Skip to content

Conversation

@RHPizzarro
Copy link

@RHPizzarro RHPizzarro commented Sep 10, 2021

When trying to install DateTimeX::Easy on both CentOS and AWS Linux containers starting around (2021-07-08), the following test error is encountered (note: you can install with the force option, but of course this is not preferrable):

#   Failed test at t/03-parse.t line 32.
#          got: 'America/New_York'
#     expected: '-0500'
# Looks like you failed 1 test of 18.
t/03-parse.t ..... Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/18 subtests

NOTE: Thanks to Petr Pisar for the following investigation / fix
After upgrading DateTime-Format-Flexible from 0.33 to 0.34, a test fails like this:

t/03-parse.t ..... 1/?
#   Failed test at t/03-parse.t line 32.
#          got: 'America/New_York'
#     expected: '-0500'
# Looks like you failed 1 test of 18.
t/03-parse.t ..... Dubious, test returned 1 (wstat 256, 0x100)

That's because new DateTime::Format::Flexible started to support time zones enclosed in parenthesis and the teste string was "Wed, 9 Nov 1994 09:50:32 -0500 (EST)".

I conclude that the test is broken because $DateTime->time_zone->name is allowed to return a label instead of an offest. So I corrected thetest to compare an offset value.

The reason why this emerged now is that DateTimeX::Easy ignores almost all DateTime::Format::Flexible return values because it plays with time zones a bad way.

After upgrading DateTime-Format-Flexible from 0.33 to 0.34, a test
fails like this:

    t/03-parse.t ..... 1/?
    #   Failed test at t/03-parse.t line 32.
    #          got: 'America/New_York'
    #     expected: '-0500'
    # Looks like you failed 1 test of 18.
    t/03-parse.t ..... Dubious, test returned 1 (wstat 256, 0x100)

That's because new DateTime::Format::Flexible started to support
time zones enclosed in parenthesis and the teste string was
"Wed, 9 Nov 1994 09:50:32 -0500 (EST)".

I conclude that the test is broken because $DateTime->time_zone->name
is allowed to return a label instead of an offest. So I corrected the
test to compare an offset value.

The reason why this emerged now is that DateTimeX::Easy ignores almost
all DateTime::Format::Flexible return values because it plays with
time zones a bad way.
@RHPizzarro
Copy link
Author

Note: this bug was reported here: https://rt.cpan.org/Public/Bug/Display.html?id=137397

@michael-stevens
Copy link

Also seeing this issue, would love to see this fixed merged.

@jjn1056
Copy link

jjn1056 commented Nov 7, 2021

Just an update for whoever is watching, I emailed the author today to see if they had any thoughts on this, and I also offered to take COMAINT on the project if they are no longer in Perl and would rather hand this off. I'll dig back in a bit if I hear something.

@RHPizzarro
Copy link
Author

Thanks for the follow up @jjn1056 - I tried reaching out to the module owner (robertkrimen Robert Krimen) with no response. I also reached out to someone who has contributed to this repo (Leo Lapworth) as well and he did reply with this response about adopting a module so I think that is our best option at this point:

I don't have release permissions for that module...

I'd suggest look at https://www.cpan.org/misc/cpan-faq.html#How_maintain_module and https://www.cpan.org/misc/cpan-faq.html#How_adopt_module

@abraxxa
Copy link

abraxxa commented Jan 19, 2022

We're affected by this issue as well.

@jjn1056: did you get a response from Robert?

@jjn1056
Copy link

jjn1056 commented Apr 22, 2022

@abraxxa @RHPizzarro ok so I now have comaint on the CPAN module. If one of you can fork this Repo and prep a tarball and get it to me I can put it on CPAN

@abraxxa
Copy link

abraxxa commented May 2, 2022

Will this stay the primary git repo for this dist or who will fork and maintain it? If noone steps up we should get rid of this dependency in DBIx::Class::Schema::PopulateMore.

@jjn1056
Copy link

jjn1056 commented Aug 7, 2022

@abraxxa @RHPizzarro This is fixed on my fork and released. Please shout out if you seem more trouble -jn

@RHPizzarro
Copy link
Author

RHPizzarro commented Aug 8, 2022 via email

@RHPizzarro
Copy link
Author

Hi again @jjn1056 - one small thing on the new distribution. The UID/GIDs of the files in the distribution are really high values (9348308, 2174114). We ran into an error with deploying a Docker container that uses the new package where we didn't cleanup the source .gz package on build. Since these UID's could not be mapped to the to the user namespace the Docker pull failed. Our fix was easy...just delete the source .gz files, but something maybe worth cleaning up.

Thanks again for making this fix happen!

@abraxxa
Copy link

abraxxa commented Aug 23, 2022

Thanks @jjn1056 !
It seems you missed the version number in the Changes file though.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants