Skip Menu |

This queue is for tickets about the Term-ReadLine-Gnu CPAN distribution.

Maintainer(s)' notes

When you report a bug, please provide the following information;

- output of
	perl -V
	perl Makefile.PL verbose
	make test TEST_VERBOSE=1
	perl -Mblib t/00checkver.t
	echo $TERM
- terminal emulator which you are using
- compiler which is used to compile the GNU Readline Library (libreadline.a) if you can know.
Read INSTALL in the distribution for more details.

Report information
The Basics
Id: 114235
Status: resolved
Priority: 0/
Queue: Term-ReadLine-Gnu

People
Owner: HAYASHI [...] cpan.org
Requestors: SREZIC [...] cpan.org
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: 1.31
Fixed in: 1.32



Subject: t/utf8_binary.t fails with perl 5.8.9
Test output is: Use of uninitialized value in print at t/utf8_binary.t line 72, <$in> line 1. Use of uninitialized value in string eq at t/utf8_binary.t line 74, <$in> line 1. # Failed test 4 in t/utf8_binary.t at line 74 # t/utf8_binary.t line 74 is: ok($line eq "漢字2"); t/utf8_binary.t .. Failed 1/5 subtests Wide character in print at t/utf8_text.t line 76. The same test works fine on the same linux system with perl 5.10.1 .. 5.24.0-RC5.
Hi, Thank you for your report. I checked CPANTESTERS log. I see 12 passes with perl 5.8.9 and 1 pass with perl 5.8.2. Only 2 fails with perl 5.8.9. Both are yours. http://www.cpantesters.org/cpan/report/183b31ee-14e5-11e6-bddd-8c34573ac0e6 This is the fail for this ticket. http://www.cpantesters.org/cpan/report/c55248a0-e389-11e5-9d60-b001fdd2507e This is the fail on one of t/readline.t tests which occurs only on your environment. 64/66 fails on Term:ReadLine:Gnu-1.31 are failed on the error. Show quoted text
> t/00checkver.t ... ok > t/callback.t ..... ok > t/history.t ...... ok > t/readline.t ..... > Failed 1/140 subtests > t/utf8_binary.t .. ok > The locale 'en_US.utf8' is not supported. Skipped... > t/utf8_text.t .... ok
Interestingly http://www.cpantesters.org/cpan/report/183b31ee-14e5-11e6-bddd-8c34573ac0e6 passes the t/readline.t test. Show quoted text
> t/00checkver.t ... ok > t/callback.t ..... ok > t/history.t ...... ok > ok Try `/bbbike/perl-5.8.9/bin/perl -Mblib t/readline.t verbose', if you will. > t/readline.t ..... ok > Use of uninitialized value in print at t/utf8_binary.t line 72, <$in> line 1. > Use of uninitialized value in string eq at t/utf8_binary.t line 74, <$in> line 1. > # Failed test 4 in t/utf8_binary.t at line 74 > # t/utf8_binary.t line 74 is: ok($line eq "漢字2"); > t/utf8_binary.t .. > Failed 1/5 subtests > Wide character in print at t/utf8_text.t line 76. > t/utf8_text.t .... ok
Do you have any idea or hints? Regards, On Sun, 8 May 2016 07:07:47 GMT, SREZIC wrote: Show quoted text
> Test output is: > > Use of uninitialized value in print at t/utf8_binary.t line 72, <$in> > line 1. > Use of uninitialized value in string eq at t/utf8_binary.t line 74, > <$in> line 1. > # Failed test 4 in t/utf8_binary.t at line 74 > # t/utf8_binary.t line 74 is: ok($line eq "漢字2"); > t/utf8_binary.t .. > Failed 1/5 subtests > Wide character in print at t/utf8_text.t line 76. > > > The same test works fine on the same linux system with perl 5.10.1 .. > 5.24.0-RC5.
