CPAN
view release on metacpan or search on metacpan
Thu Sep 6 04:39:33 2007 TURNERJW/Tk-TextHighlight-1.0.2.tar.gz
Thu Sep 6 03:12:29 2007 PANGJ/IP-ChinaISP.tgz
Thu Sep 6 02:42:05 2007 JOHANL/Devel-PerlySense-0.01_09.tar.gz
Thu Sep 6 02:16:41 2007 JOHANL/Devel-PerlySense-0.01_08.tar.gz
Thu Sep 6 00:48:24 2007 DAGOLDEN/CPAN-Reporter-0.99_07.tar.gz
Wed Sep 5 23:52:49 2007 DDICK/Net-DNS-DynDNS-0.95.tar.gz
Wed Sep 5 23:42:30 2007 CHRMUE/Socket-Class-1.1.0.tar.gz
Wed Sep 5 23:09:40 2007 PAJOUT/XML-Trivial-0.02.tar.gz
Wed Sep 5 23:08:13 2007 DAGOLDEN/CPAN-Reporter-0.99_06.tar.gz
Wed Sep 5 22:57:54 2007 DSNOPEK/POE-Component-MessageQueue-0.1.6.tar.gz
Wed Sep 5 21:33:13 2007 CORION/HTML-Display-0.37.tar.gz
Wed Sep 5 21:18:01 2007 KEN/XML-XML2JSON-0.01.tar.gz
Wed Sep 5 21:16:35 2007 GWYN/Text-Diff-Parser-0.07.tar.gz
Wed Sep 5 19:33:12 2007 BINGOS/POE-Component-Server-DNS-0.11.tar.gz
Wed Sep 5 19:15:18 2007 ALEXMASS/InSilicoSpectro-Databanks-0.0.21.tar.gz
Wed Sep 5 18:29:47 2007 BLBLACK/POE-Loop-Event_Lib-0.001_01.tar.gz
Wed Sep 5 17:27:54 2007 SMUELLER/Archive-Unzip-Burst-0.02_01.tar.gz
Wed Sep 5 17:17:35 2007 URKLE/DJabberd-VCard-LDAP-0.03.tar.gz
Wed Sep 5 17:07:16 2007 KARJALA/MyXML/XML-MyXML-0.0985.tar.gz
Wed Sep 5 16:10:25 2007 ISAAC/Business-PayPal-EWP-1.00.tar.gz
Wed Sep 5 16:03:54 2007 DGL/Net-CouchDb-0.01.tar.gz
Wed Sep 5 15:47:17 2007 JDHEDDEN/threads-shared-1.13.tar.gz
Wed Sep 5 15:38:10 2007 ISAAC/Business-PayPal-EWP-0.03.tar.gz
Wed Sep 5 15:22:47 2007 JDHEDDEN/threads-1.65.tar.gz
Wed Sep 5 15:18:46 2007 JSCHNEID/DBIx-ORM-Declarative-0.21.tar.gz
Wed Sep 5 14:55:53 2007 JSCHNEID/DBIx-ORM-Declarative-0.20.tar.gz
Wed Sep 5 14:48:06 2007 YAPPO/Class-Component-0.06.tar.gz
Wed Sep 5 14:25:13 2007 TBUSCH/Text-Scan-0.29.tar.gz
Here is still one of the broken bs file:
-rw-rw-rw- 23942 sand sand 2217 Sep 6 22:09 /home/src/perl/repoperls/installed-perls/maint-5.8/pFrR8O4/perl-5.8.0@31223/lib/5.8.8/i686-linux-thread-multi-64int/auto/File/Glob/Glob.bs
And I find 163 identical *.bs files scattered all over the place with
the same timestamp. And they all start with
/home/src/perl/repoperls/installed-perls/maint-5.8/p4wjLcy/perl-5.8.0@22552/lib/site_perl/5.8.3/i686-linux-thread-multi-64int/Data/Dump/Streamer.pm
/home/src/perl/repoperls/installed-perls/maint-5.8/p4wjLcy/perl-5.8.0@22552/lib/site_perl/5.8.3/i686-linux-thread-multi-64int/Data/Dump/Streamer/_/Printers.pm
/home/src/perl/repoperls/installed-perls/maint-5.8/p4wjLcy/perl-5.8.0@22552/lib/site_perl/5.8.3/i686-linux-thread-multi-64int/auto/Data/Dump/Streamer/Streamer.bs
/home/src/perl/repoperls/installed-perls/maint-5.8/p4wjLcy/perl-5.8.0@22552/lib/site_perl/5.8.3/i686-linux-thread-multi-64int/auto/Data/Dump/Streamer/Streamer.so
/home/src/perl/repoperls/installed-perls/maint-5.8/p4wjLcy/perl-5.8.0@22552/man/man3/Data::Dump::Streamer.3
/home/src/perl/repoperls/installed-perls/maint-5.8/pFrR8O4/perl-5.8.0@31223/lib/site_perl/5.8.8/i686-linux-thread-multi-64int/Data/Dump/Streamer.pm
/home/src/perl/repoperls/installed-perls/maint-5.8/pFrR8O4/perl-5.8.0@31223/lib/site_perl/5.8.8/i686-linux-thread-multi-64int/Data/Dump/Streamer/_/Printers.pm
/home/src/perl/repoperls/installed-perls/maint-5.8/pFrR8O4/perl-5.8.0@31223/lib/site_perl/5.8.8/i686-linux-thread-multi-64int/auto/Data/Dump/Streamer/Streamer.bs
/home/src/perl/repoperls/installed-perls/maint-5.8/pFrR8O4/perl-5.8.0@31223/lib/site_perl/5.8.8/i686-linux-thread-multi-64int/auto/Data/Dump/Streamer/Streamer.so
[...]
