Eric Wilhelm wrote:
> Why is it "die 'OS unsupported'" in one case
My understanding is that there was already special-case code to support
that, but adding loads more bits of magic text would be unwieldy. It
would also require everyone to upgrade whenever new magic text appears
on the scene, whi
# from David Cantrell
# on Monday 22 October 2007 12:50:
>> then "inventing" a
>> Makefile.PL and trying to run that.
>
>Given the above ubiquitous bug in Build.PL, and that Module::Build can
>be a real pain in the arse to get working properly,
Othe
On Mon, Oct 22, 2007 at 10:52:47AM -0700, Eric Wilhelm wrote:
> # from David Cantrell
> # on Monday 22 October 2007 09:58:
> >> This is a specific sort of failure, and as far as I'm concerned it
> >> is an improperly configured machine.
> >How is it improperly configured?
> It is ignoring Build.PL
chromatic wrote:
>> This is a specific sort of failure, and as far as I'm concerned it is an
>> improperly configured machine. I'm not going to include a redundant
>> Makefile.PL just to support people who don't upgrade their tools. It's
>> no different than not supporting perl 5.005.
I believe
# from David Cantrell
# on Monday 22 October 2007 09:58:
>> This is a specific sort of failure, and as far as I'm concerned it
>> is an improperly configured machine.
>
>How is it improperly configured?
It is ignoring Build.PL (and the requirement for Module::Build in the
META.yml (and the instr
On Mon, Oct 22, 2007 at 09:50:42AM -0700, Eric Wilhelm wrote:
> # from David Cantrell
> >And if the build process had a create a Makefile.PL for you, then in
> > my opinion it's pretty important to let you know about that if it
> > causes problems. It shows that the module wasn't properly packaged,
# from chromatic
# on Monday 22 October 2007 10:15:
>On Monday 22 October 2007 09:50:42 Eric Wilhelm wrote:
>> I'm not going to include a
>> redundant Makefile.PL just to support people who don't upgrade their
>> tools. It's no different than not supporting perl 5.005.
>
>Normally I'm all for tha
On Monday 22 October 2007 09:50:42 Eric Wilhelm wrote:
> This is a specific sort of failure, and as far as I'm concerned it is an
> improperly configured machine. I'm not going to include a redundant
> Makefile.PL just to support people who don't upgrade their tools. It's
> no different than not
# from David Cantrell
# on Monday 22 October 2007 08:48:
>And if the build process had a create a Makefile.PL for you, then in
> my opinion it's pretty important to let you know about that if it
> causes problems. It shows that the module wasn't properly packaged,
> for some definition of properl
On Sun, Oct 21, 2007 at 09:59:11AM -0700, Eric Wilhelm wrote:
> It would be nice if the reports were somehow queryable WRT machine
> config and failure mode. Sorting through a slew of reports is a real
> pain when some have no usable information (like an error message) and
> others are just po
On Mon, Oct 22, 2007 at 08:50:16AM -0400, David Golden wrote:
> I like that idea -- it would allow (theoretically) some analysis of
> PASS and UNTESTABLE to find which actual prerequisites cause things to
> fail on different platforms.
This information will be available once I get round to adding
On 10/22/07, Eric Wilhelm <[EMAIL PROTECTED]> wrote:
> # from David Golden
> # on Sunday 21 October 2007 20:18:
>
> >Your META.yml says your module needs "Foo::Bar 1.23". Someone doesn't
> >have Foo::Bar installed. Your tests fail. Did they fail because
> >Foo::Bar was missing? Or because one or mo
* David Cantrell <[EMAIL PROTECTED]> [2007-10-22 13:40]:
> I hope, however, that the CP6AN will make it obsolete by
> having:
And as an aside, 6PAN will hopefully have a `configure_requires`
mechanism right out of the gate; its hitherto absence from the
Perl 5 toolchain is a critical hole.
Regard
On Sun, Oct 21, 2007 at 04:48:23PM +0200, demerphq wrote:
> As an aside, it seems to me that both Devel::CheckLib and
> configure_requires suffer from a fatal flaw in that they do not solve
> the problem for existing modules.
Agreed.
There are actually two problems to solve:
* not sending bogus
14 matches
Mail list logo