On 2016-05-08 11:16:28, HAYASHI wrote: Show quoted text
> Hi, > > Thank you for your report. > > I checked CPANTESTERS log. I see 12 passes with perl 5.8.9 and 1 pass > with perl 5.8.2. > Only 2 fails with perl 5.8.9. Both are yours. > > http://www.cpantesters.org/cpan/report/183b31ee-14e5-11e6-bddd- > 8c34573ac0e6 > This is the fail for this ticket. > > http://www.cpantesters.org/cpan/report/c55248a0-e389-11e5-9d60- > b001fdd2507e > This is the fail on one of t/readline.t tests which occurs only on > your environment. 64/66 fails on Term:ReadLine:Gnu-1.31 are failed on > the error. >
> > t/00checkver.t ... ok > > t/callback.t ..... ok > > t/history.t ...... ok > > t/readline.t ..... > > Failed 1/140 subtests > > t/utf8_binary.t .. ok > > The locale 'en_US.utf8' is not supported. Skipped... > > t/utf8_text.t .... ok
> > Interestingly http://www.cpantesters.org/cpan/report/183b31ee-14e5- > 11e6-bddd-8c34573ac0e6 passes the t/readline.t test. >
> > t/00checkver.t ... ok > > t/callback.t ..... ok > > t/history.t ...... ok > > ok Try `/bbbike/perl-5.8.9/bin/perl -Mblib t/readline.t verbose', > > if you will. > > t/readline.t ..... ok > > Use of uninitialized value in print at t/utf8_binary.t line 72, <$in> > > line 1. > > Use of uninitialized value in string eq at t/utf8_binary.t line 74, > > <$in> line 1. > > # Failed test 4 in t/utf8_binary.t at line 74 > > # t/utf8_binary.t line 74 is: ok($line eq "漢字2"); > > t/utf8_binary.t .. > > Failed 1/5 subtests > > Wide character in print at t/utf8_text.t line 76. > > t/utf8_text.t .... ok
> > Do you have any idea or hints? > > Regards, > > On Sun, 8 May 2016 07:07:47 GMT, SREZIC wrote:
> > Test output is: > > > > Use of uninitialized value in print at t/utf8_binary.t line 72, <$in> > > line 1. > > Use of uninitialized value in string eq at t/utf8_binary.t line 74, > > <$in> line 1. > > # Failed test 4 in t/utf8_binary.t at line 74 > > # t/utf8_binary.t line 74 is: ok($line eq "漢字2"); > > t/utf8_binary.t .. > > Failed 1/5 subtests > > Wide character in print at t/utf8_text.t line 76. > > > > > > The same test works fine on the same linux system with perl 5.10.1 .. > > 5.24.0-RC5.
As you can see in http://matrix.cpantesters.org/?dist=Term-ReadLine-Gnu%201.31;perl=5.8.9;reports=1 , there are no pass reports for the combination linux + perl 5.8.9. And the reason why it passes on my freebsd systems is simple: $ perl5.8.9 -Mblib t/utf8_binary.t 1..5 # Running under perl version 5.008009 for freebsd # Current time local: Sun May 15 18:19:01 2016 # Current time GMT: Sun May 15 16:19:01 2016 # Using Test.pm version 1.25 # I'm testing Term::ReadLine::Gnu version 1.31 ok 1 The locale 'en_US.utf8' is not supported. Skipped... ok 2 ok 3 ok 4 ok 5 On the linux system the en_US.utf8 locale exists, so the failing tests are not skipped. If I change the two occurences of "en_US.utf8" to "en_US.UTF-8" (which exists on this freebsd system), then the t/utf8_binary.t test fails on the freebsd system, too.
Hi, Thank you for your help. Yesterday I build perl 5.8.9 and could also reproduce the error. I've checked in the fix. https://sourceforge.net/p/perl-trg/code/517/ BTW as I wrote Show quoted text
> http://www.cpantesters.org/cpan/report/c55248a0-e389-11e5-9d60-b001fdd2507e This is the fail on one of t/readline.t tests which occurs only on your environment. 64/66 fails on Term:ReadLine:Gnu-1.31 are failed on the error.
a test for cursor movement fails only on your environment. But it passed in http://www.cpantesters.org/cpan/report/183b31ee-14e5-11e6-bddd-8c34573ac0e6 Do you see any reason of them? Attached is a test script which I sent to you before to debug the issue. Could you run it on the environment where t/readline.t fails and send the output to me? Thanks!