What was smoking with 31223 at 22:09? I just searched for my reports on
rpm-build-perl on http://www.nntp.perl.org/group/perl.cpan.testers/ and
they seem to be fine.
In my eyes there is only one candidate at the moment: DDS.
In the meantime Max Maischain (CORION) complained as well.
Update 2007-09-08: I just discovered that not only the 10 perls were
broken that I'm using to smoke but many, many others as well. The
timestamp of the file is 2007-09-06T22:09, they have 2217 bytes and have
the same md5sum and although I have already deleted plenty of them I
still count 22049 of these files. Somebody was really busy.
Hrm. Now I have a suspect. My trimtrees is most likely to blame. These
files will have the same inode and somebody only changed one random
zero-byte file without unlinking it first and all this happened. Yes?
Indeed, the inode is 1855345.
So I'll have to fix trimtrees:-(
2007-09-06 Andreas J. Koenig <andreas.koenig.7os6VVqR@franz.ak.mind.de>
* Something broke my parser for the output of the test runs.
residuum.yml are the unparsed lines. Ahhh, it's the new //hint// line.
FIXED.
% ls -l logs/megainstall.200709*.d/residuum.yml
-rw-rw-r-- 1 sand sand 86009 Sep 1 08:05 logs/megainstall.20070901T0114.d/residuum.yml
-rw-rw-r-- 1 sand sand 84530 Sep 1 23:34 logs/megainstall.20070901T1206.d/residuum.yml
-rw-rw-r-- 1 sand sand 71977 Sep 3 03:48 logs/megainstall.20070902T1240.d/residuum.yml
-rw-rw-r-- 1 sand sand 4168125 Sep 5 13:18 logs/megainstall.20070905T0107.d/residuum.yml
-rw-rw-r-- 1 sand sand 4307567 Sep 6 03:28 logs/megainstall.20070905T1716.d/residuum.yml
2007-09-04 Andreas J. Koenig <andreas.koenig.7os6VVqR@franz.ak.mind.de>
* Bug in Pod::XXX: when writing manpage for
TOBY/Geo-Coordinates-OSGB-2.00.tar.gz there is a pm file with 11 MB.
Some perls do that without hesitating but others (and they seem to be
the more recent ones) never seem to finish.
2007-09-02 Andreas J. Koenig <andreas.koenig.7os6VVqR@franz.ak.mind.de>
* Bug: with a Makefile.PL that has a timestamp in the future, we cannot
proceed very far. Because the generated Makefile will be outdated.
Likewise with all other files in a distro.
* Todo: config option to enable code refs in YAML. (LoadCode)
rafl: DONE
2007-08-30 Andreas Koenig <k@andreas-koenigs-computer.local>
* hashed out on the bus with anonymous, how the endless loop of RECENT
testing shall look like: one never dying job produces POE sessions that
shall test some distro with some perl. This queue is always keeping two
or three jobs active in parallel and watches stdout+stderr. If for more
than 10 minutes no output is produced the job gets a SIGSTOP and it is
moved to the unlimited(?) heap of hangers and the next job from the
queue is made active. Somebody provides us an interface to send stdin to
one of the hangers. As soon as this is done that job gets queued again
and is not considered a hanger anymore. The queuerunner will then send
it a SIGCONT to activate it. Somebody provides us an interface to list
all hangers, and one to show the output of one hanger.
http://gumbynet.org.uk/hackintro/
http://gumbynet.org.uk/hackathon/
* todo: commit bit for florian
2007-08-29 Andreas Koenig <k@andreas-koenigs-computer.local>
* bug from tina: new debian without gcc. Then install Bundle::CPAN and
it leads to the installation of a Compress::Zlib that does not work
( run in 1.348 second using v1.01-cache-2.11-cpan-5a3173703d6 )