CPAN
view release on metacpan or search on metacpan
----EOF ($?='0')----
Need a perl between 17639 and 17705
(but 17641, 17642, 17644, 17654, 17655, 17656, 17671, 17672, 17673, 17674, 17675, 17676, 17677, 17678, 17679, 17680, 17681, 17682, 17683, 17684, 17685, 17686, 17687, 17688, 17689, 17690, 17691, 17692, 17693, 17694, 17695, 17696, 17697, 17698, 17699, ...
No useable patch available between 17639 and 17705
Patches 17641, 17642, 17644, 17654, 17655, 17656, 17671, 17672, 17673, 17674, 17675, 17676, 17677, 17678, 17679, 17680, 17681, 17682, 17683, 17684, 17685, 17686, 17687, 17688, 17689, 17690, 17691, 17692, 17693, 17694, 17695, 17696, 17697, 17698, 1769...
pseudo hashes were extinct, when? 17725!
So it isn't a pseudo hash thing? well, 17705 failed because XML::Twig
was not working then. But when I 'force notest install XML::Twig', then
SVG::Metadata passes its tests.
Next round of binsearch with the fixed 17705. And then I had to find out
that SVG::Metadata does not declare its dependency on HTML::Entities and
had to install that manually into some involved perls and finally I got:
----Program----
eval q{use SVG::Metadata 0.28};
print $@ ? "N/A" : "OK";
print "\n";
----Output of .../pjEfwQE/perl-5.8.0@17724/bin/perl----
OK
----EOF ($?='0')----
----Output of .../pg7Kp5D/perl-5.8.0@17725/bin/perl----
N/A
----EOF ($?='0')----
* Assertion: Environment is being set during perl {Build,Makefile}.PL?
iow: is set_perl5lib called from within get or its deputies? Yes, in
line 7193 within make
* Idea: new option to cleanup the build_dir not by size but by age.
* Todo: with the new sendmail configuration I should retry if mailtools
work now. Which perl is good for that? Stock 5.8.8 should do and note
there is a 2.00_xx! For install I can configure back to using sudo. But
all testing should be done by sand because there is the reporter DB. So
this time I want to disable prefs? No, k installs, sand tests.
Yes, even with Test::Reporter 1.27 the mail from Mail::Send seems to
have been accepted for delivery (RJBS/Test-BinaryData-0.003.tar.gz). Now
with Test::Reporter 1.30 I tried KWILLIAMS/Path-Class-0.16.tar.gz. And
finally with MARKOV/MailTools-2.00_01.tar.gz I try
JROCKWAY/Directory-Scratch-0.12.tar.gz
Mails went out through sendmail.
Mailq is empty. But the mails did not come back to me on the
subscription address. Nor to Google
http://groups.google.com/group/perl.cpan.testers/topics?lnk=srg
Let me compare the 3 mails above with the last mail before that. It was
about BPAN. A bad one with Mail::Send:
# egrep 'l6L5ZkbD030283' /var/log/mail.info
Jul 21 07:35:46 k75 sendmail[30283]: l6L5ZkbD030283: from=sand, size=5897, class=0, nrcpts=1, msgid=<200707210535.l6L5ZkbD030283@k75.linux.bogus>, relay=sand@localhost
Jul 21 07:35:46 k75 sm-mta[30284]: l6L5Zk0d030284: from=<sand@k75.linux.bogus>, size=6121, class=0, nrcpts=1, msgid=<200707210535.l6L5ZkbD030283@k75.linux.bogus>, proto=ESMTP, daemon=MSA, relay=localhost.localdomain [127.0.0.1]
Jul 21 07:35:46 k75 sendmail[30283]: l6L5ZkbD030283: to=cpan-testers@perl.org, ctladdr=sand (1005/1005), delay=00:00:00, xdelay=00:00:00, mailer=relay, pri=35897, relay=[127.0.0.1] [127.0.0.1], dsn=2.0.0, stat=Sent (l6L5Zk0d030284 Message accepted fo...
Jul 21 07:35:59 k75 sm-mta[30286]: l6L5Zk0d030284: to=<cpan-testers@perl.org>, ctladdr=<sand@k75.linux.bogus> (1005/1005), delay=00:00:13, xdelay=00:00:13, mailer=esmtp, pri=126121, relay=mx.develooper.com. [63.251.223.176], dsn=2.0.0, stat=Sent (Que...
And a good one with Net::SMTP:
# egrep 'l6L5MQN5027327' /var/log/mail.info
Jul 21 07:22:26 k75 sm-mta[27327]: l6L5MQN5027327: from=<andreas.koenig.gmwojprw@franz.ak.mind.de>, size=4884, class=0, nrcpts=1, msgid=<200707210522.l6L5MQN5027327@franz.ak.mind.de>, proto=ESMTP, daemon=MTA, relay=localhost.localdomain [127.0.0.1]
Jul 21 07:22:39 k75 sm-mta[27329]: l6L5MQN5027327: to=<cpan-testers@perl.org>, delay=00:00:13, xdelay=00:00:12, mailer=esmtp, pri=124884, relay=mx.develooper.com. [63.251.223.176], dsn=2.0.0, stat=Sent (Queued! 1184995359 qp 21633 <200707210522.l6L5M...