Subject: testSlaven.pl
# -*- perl -*- # readline.t - Test script for Term::ReadLine:GNU # # $Id: readline.t 475 2014-12-13 03:20:00Z hayashi $ # # Copyright (c) 1996 Hiroo Hayashi. All rights reserved. # # This program is free software; you can redistribute it and/or # modify it under the same terms as Perl itself. # Before `make install' is performed this script should be runnable with # `make test'. After `make install' it should work as `perl t/readline.t' BEGIN { print "1..140\n"; $n = 1; $ENV{PERL_RL} = 'Gnu'; # force to use Term::ReadLine::Gnu $ENV{LANG} = 'C'; } END {print "not ok 1\tfail to loading\n" unless $loaded;} # 'define @ARGV' is deprecated my $verbose = scalar @ARGV && ($ARGV[0] eq 'verbose'); use strict; use warnings; use vars qw($loaded $n); eval "use ExtUtils::testlib;" or eval "use lib './blib';"; use Term::ReadLine; use Term::ReadLine::Gnu qw(ISKMAP ISMACR ISFUNC RL_STATE_INITIALIZED); print "# Testing Term::ReadLine::Gnu version $Term::ReadLine::Gnu::VERSION\n"; $loaded = 1; print "ok 1\tloading\n"; $n++; # Perl-5.005 and later has Test.pm, but I define this here to support # older version. # MEMO: Since version TRL-1.10 Perl 5.7.0 has been required. So Test.pm # can be used now. my $res; my $ok = 1; sub ok { my $what = shift || ''; if ($res) { print "ok $n\t$what\n"; } else { print "not ok $n\t$what"; print @_ ? "\t@_\n" : "\n"; $ok = 0; } $n++; } ######################################################################## # test new method # stop reading ~/.inputrc not to change the default key-bindings. $ENV{'INPUTRC'} = '/dev/null'; # These tty setting affects GNU Readline key-bindings. # Set the standard bindings before rl_initialize() being called. # comment out since check_default_keybind_and_fix() takes care. # system('stty erase ^?') == 0 or warn "stty erase failed: $?"; # system('stty kill ^u') == 0 or warn "stty kill failed: $?"; # system('stty lnext ^v') == 0 or warn "stty lnext failed: $?"; # system('stty werase ^w') == 0 or warn "stty werase failed: $?"; my $t = new Term::ReadLine 'ReadLineTest'; $res = defined $t; ok('new'); my $OUT; if ($verbose) { $OUT = $t->OUT; } else { open(NULL, '>/dev/null') or die "cannot open \`/dev/null\': $!\n"; $OUT = \*NULL; $t->Attribs->{outstream} = \*NULL; } ######################################################################## # test ReadLine method $res = $t->ReadLine eq 'Term::ReadLine::Gnu'; ok('ReadLine method', "\tPackage name should be \`Term::ReadLine::Gnu\', but it is \`", $t->ReadLine, "\'\n"); ######################################################################## # test Features method my %features = %{ $t->Features }; $res = %features; ok('Features method',"\tNo additional features present.