason
	seems to be that the distro stays on the queue when we have finished
	installing it. The solution might be to remove it within the goto method.

	DONE in SVK rev. 1652

	And for YAML::Syck I find indeed the artefact AND a bug in the reanimate
	routine. I wrote $obj->{make_install} but it should be {install}.

	FIXED in SVK rev. 1653

	* Todo: verify that CPAN_RUN_SHELL_TEST_WITHOUT_EXPECT=1 still works.

	It's incredible slow. Maybe the $prompt_re is inefficient? No, the
	problem is that the more tests we have, the longer is the string that
	needs testing. The output must be split on the prompt and then tested
	one piece at a time. This is a hard job because we do not know where to
	split. 31sessions.t seems better, let's move there and go away from