Too obvious. CPAN::Reporter or Test::Reporter is unable to set the from
address when switching to Mail::Send.
Stepping through the debugger. Test::Reporter::_mail_send line 344 sets
the From correctly. Later Mail::Send passes itself over to Mail::Mailer
DB<14> n
Mail::Send::open(/usr/local/perl-5.8.8/lib/site_perl/5.8.8/Mail/Send.pm:57):
57: Mail::Mailer->new(@_)->open($self);
DB<14> x \@_
0 ARRAY(0x85b1c74)
empty array
DB<15> x $self
0 Mail::Send=HASH(0xbe3ac30)
'From' => ARRAY(0xbe3c75c)
0 'andreas.koenig.gmwojprw@franz.ak.mind.de'
'Subject' => ARRAY(0xbe3c7d4)
0 'PASS WebService-Mogo2-0.01 i686-linux-64int 2.6.14'
'To' => ARRAY(0xbe3a9e4)
0 'cpan-testers@perl.org'
'X-Reported-Via' => ARRAY(0xbe3be14)
0 'Test::Reporter 1.30, via CPAN::Reporter 0.46'
Aha. I have sendmail installed and despite Mail::Send and Mail::Mailer
carefully keep the $sender and $hdr intact, Mail::Mailer::sendmail
throws it away because it cannot handle it.
The documentation of Test::Reporter tells me that I can use
$tr->mail_send_args(['smtp']); perhaps. So let's try. With MailTool
2.00_01 I must write a bugreport that there is a syntax error.
http://rt.cpan.org/Ticket/Display.html?id=28411
With current version 1.77 there is a failure:
Sending test report with 'pass' to cpan-testers@perl.org
(in cleanup) Undefined subroutine &Mail::Mailer::smtp::close called at /usr/local/perl-5.8.8/lib/site_perl/5.8.8/Mail/Mailer.pm line 204.
Catching error: 'Undefined subroutine &Mail::Mailer::smtp::close called at /usr/local/perl-5.8.8/lib/site_perl/5.8.8/Mail/Mailer.pm line 133.
(in cleanup) Undefined subroutine &Mail::Mailer::smtp::close called at /usr/local/perl-5.8.8/lib/site_perl/5.8.8/Mail/Mailer.pm line 204.
' at lib/CPAN.pm line 275
CPAN::shell() called at -e line 1
So maybe the mail was sent but it certainly was a die() before the
bookkeeping was done:-( No, from the logfile it seems there was no mail
sent at all.
So I try again with 2.00_01 and insert an "or" before the return to fix
the syntax. This dies without an error message.
So I come to the conclusion that I must fix Test::Reporter for me.
IS this all I need?
- use constant FAKE_NO_MAIL_SEND => 0; # for debugging only
+ use constant FAKE_NO_MAIL_SEND => 1; # for debugging only
YES!
2007-07-19 Andreas J. Koenig <andreas.koenig.7os6VVqR@franz.ak.mind.de>
http://rt.cpan.org/Ticket/Display.html?id=27721
* MLEHMANN/AnyEvent-2.52.tar.gz Started dying on 2007-06-21T21:46 with
'X connection to localhost:10.0 broken (explicit kill or server
shutdown).' But not always.
* JRED/Event-ExecFlow-0.63.tar.gz?
Depends on AnyEvent.
* DBRIAN/XML-SimpleObject-0.53.tar.gz? Does not declare prereq XML::Parser
http://rt.cpan.org/Ticket/Display.html?id=27109 (six weeks ago)
2007-06-23 Andreas J. Koenig <andreas.koenig.7os6VVqR@franz.ak.mind.de>
* CHROMATIC/Crypt-CipherSaber-1.00.tar.gz had a dropout at 31444 need to
look closer
* Off topic Todo: Devel::Symdump fails recur test with 5.00504.
But succeeds in the debugger.
* DMAKI/File-Extract-0.06.tar.gz is dependent on Spreadsheet::Read and
therefore has mixed results.
* Refactor the distroprefs with one file for all the "standard" MI
distros. DONE but as of 31446 not yet tested.
2007-06-22 Andreas J. Koenig <andreas.koenig.7os6VVqR@franz.ak.mind.de>
* Glib: I once put a patch for Glib into /tmp/ and then forgot and then
lost the patch. Today I reconstructed what might have been in there so
that the distroprefs get useable again. Similarly I had to put the Gtk2
patch from patches/ to /tmp/.