\n"); ######################################################################## # test Attribs method my $a = $t->Attribs; $res = defined $a; ok('Attrib method'); ######################################################################## # convert control charactors to printable charactors (ex. "\cx" -> '\C-x') sub toprint { join('', map{$_ eq "\e" ? '\M-': ord($_)<32 ? '\C-'.lc(chr(ord($_)+64)) : $_} (split('', $_[0]))); } my ($INSTR, $line); # simulate key input by using a variable 'rl_getc_function' $a->{getc_function} = sub { unless (length $INSTR) { print $OUT "Error: getc_function: insufficient string, \`\$INSTR\'."; undef $a->{getc_function}; return 0; } my $c = substr $INSTR, 0, 1; # the first char of $INSTR $INSTR = substr $INSTR, 1; # rest of $INSTR return ord $c; }; # This is required after GNU Readline Library 6.3. $a->{input_available_hook} = sub { return 1; }; # check some key binding used by following test sub is_boundp { my ($seq, $fname) = @_; my ($fn, $type) = $t->function_of_keyseq($seq); if ($fn) { return ($t->get_function_name($fn) eq $fname && $type == ISFUNC); } else { warn ("No function is bound for sequence \`", toprint($seq), "\'. \`$fname\' is expected,"); return 0; } } sub check_default_keybind_and_fix { my ($seq, $fname) = @_; if (is_boundp($seq, $fname)) { print "ok $n\t$fname is bound to " . toprint($seq) . "\n"; } else { # Try to fix the binding. But tty setting seems have precedence. $t->set_key($seq, $fname); if (is_boundp($seq, $fname)) { print "ok $n\tThe default keybinding for $fname was changed. Fixed.\n"; print "$fname is bound to " . toprint($seq) . "\n"; } else { print "not ok $n\t$fname cannot be bound to " . toprint($seq) . "\n"; $ok = 0; } } $n++; } check_default_keybind_and_fix("\cM", 'accept-line'); check_default_keybind_and_fix("\cF", 'forward-char'); check_default_keybind_and_fix("\cB", 'backward-char'); check_default_keybind_and_fix("\ef", 'forward-word'); check_default_keybind_and_fix("\eb", 'backward-word'); check_default_keybind_and_fix("\cE", 'end-of-line'); check_default_keybind_and_fix("\cA", 'beginning-of-line'); check_default_keybind_and_fix("\cH", 'backward-delete-char'); check_default_keybind_and_fix("\cD", 'delete-char'); check_default_keybind_and_fix("\cI", 'complete'); $INSTR = "abcdefgh\cM"; $line = $t->readline("self insert> "); $res = $line eq 'abcdefgh'; ok('self insert', $line); $INSTR = "\cAe\cFf\cBg\cEh\cH ij kl\eb\ebm\cDn\cM"; $line = $t->readline("cursor move> ", 'abcd'); # default string $res = $line eq 'eagfbcd mnj kl'; ok('cursor move', $line); $INSTR = "\cM"; print $t->readline("cursor move> ", 'abcd'), "\n"; $INSTR = "\cAe\cM"; print $t->readline("cursor move> ", 'abcd'), "\n"; $INSTR = "\cAe\cFf\cM"; print $t->readline("cursor move> ", 'abcd'), "\n"; $INSTR = "\cAe\cFf\cBg\cM"; print $t->readline("cursor move> ", 'abcd'), "\n"; $INSTR = "\cAe\cFf\cBg\cEh\cM"; print $t->readline("cursor move> ", 'abcd'), "\n"; $INSTR = "\cAe\cFf\cBg\cEh\cH\cM"; print $t->readline("cursor move> ", 'abcd'), "\n"; $INSTR = "\cAe\cFf\cBg\cEh\cH ij kl\cM"; print $t->readline("cursor move> ", 'abcd'), "\n"; $INSTR = "\cAe\cFf\cBg\cEh\cH ij kl\eb\ebm\cM"; print $t->readline("cursor move> ", 'abcd'), "\n"; $INSTR = "\cAe\cFf\cBg\cEh\cH ij kl\eb\ebm\cD\cM"; print $t->readline("cursor move> ", 'abcd'), "\n"; $INSTR = "\cAe\cFf\cBg\cEh\cH ij kl\eb\ebm\cDn\cM"; print $t->readline("cursor move> ", 'abcd'), "\n";
RT-Send-CC: srezic [...] iconmobile.net
Hi, Show quoted text
> Do you see any reason of them?
I converted http://www.cpantesters.org/distro/T/Term-ReadLine-Gnu.json to CSV format, and filter your result and sort it. The attached Excel is the result. I found interesting thing. We have pass and fail with almost same configuration and on a same day. The only difference is PERL_CPAN_REPORTER_CONFIG setting. Example pass http://www.cpantesters.org/cpan/report/aa711a18-14ec-11e6-a1d1-ba3175687904 fail http://www.cpantesters.org/cpan/report/df3251cc-14ec-11e6-a1d1-ba3175687904 ... - PERL_CPAN_REPORTER_CONFIG = /run/user/1023/cpansmoker/2016030600/cpanreporter_000_config.ini + PERL_CPAN_REPORTER_CONFIG = /run/user/1023/cpansmoker/2016030609/cpanreporter_000_config.ini ... Could your check the difference between the two configuration files? Regards,
Subject: CPANTESTERS.xlsx
Download CPANTESTERS.xlsx
application/vnd.openxmlformats-officedocument.spreadsheetml.sheet 988.1k

Message body not shown because it is not plain text.

Subject: diff.txt
--- pass 2016-05-20 00:53:41.731917300 +0900 +++ fail 2016-05-20 00:53:55.058329900 +0900 @@ -3,8 +3,8 @@ Raw Back From: Slaven Rezić SREZIC -Subject: PASS Term-ReadLine-Gnu-1.31 v5.20.1 GNU/Linux -Date: 2016-03-06T00:51:00Z +Subject: FAIL Term-ReadLine-Gnu-1.31 v5.20.1 GNU/Linux +Date: 2016-03-06T10:54:34Z This distribution has been tested as part of the CPAN Testers project, supporting the Perl programming language. See @@ -18,8 +18,14 @@ This is a computer-generated report for Term-ReadLine-Gnu-1.31 on perl 5.20.1, created by CPAN-Reporter-1.2015. -Thank you for uploading your work to CPAN. Congratulations! -All tests were successful. +Thank you for uploading your work to CPAN. However, there was a problem +testing your distribution. + +If you think this report is invalid, please consult the CPAN Testers Wiki +for suggestions on how to avoid getting FAIL reports for missing library +or binary dependencies, unsupported operating systems, and so on: + +http://wiki.cpantesters.org/wiki/CPANAuthorNotes Sections of this report: @@ -44,17 +50,24 @@ Running Mkbootstrap for Term::ReadLine::Gnu () chmod 644 "Gnu.bs" -PERL_DL_NONLAZY=1 "/opt/perl-5.20.1/bin/perl5.20.1" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t +PERL_DL_NONLAZY=1 "/opt/perl-5.20.1/bin/perl" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/00checkver.t ... ok t/callback.t ..... ok t/history.t ...... ok -ok Try `/opt/perl-5.20.1/bin/perl5.20.1 -Mblib t/readline.t verbose', if you will. -t/readline.t ..... ok +t/readline.t ..... +Failed 1/140 subtests t/utf8_binary.t .. ok t/utf8_text.t .... ok -All tests successful. -Files=6, Tests=244, 1 wallclock secs ( 0.16 usr 0.02 sys + 1.04 cusr 0.08 csys = 1.30 CPU) -Result: PASS + +Test Summary Report +------------------- +t/readline.t (Wstat: 0 Tests: 140 Failed: 1) + Failed test: 109 +Files=6, Tests=244, 3 wallclock secs ( 0.19 usr 0.01 sys + 0.99 cusr 0.04 csys = 1.23 CPU) +Result: FAIL +Failed 1/6 test programs. 