Bug #91719 for Git-Repository-Plugin-Log: Test failure on NetBSD and FreeBSD 10

Sun Dec 29 15:48:52 2013 SREZIC [...] cpan.org - Ticket created

Subject:

Test failure on NetBSD and FreeBSD 10

The module has a test failure in t/41-multiline-log.t on NetBSD and FreeBSD 10 systems (but not FreeBSD 9). A sample fail report: http://www.cpantesters.org/cpan/report/44a74c42-492a-11e3-8a37-e2882d12fa71 Maybe it's related to the git version? In this case it would probably be good to output the git version in the test suite, so one can analyze the problem with CPAN::Testers::ParseReport... Regards, Slaven

Mon Dec 30 17:22:24 2013 book [...] cpan.org - Correspondence added

Subject:	Re: [rt.cpan.org #91719] Test failure on NetBSD and FreeBSD 10
Date:	Mon, 30 Dec 2013 23:22:11 +0100
To:	Slaven_Rezic via RT <bug-Git-Repository-Plugin-Log [...] rt.cpan.org>
From:	book [...] cpan.org

On Sun, Dec 29, 2013 at 03:48:52PM -0500, Slaven_Rezic via RT wrote: Show quoted text

> > The module has a test failure in t/41-multiline-log.t on NetBSD and FreeBSD 10 systems (but not FreeBSD 9). A sample fail report: http://www.cpantesters.org/cpan/report/44a74c42-492a-11e3-8a37-e2882d12fa71

Actually, it's related to the shift-jis encoding, that is used for that specific commit (in the commit message). I'm not exactly sure what happens with FreeBSD. Is there a way to know which encodings are available? Maybe "shift-jis" is not the right name? Someone told me the canonical name was "SJIS". It could also be that the encoding is not installed on those systems? Then I would need a way to check its availability. Show quoted text

> Maybe it's related to the git version? In this case it would probably be good to output the git version in the test suite, so one can analyze the problem with CPAN::Testers::ParseReport... >

Indeed, the version is missing in the report. I'll add it in the next release. -- Philippe Bruhat (BooK) You can always buy another house but you cannot put a price on a home. (Moral from Groo The Wanderer #63 (Epic))

Mon Dec 30 17:22:24 2013 The RT System itself - Status changed from 'new' to 'open'

Mon Dec 30 19:14:01 2013 SREZIC [...] cpan.org - Correspondence added

On 2013-12-30 17:22:24, BOOK wrote: Show quoted text

> On Sun, Dec 29, 2013 at 03:48:52PM -0500, Slaven_Rezic via RT wrote:

> > > > The module has a test failure in t/41-multiline-log.t on NetBSD and > > FreeBSD 10 systems (but not FreeBSD 9). A sample fail report: > > http://www.cpantesters.org/cpan/report/44a74c42-492a-11e3-8a37- > > e2882d12fa71

> > Actually, it's related to the shift-jis encoding, that is used for > that specific commit (in the commit message). I'm not exactly sure > what happens with FreeBSD. Is there a way to know which encodings > are available?

"locale -a" is returning a list of locales. Works on Linux and FreeBSD. Show quoted text

> Maybe "shift-jis" is not the right name? Someone told > me the canonical name was "SJIS".

This system has "ja_JP.SJIS" available. Regards, Slaven Show quoted text

> > Maybe it's related to the git version? In this case it would probably > > be good to output the git version in the test suite, so one can > > analyze the problem with CPAN::Testers::ParseReport... > >

> > Indeed, the version is missing in the report. I'll add it in the next > release.

Tue Dec 31 11:38:20 2013 book [...] cpan.org - Correspondence added

Subject:	Re: [rt.cpan.org #91719] Test failure on NetBSD and FreeBSD 10
Date:	Tue, 31 Dec 2013 17:38:07 +0100
To:	Slaven_Rezic via RT <bug-Git-Repository-Plugin-Log [...] rt.cpan.org>
From:	book [...] cpan.org

On Mon, Dec 30, 2013 at 07:14:01PM -0500, Slaven_Rezic via RT wrote: Show quoted text

> > > > Actually, it's related to the shift-jis encoding, that is used for > > that specific commit (in the commit message). I'm not exactly sure > > what happens with FreeBSD. Is there a way to know which encodings > > are available?

> > "locale -a" is returning a list of locales. Works on Linux and FreeBSD.

Yes. But that's only the locales (especially translations) configured for the users. That's different from the encodings supported, I think. Show quoted text

> > Maybe "shift-jis" is not the right name? Someone told > > me the canonical name was "SJIS".

> > This system has "ja_JP.SJIS" available.

I suppose this could be the issue for me. I've regenerated a new test repository using "SJIS" instead of "shift-jis" as the encoding name. Do you mind trying to run the test suite of the "sjis" branch? https://github.com/book/Git-Repository-Plugin-Log/tree/sjis Thanks, -- Philippe Bruhat (BooK) A little work with your brain can save you a lot of work with your body. (Moral from Groo The Wanderer #40 (Epic))