I would also like to point out to the CPAN Testers Matrix page, where
it's much easier to separate "useful" from "useless" reports. See

http://bbbike.radzeit.de/~slaven/cpantestersmatrix.cgi?dist=OpenOffice-OODoc&maxver=

or just follow the link from the distribution's page at
search.cpan.org or kobesearch.cpan.org.

Regards,
     Slaven

"David Golden" <[email protected]> writes:

> Dear Jean-Marie,
>
> The "NA" grade is defined to means "Not applicable -- not supported on this
> platform or perl" -- so an NA result from a 5.6 tester is correct and
> informative for anyone curious if your module runs on 5.6.
>
> I'm sorry to hear that you see it as "pollution" -- it is actually a
> confirmation of the fact that your module is well-written and is explicit
> about the version of Perl it requires, rather than letting a 5.6 user try it
> and fail tests instead.
>
> While I encourage you to leave the NA reports alone as useful information,
> in the spirit of TIMTOWTDI, I will mention that you can avoid having reports
> sent in any particular case by using the "exit 0" hack.  In this case, put
> the following at the top of your Makefile.PL before (or in place of) "use
> 5.008":
>
>     BEGIN { exit 0 if $] < 5.008 }
>
> If you exit with a 0 exit value and do not create a Makefile, then both
> CPAN.pm and CPANPLUS will just stop processing and no CPAN Testers report is
> sent.
>
> Sincerely,
> David Golden
>
> On Wed, Dec 10, 2008 at 5:50 AM, Ask Bjørn Hansen <[email protected]> wrote:
>
>>
>> On Dec 9, 2008, at 2:37, JMGDOC wrote:
>>
>> Hi Jean-Marie,
>>
>> I cc'ed a couple of people who'll know the answers to your questions.  :-)
>>
>>  - ask
>>
>>  Hello,
>>>
>>> I know my question has probably been asked and answered many times but...
>>> I regularly see "N/A" automated test results on cpantesters for tests that
>>> should never be done. More precisely, while my OpenOffice::OODoc module
>>> explicitly requires Perl 5.8, a few tests with Perl 5.6 are published each
>>> time I upload a new release. These tests pollute the cpantesters' page
>>> with
>>> useless and annoying N/As.
>>>
>>> So, is there a way to prevent a automated test to be executed/published
>>> when
>>> the requirements are not met ?
>>>
>>> Thanks for your help
>>> Regards,
>>>
>>> --
>>> Jean-Marie Gouarné (JMGDOC)
>>> http://search.cpan.org/dist/OpenOffice-OODoc
>>>
>>
>>

-- 
Slaven Rezic - slaven <at> rezic <dot> de

    tkrevdiff - graphical display of diffs between revisions (RCS, CVS or SVN)
    http://ptktools.sourceforge.net/#tkrevdiff

Reply via email to