1/244 subtests failed. +Makefile:1065: recipe for target 'test_dynamic' failed +make: *** [test_dynamic] Error 255 ------------------------------ PREREQUISITES @@ -92,21 +105,21 @@ PATH = /usr/local/bin:/usr/bin:/bin:/usr/local/sbin:/usr/sbin:/sbin:/home/cpansand/bin/linux-gnu:/home/cpansand/bin/sh:/home/cpansand/bin:/usr/games:/home/cpansand/devel:/home/eserte/src/srezic-misc/scripts PERL5LIB = PERL5OPT = - PERL5_CPANPLUS_IS_RUNNING = 26956 - PERL5_CPAN_IS_RUNNING = 26956 - PERL5_CPAN_IS_RUNNING_IN_RECURSION = 26831,26956 + PERL5_CPANPLUS_IS_RUNNING = 1438 + PERL5_CPAN_IS_RUNNING = 1438 + PERL5_CPAN_IS_RUNNING_IN_RECURSION = 353,1438 PERLDOC = -MPod::Perldoc::ToTextOverstrike PERL_BATCH = yes PERL_CANARY_STABILITY_NOPROMPT = 1 - PERL_CPAN_REPORTER_CONFIG = /run/user/1023/cpansmoker/2016030600/cpanreporter_000_config.ini + PERL_CPAN_REPORTER_CONFIG = /run/user/1023/cpansmoker/2016030609/cpanreporter_000_config.ini PERL_EXTUTILS_AUTOINSTALL = --defaultdeps SHELL = /bin/zsh TERM = screen - TMPDIR = /run/user/1023/cpansmoker/2016030600 + TMPDIR = /run/user/1023/cpansmoker/2016030609 Perl special variables (and OS-specific diagnostics, for MSWin32): - $^X = /opt/perl-5.20.1/bin/perl5.20.1 + $^X = /opt/perl-5.20.1/bin/perl $UID/$EUID = 1023 / 1023 $GID = 1023 1023 $EGID = 1023 1023 @@ -182,13 +195,13 @@ %ENV: PERL5LIB="" PERL5OPT="" - PERL5_CPANPLUS_IS_RUNNING="26956" - PERL5_CPAN_IS_RUNNING="26956" - PERL5_CPAN_IS_RUNNING_IN_RECURSION="26831,26956" + PERL5_CPANPLUS_IS_RUNNING="1438" + PERL5_CPAN_IS_RUNNING="1438" + PERL5_CPAN_IS_RUNNING_IN_RECURSION="353,1438" PERLDOC="-MPod::Perldoc::ToTextOverstrike" PERL_BATCH="yes" PERL_CANARY_STABILITY_NOPROMPT="1" - PERL_CPAN_REPORTER_CONFIG="/run/user/1023/cpansmoker/2016030600/cpanreporter_000_config.ini" + PERL_CPAN_REPORTER_CONFIG="/run/user/1023/cpansmoker/2016030609/cpanreporter_000_config.ini" PERL_EXTUTILS_AUTOINSTALL="--defaultdeps" @INC: /opt/perl-5.20.1/lib/site_perl/5.20.1/x86_64-linux
On 2016-05-19 12:22:52, HAYASHI wrote: Show quoted text
> Hi, >
> > Do you see any reason of them?
> > I converted http://www.cpantesters.org/distro/T/Term-ReadLine-Gnu.json > to CSV format, and filter your result and sort it. The attached Excel > is the result. > > I found interesting thing. > We have pass and fail with almost same configuration and on a same > day. The only difference is PERL_CPAN_REPORTER_CONFIG setting. > > Example > > pass http://www.cpantesters.org/cpan/report/aa711a18-14ec-11e6-a1d1- > ba3175687904 > fail http://www.cpantesters.org/cpan/report/df3251cc-14ec-11e6-a1d1- > ba3175687904 > > ... > - PERL_CPAN_REPORTER_CONFIG = > /run/user/1023/cpansmoker/2016030600/cpanreporter_000_config.ini > + PERL_CPAN_REPORTER_CONFIG = > /run/user/1023/cpansmoker/2016030609/cpanreporter_000_config.ini > ... > > Could your check the difference between the two configuration files?
The content of this config file is always the same. You can only tell that the first one was created roughly nine hours before the second one, that's all. Regards, Slaven
Subject: Re: [rt.cpan.org #114235] t/utf8_binary.t fails with perl 5.8.9
Date: Fri, 20 May 2016 17:02:18 +0900
To: bug-Term-ReadLine-Gnu [...] rt.cpan.org
From: 林宏雄 <hiroo.hayashi [...] computer.org>
Hi, Thank you for your reply. Show quoted text
>> Could your check the difference between the two configuration files?
