develooper Front page | perl.module-authors | Postings from June 2008

Re: New CPANTS metrics

Thread Previous | Thread Next
Eric Roode
June 10, 2008 09:05
Re: New CPANTS metrics
Message ID:
On Tue, Jun 10, 2008 at 11:47 AM, Andy Lester <> wrote:
> On Jun 10, 2008, at 10:41 AM, David Cantrell wrote:
>> If you see no value in it, just ignore it.  I'm sure it will do wonders
>> for your blood pressure.
> I guess that's my very point.  Here's this entire subsystem that exists to
> supposedly give information to authors and potential users about the
> relative quality of the code, and yet the attitude that comes out is "Eh, we
> like it, you don't have to like it."
> Take a lesson from Perl::Critic and explain the reasoning behind the
> policies.

Well put, Andy.

> It says that Mech's META.yml doesn't conform to a known spec (At least
> that's what I think the arcane code in the hover box tells me). So what? Why
> do I as an author care?

CPANTS says that most of my modules have a META.yml that doesn't
conform to a known spec.  Yet I created the META.yml for all of these
modules with 'perl Build dist' or 'make dist'.  I know that META.yml
is useful for machine-driven testing and distribution mechanisms....
but if Module::Build or ExtUtils::MakeMaker isn't making conformant
META.yml files, I can't do much about that.

> It's only overstating slightly to say that the entire CPANTS structure seems
> to be built upon the premise of "These are things that should be a certain
> way because I say so," whoever "I" may happen to be.
> That's not to say that the things checked for aren't worthwhile, but nothing
> says WHY they are worthwhile.
> Further, and worse, it's presented as "You should just know this stuff and
> appreciate us for telling you."

I concur on these observations -- I get the same impression that Andy does.

-- Eric

Thread Previous | Thread Next Perl Programming lists via nntp and http.
Comments to Ask Bjørn Hansen at | Group listing | About