Test fails three days after DST switch [rt.cpan.org #130875] #12
Description
Migrated from rt.cpan.org#130875 (status was 'open')
Requestors:
From srezic@cpan.org on 2019-11-02 09:30:31
:
Test number 95 fails three days after a DST switch to normal date. This can be reproduced with the help of Time::Fake:
$ env TZ=Europe/Berlin perl5.28.1 -MTime::Fake=1572399180 -Mblib t/date.t
...
not ok 95
# Test 95 got: "1572136380" (t/date.t at line 89)
# Expected: "1572139980"
# t/date.t line 89 is: ok($t, $time);
...
Here's a related CPAN Testers report:
http://www.cpantesters.org/cpan/report/e487ea72-fae1-11e9-aff9-042b1f24ea8f
From olaf@wundersolutions.com on 2019-11-03 17:00:29
:
On Sat Nov 02 05:30:31 2019, SREZIC wrote:
> Test number 95 fails three days after a DST switch to normal date.
> This can be reproduced with the help of Time::Fake:
>
> $ env TZ=Europe/Berlin perl5.28.1 -MTime::Fake=1572399180 -Mblib
> t/date.t
> ...
> not ok 95
> # Test 95 got: "1572136380" (t/date.t at line 89)
> # Expected: "1572139980"
> # t/date.t line 89 is: ok($t, $time);
> ...
>
> Here's a related CPAN Testers report:
> http://www.cpantesters.org/cpan/report/e487ea72-fae1-11e9-aff9-
> 042b1f24ea8f
Does this patch fix the issue? https://github.com/gisle/http-date/pull/4/files If so, I can look at getting a new release out soon.
From srezic@cpan.org on 2019-11-03 19:59:30
:
On 2019-11-03 12:00:29, OALDERS wrote:
> On Sat Nov 02 05:30:31 2019, SREZIC wrote:
> > Test number 95 fails three days after a DST switch to normal date.
> > This can be reproduced with the help of Time::Fake:
> >
> > $ env TZ=Europe/Berlin perl5.28.1 -MTime::Fake=1572399180 -Mblib
> > t/date.t
> > ...
> > not ok 95
> > # Test 95 got: "1572136380" (t/date.t at line 89)
> > # Expected: "1572139980"
> > # t/date.t line 89 is: ok($t, $time);
> > ...
> >
> > Here's a related CPAN Testers report:
> > http://www.cpantesters.org/cpan/report/e487ea72-fae1-11e9-aff9-
> > 042b1f24ea8f
>
> Does this patch fix the issue? https://github.com/gisle/http-
> date/pull/4/files If so, I can look at getting a new release out
> soon.
Looks good to me, at least it passes with the Time::Fake hack above.