Show quoted text
> The content of this config file is always the same. You can only tell
that the first one was created roughly nine hours before the second one, that's all. And we can say that the test-109 in t/readline fails (almost) once every two runs even under the exactly same condition only on your environment. If we cannot solve this issue, I have to skip the test on CPAN Testers test. According to the section "Can I tell if my module is being tested by an automated client (a.k.a. 'smoker')?" in http://wiki.cpantesters.org/wiki/CPANAuthorNotes, $ENV{AUTOMATED_TESTING} should be set, it is not set on your environment. Could you consider to set the environment variable on your environment, too? Regards, 2016-05-20 15:50 GMT+09:00 Slaven_Rezic via RT < bug-Term-ReadLine-Gnu@rt.cpan.org>: Show quoted text
> Queue: Term-ReadLine-Gnu > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=114235 > > > On 2016-05-19 12:22:52, HAYASHI wrote:
> > Hi, > >
> > > Do you see any reason of them?
> > > > I converted http://www.cpantesters.org/distro/T/Term-ReadLine-Gnu.json > > to CSV format, and filter your result and sort it. The attached Excel > > is the result. > > > > I found interesting thing. > > We have pass and fail with almost same configuration and on a same > > day. The only difference is PERL_CPAN_REPORTER_CONFIG setting. > > > > Example > > > > pass http://www.cpantesters.org/cpan/report/aa711a18-14ec-11e6-a1d1- > > ba3175687904 > > fail http://www.cpantesters.org/cpan/report/df3251cc-14ec-11e6-a1d1- > > ba3175687904 > > > > ... > > - PERL_CPAN_REPORTER_CONFIG = > > /run/user/1023/cpansmoker/2016030600/cpanreporter_000_config.ini > > + PERL_CPAN_REPORTER_CONFIG = > > /run/user/1023/cpansmoker/2016030609/cpanreporter_000_config.ini > > ... > > > > Could your check the difference between the two configuration files?
> > The content of this config file is always the same. You can only tell that > the first one was created roughly nine hours before the second one, that's > all. > > Regards, > Slaven >
-- Hiroo Hayashi
I released 1.32. 1.31 has another UTF8 issue on perl 5.8.x. In addition to locale fix ("en_US.utf8" to "en_US.UTF-8") I fixed it, too. And the test-109 in t/readline is skipped if AUTOMATED_TESTING is set or PERL_CPAN_REPORTER_CONFIG is defined.
RT-Send-CC: hiroo.hayashi [...] computer.org
On 2016-06-08 00:06:27, HAYASHI wrote: Show quoted text
> I released 1.32. > 1.31 has another UTF8 issue on perl 5.8.x. In addition to locale fix > ("en_US.utf8" to "en_US.UTF-8") I fixed it, too. > > And the test-109 in t/readline is skipped if AUTOMATED_TESTING is set > or PERL_CPAN_REPORTER_CONFIG is defined.
This is not a good idea. Tests should never get skipped if it happens to run on a "smoker", and possibly fail if running on a non-smoker system. If you have a test which is spurious failing then you should first use the $TODO variable. If you know that the test is failing just on some systems, then the $TODO variable may be set conditional to these systems (e.g. $TODO = "spurious fails" if $^O eq 'freebsd' && $] < 5.010). Regards, Slaven
Hi, Show quoted text
> This is not a good idea.
I agree with you. But I have no better idea. Show quoted text
> Tests should never get skipped if it happens > to run on a "smoker", and possibly fail if running on a non-smoker > system.
I don't know what "smoker" is. Show quoted text
> If you have a test which is spurious failing then you should > first use the $TODO variable. If you know that the test is failing > just on some systems, then the $TODO variable may be set conditional > to these systems (e.g. $TODO = "spurious fails" if $^O eq 'freebsd' && > $] < 5.010).
I thought of the TODO block, but did not use it. Because I have nothing TODO about this issue any more. from https://rt.cpan.org/Ticket/Display.html?id=114235#txn-1628198 Show quoted text
> The content of this config file is always the same. You can only tell that the first one was created roughly nine hours before the second one, that's all.
How can I express this condition? Regards,