Skip Menu |
 

This queue is for tickets about the App-TimeTracker CPAN distribution.

Report information
The Basics
Id: 125360
Status: resolved
Priority: 0/
Queue: App-TimeTracker

People
Owner: Nobody in particular
Requestors: jkeenan [...] cpan.org
Cc:
AdminCc:

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



Subject: multiple test failures with perl-5.27.11 on FreeBSD-11.1
Download (untitled) / with headers
text/plain 3.1k
Today I attempted to install App-TimeTracker against perl-5.27.11 on a FreeBSD-11.1 VM sitting on a FreeBSD-11 host. I encountered test failures similar to those we have seen in the past year in Dist::Zilla (https://github.com/rjbs/Dist-Zilla/issues/586) and XML::Atom::Simple::Feed (https://github.com/ap/XML-Atom-SimpleFeed/issues/3). These failures manifest as an exception, "Cannot determine local time zone", thought to be thrown by DateTime::TimeZone::Local. The failures occur in environments such as virtual machines and chroot-ed directory trees. Here is the output I got from the cpanm build.log: ##### Building App-TimeTracker t/00-load.t ...................... ok Cannot determine local time zone t/Command/core.t ................. Dubious, test returned 2 (wstat 512, 0x200) No subtests run # Test initial bug report Cannot determine local time zone t/Command/rt_73859_daychange.t ... Dubious, test returned 2 (wstat 512, 0x200) No subtests run t/Proto/config.t ................. ok t/Proto/run.t .................... ok t/Task/current.t ................. ok t/Task/helpers.t ................. ok Cannot determine local time zone t/Task/start.t ................... Dubious, test returned 2 (wstat 512, 0x200) No subtests run t/Task/storage_location.t ........ ok Cannot determine local time zone t/TimeTracker/datetime_coerce.t .. Dubious, test returned 2 (wstat 512, 0x200) No subtests run Cannot determine local time zone # Tests were run but no plan was declared and done_testing() was not seen. # Looks like your test exited with 2 just after 5. t/TimeTracker/find_task_files.t .. Dubious, test returned 2 (wstat 512, 0x200) All 5 subtests passed Cannot determine local time zone t/TimeTracker/helpers.t .......... Dubious, test returned 2 (wstat 512, 0x200) No subtests run t/Task/start.t ................... Dubious, test returned 2 (wstat 512, 0x200) No subtests run t/Task/storage_location.t ........ ok Cannot determine local time zone t/TimeTracker/datetime_coerce.t .. Dubious, test returned 2 (wstat 512, 0x200) No subtests run Cannot determine local time zone # Tests were run but no plan was declared and done_testing() was not seen. # Looks like your test exited with 2 just after 5. t/TimeTracker/find_task_files.t .. Dubious, test returned 2 (wstat 512, 0x200) All 5 subtests passed Cannot determine local time zone t/TimeTracker/helpers.t .......... Dubious, test returned 2 (wstat 512, 0x200) No subtests run Test Summary Report ------------------- t/Command/core.t (Wstat: 512 Tests: 0 Failed: 0) Non-zero exit status: 2 ##### I have found that assigning a valid timezone value to local $ENV{TZ} enables these tests to PASS. Although that doesn't solve the underlying problem, it *does* have the advantage that it prevents automated installers like cpan to not choke when asked to install your module's reverse dependencies. It will also enable "CPAN-river-focused" automated testing of CPAN modules against perl 5 blead to proceed -- which is a very important consideration now that we're about to release perl-5.28.0. I will submit a p.r. which I urge you to evaluate, apply and release to CPAN. Thank you very much. Jim Keenan
Subject: Re: [rt.cpan.org #125360] multiple test failures with perl-5.27.11 on FreeBSD-11.1
Date: Mon, 21 May 2018 14:52:39 +0200
To: James E Keenan via RT <bug-App-TimeTracker [...] rt.cpan.org>
From: Thomas Klausner <domm [...] plix.at>
Download (untitled) / with headers
text/plain 929b
Hi! On Sat, May 19, 2018 at 10:44:58PM -0400, James E Keenan via RT wrote: Show quoted text
> I have found that assigning a valid timezone value to local $ENV{TZ} > enables these tests to PASS. Although that doesn't solve the > underlying problem, it *does* have the advantage that it prevents > automated installers like cpan to not choke when asked to install your > module's reverse dependencies. It will also enable > "CPAN-river-focused" automated testing of CPAN modules against perl 5 > blead to proceed -- which is a very important consideration now that > we're about to release perl-5.28.0.