2007-06-20 Andreas J. Koenig <andreas.koenig.7os6VVqR@franz.ak.mind.de>
* ZOOLEIKA/SVG-Plot-0.06.tar.gz
http://rt.cpan.org/Ticket/Display.html?id=27652
* MRAMBERG/Catalyst-Plugin-Unicode-0.3.tar.gz is among the new fails but
I don't see a fail report on cpantesters. Ah, he did not include the
inc/ directory so need Module::Install installed. Does not build, cannot
write a Reporter mail. Already reported by DANIELTWC
https://rt.cpan.org/Ticket/Display.html?id=27633
* ARANDAL/parrot-0.4.13.tar.gz does not even finish make:-(
2007-06-19 Andreas J. Koenig <andreas.koenig.7os6VVqR@franz.ak.mind.de>
* DBI 1.57 needs a binary search and I cannot compile perls between
20373 and 20575. They all fail with
cc -L/usr/local/lib -o miniperl \
miniperlmain.o opmini.o libperl.a -lnsl -ldl -lm -lcrypt -lutil -lc
libperl.a(mg.o): In function `Perl_magic_set':
/home/src/perl/repoperls/perl-p-5.8.0@20573/mg.c:2215: undefined reference to `setruid'
/home/src/perl/repoperls/perl-p-5.8.0@20573/mg.c:2269: undefined reference to `setrgid'
/home/src/perl/repoperls/perl-p-5.8.0@20573/mg.c:2364: undefined reference to `setproctitle'
libperl.a(pp_sys.o): In function `Perl_pp_fteread':
/home/src/perl/repoperls/perl-p-5.8.0@20573/pp_sys.c:2949: undefined reference to `eaccess'
libperl.a(pp_sys.o): In function `Perl_pp_ftewrite':
/home/src/perl/repoperls/perl-p-5.8.0@20573/pp_sys.c:2976: undefined reference to `eaccess'
libperl.a(pp_sys.o): In function `Perl_pp_fteexec':
/home/src/perl/repoperls/perl-p-5.8.0@20573/pp_sys.c:3003: undefined reference to `eaccess'
collect2: ld returned 1 exit status
So we could try to replace mg.c and pp_sys.c with later versions? We
could look into our history if we ever had such an encounter -- No.
So we need to document that because we will have to go into this region
again.
Where were mg.c changes?
./bin/blame-helper.pl --diffdir APC/5.9.0/diffs --lower 20400 --upper 24009 mg.c
As expected, I cannot compile 20829 either. As all this is highly
unlikely to help we should rather look into config.sh:
Maybe I should try to add to --config '-Ud_eaccess -Ud_setruid
-Ud_setrgid -Ud_setproctitle'? I think I did that before. Because we
have the directories already I edit the config.sh to the same effect and
run make again. I do this in 20474. Yes this is the way to go.
Lets write this in perl:
perl -pi~ -e 's/define/undef/ if /^d_(eaccess|setruid|setrgid|setproctitle)=/' config.sh
----Program----
eval q{use DBI 1.57};
print $@ ? "N/A" : "OK";
print "\n";
----Output of .../pCkomOb/perl-5.8.0@20556/bin/perl----
OK
----EOF ($?='0')----
----Output of .../pdIogQu/perl-5.8.0@20559/bin/perl----
N/A
----EOF ($?='0')----
Change 20559 by rgs@rgs-home on 2003/08/07 19:59:18
Subject: Fwd: Re: [PATCH] flags set incorrectly on abandoned lexicals
From: Dave Mitchell <davem@fdgroup.com>
Date: Thu, 7 Aug 2003 15:53:35 +0100
Message-ID: <20030807145335.GF26683@fdgroup.com>
t/86gofer_fail............ok 21/0Unexpected failure: DBD::Gofer::db do failed: A
ssertion !((sv)->sv_flags & 0x00000100) failed: file "pad.c", line 1411 at /home
/sand/.cpan/build/DBI-1.57-QQq6PO/blib/lib/DBI/Gofer/Execute.pm line 624. at t/8
6gofer_fail.t line 129.
# Looks like your test died just after 22.
t/86gofer_fail............dubious
Test returned status 255 (wstat 65280, 0xff00)
after all the subtests completed successfully
2007-06-18 Andreas J. Koenig <andreas.koenig.7os6VVqR@franz.ak.mind.de>
* SVG-Metadata-0.28 might be a pseudo hash issue.
* TANIGUCHI/Unicode-RecursiveDowngrade-0.03.tar.gz is the next candidate.
And JUERD/encoding-split-0.01.tar.gz?
----Program----
eval q{use Unicode::RecursiveDowngrade 0.03};
print $@ ? "N/A" : "OK";
print "\n";
----Output of .../pt5LKrb/perl-5.8.0@24009/bin/perl----
OK
----EOF ($?='0')----
----Output of .../pAnzsYU/perl-5.8.0@24010/bin/perl----
N/A
----EOF ($?='0')----
----Program----
# does not define a VERSION, so we must test without
eval q{use encoding::source 'latin1'};
print $@ ? "N/A" : "OK";
print "\n";
( run in 0.810 second using v1.01-cache-2.11-cpan-97f6503c9c8 )