I've added some local $ENV{TZ} = 'UTC' to various tests in commit ea587270b790131f94965e6ff25dea0f5fc00219 Can you please check if this works for you? Thanks for the feedback & greetings, domm -- #!/usr/bin/perl http://domm.plix.at for(ref bless{},just'another'perl'hacker){s-:+-$"-g&&print$_.$/}
Download (untitled) / with headers
text/plain 2.2k
On Mon May 21 08:52:52 2018, domm@plix.at wrote: Show quoted text
> Hi! > > On Sat, May 19, 2018 at 10:44:58PM -0400, James E Keenan via RT wrote: > >
> > I have found that assigning a valid timezone value to local $ENV{TZ} > > enables these tests to PASS. Although that doesn't solve the > > underlying problem, it *does* have the advantage that it prevents > > automated installers like cpan to not choke when asked to install your > > module's reverse dependencies. It will also enable > > "CPAN-river-focused" automated testing of CPAN modules against perl 5 > > blead to proceed -- which is a very important consideration now that > > we're about to release perl-5.28.0.
> > I've added some local $ENV{TZ} = 'UTC' to various tests in commit > ea587270b790131f94965e6ff25dea0f5fc00219 > > Can you please check if this works for you? > > Thanks for the feedback & greetings, > domm >
Domm, thanks for your rapid response. This appears to work. Because your distro has a dependency on Dist-Zilla, and because the ultimate problem is somewhere in or above Dist-Zilla, I wasn't able to thoroughly test your commit (i.e., unable to dzil build, dzil test). However, I got this: ##### [pmsg-vm: App-TimeTracker] $ this_prove -lv t/*.t t/00-load.t .. ok 1 - require App::TimeTracker::Command::Core; ok 2 - require App::TimeTracker::Command::Git; ok 3 - require App::TimeTracker::Command::Overtime; ok 4 - require App::TimeTracker::Command::Post2IRC; ok 5 - require App::TimeTracker::Command::RT; ok 6 - require App::TimeTracker::Command::SyncViaGit; ok 7 - require App::TimeTracker::Command::TextNotify; ok 8 - require App::TimeTracker::Data::Task; ok 9 - require App::TimeTracker::Proto; ok 10 - require App::TimeTracker::Utils; 1..10 ok All tests successful. Files=1, Tests=10, 0 wallclock secs ( 0.01 usr 0.02 sys + 0.62 cusr 0.18 csys = 0.83 CPU) Result: PASS [pmsg-vm: App-TimeTracker] $ git show | head -1 commit ea587270b790131f94965e6ff25dea0f5fc00219 ##### So I think this is CPAN-releasable. The first perl-5.28.0 RC tarballs have just come out, so I'll be busy with them today. However, I hope to soon resume the search for the underlying bug. If we detect that, some of these work-arounds will be removable in a future release. Thank you very much. Jim Keenan
Download (untitled) / with headers
text/plain 2.7k
On Mon May 21 15:43:41 2018, JKEENAN wrote: Show quoted text
> On Mon May 21 08:52:52 2018, domm@plix.at wrote:
> > Hi! > > > > On Sat, May 19, 2018 at 10:44:58PM -0400, James E Keenan via RT > > wrote: > > > >
> > > I have found that assigning a valid timezone value to local > > > $ENV{TZ} > > > enables these tests to PASS. Although that doesn't solve the > > > underlying problem, it *does* have the advantage that it prevents > > > automated installers like cpan to not choke when asked to install > > > your > > > module's reverse dependencies. It will also enable > > > "CPAN-river-focused" automated testing of CPAN modules against perl > > > 5 > > > blead to proceed -- which is a very important consideration now > > > that > > > we're about to release perl-5.28.0.
> > > > I've added some local $ENV{TZ} = 'UTC' to various tests in commit > > ea587270b790131f94965e6ff25dea0f5fc00219 > > > > Can you please check if this works for you? > > > > Thanks for the feedback & greetings, > > domm > >
> > Domm, thanks for your rapid response. > > This appears to work. Because your distro has a dependency on Dist- > Zilla, and because the ultimate problem is somewhere in or above Dist- > Zilla, I wasn't able to thoroughly test your commit (i.e., unable to > dzil build, dzil test). However, I got this: > > ##### > [pmsg-vm: App-TimeTracker] $ this_prove -lv t/*.t > t/00-load.t .. > ok 1 - require App::TimeTracker::Command::Core; > ok 2 - require App::TimeTracker::Command::Git; > ok 3 - require App::TimeTracker::Command::Overtime; > ok 4 - require App::TimeTracker::Command::Post2IRC; > ok 5 - require App::TimeTracker::Command::RT; > ok 6 - require App::TimeTracker::Command::SyncViaGit; > ok 7 - require App::TimeTracker::Command::TextNotify; > ok 8 - require App::TimeTracker::Data::Task; > ok 9 - require App::TimeTracker::Proto; > ok 10 - require App::TimeTracker::Utils; > 1..10 > ok > All tests successful. > Files=1, Tests=10, 0 wallclock secs ( 0.01 usr 0.02 sys + 0.62 cusr > 0.18 csys = 0.83 CPU) > Result: PASS > [pmsg-vm: App-TimeTracker] $ git show | head -1 > commit ea587270b790131f94965e6ff25dea0f5fc00219 > ##### > > So I think this is CPAN-releasable. The first perl-5.28.0 RC tarballs > have just come out, so I'll be busy with them today. However, I hope > to soon resume the search for the underlying bug. If we detect that, > some of these work-arounds will be removable in a future release. > > Thank you very much. > Jim Keenan
I just tried the current version of App::TimeTracker (3.003) on FreeBSD 12.1 (running in a Virtualbox VM on a Debian buster host system) and with perl 5.28.3 and Dist::Zilla 6.015. I was unable to reproduce the error mentioned in this ticket and therefore assume that the issue no longer exists. @kid51: would you be able to confirm this? Thanks! Cheers, ptc
On Mon Oct 05 07:32:53 2020, PTC wrote: Show quoted text
> On Mon May 21 15:43:41 2018, JKEENAN wrote:
> > On Mon May 21 08:52:52 2018, domm@plix.at wrote:
> > > Hi! > > > > > > On Sat, May 19, 2018 at 10:44:58PM -0400, James E Keenan via RT > > > wrote: > > > > > >
> > > > I have found that assigning a valid timezone value to local > > > > $ENV{TZ} > > > > enables these tests to PASS. Although that doesn't solve the > > > > underlying problem, it *does* have the advantage that it prevents > > > > automated installers like cpan to not choke when asked to install > > > > your > > > > module's reverse dependencies. It will also enable > > > > "CPAN-river-focused" automated testing of CPAN modules against > > > > perl > > > > 5 > > > > blead to proceed -- which is a very important consideration now > > > > that > > > > we're about to release perl-5.28.0.
> > > > > > I've added some local $ENV{TZ} = 'UTC' to various tests in commit > > > ea587270b790131f94965e6ff25dea0f5fc00219 > > > > > > Can you please check if this works for you? > > > > > > Thanks for the feedback & greetings, > > > domm > > >
> > > > Domm, thanks for your rapid response. > > > > This appears to work. Because your distro has a dependency on Dist- > > Zilla, and because the ultimate problem is somewhere in or above > > Dist- > > Zilla, I wasn't able to thoroughly test your commit (i.e., unable to > > dzil build, dzil test). However, I got this: > > > > ##### > > [pmsg-vm: App-TimeTracker] $ this_prove -lv t/*.t > > t/00-load.t .. > > ok 1 - require App::TimeTracker::Command::Core; > > ok 2 - require App::TimeTracker::Command::Git; > > ok 3 - require App::TimeTracker::Command::Overtime; > > ok 4 - require App::TimeTracker::Command::Post2IRC; > > ok 5 - require App::TimeTracker::Command::RT; > > ok 6 - require App::TimeTracker::Command::SyncViaGit; > > ok 7 - require App::TimeTracker::Command::TextNotify; > > ok 8 - require App::TimeTracker::Data::Task; > > ok 9 - require App::TimeTracker::Proto; > > ok 10 - require App::TimeTracker::Utils; > > 1..10 > > ok > > All tests successful. > > Files=1, Tests=10, 0 wallclock secs ( 0.01 usr 0.02 sys + 0.62 > > cusr > > 0.18 csys = 0.83 CPU) > > Result: PASS > > [pmsg-vm: App-TimeTracker] $ git show | head -1 > > commit ea587270b790131f94965e6ff25dea0f5fc00219 > > ##### > > > > So I think this is CPAN-releasable. The first perl-5.28.0 RC > > tarballs > > have just come out, so I'll be busy with them today. However, I hope > > to soon resume the search for the underlying bug. If we detect that, > > some of these work-arounds will be removable in a future release. > > > > Thank you very much. > > Jim Keenan
> > > I just tried the current version of App::TimeTracker (3.003) on > FreeBSD 12.1 (running in a Virtualbox VM on a Debian buster host > system) and with perl 5.28.3 and Dist::Zilla 6.015. I was unable to > reproduce the error mentioned in this ticket and therefore assume that > the issue no longer exists. > > @kid51: would you be able to confirm this? Thanks! >
Yes, I was able to install App::TimeTracker against perl-5.30 on FreeBSD-12 using cpanm as the installer.


This service is sponsored and maintained by Best Practical Solutions and runs on Perl.org infrastructure.

Please report any issues with rt.cpan.org to rt-cpan-admin@bestpractical.com.