On Mon, Feb 10, 2003 at 03:23:47PM +0000, Leon Brocard said:
> But if you enforce more controls on CPAN, surely people are less
> likely to upload code?

IS this necessarily a bad thing though? Would less high quality code be
better than more low quality code? To be honest I couldn't tell you the
answer although I recently saw someone complain that "all the useful
stuff from CPAN has now been reimplemented by the core Python developers
and it's better quality too" or something similar.

> At the moment CPAN is a seperate entity than the modules list. You
> could argue that everyone ignores the modules list thesedays thanks to
> search.cpan.org and CPANPLUS.

This is the same argument as 'I don't need bookmarks because I have
Google' to a certain extent but, unlike the Google argument I'm mildly
inclined to agree as long as there was a better search. 

  
> To do this properly would need perl to be changed.

Would it? I'm talking about the problem Andy had in that he wanted to
have two versions of Template::Toolkit  - one for Apache and one for
Apache2. I think solving that would involve a different CPAN structure
not a different Perl.

How hard would changing the Perl was of loading and installing modules
be so that you have

$PERL_LIB_PATH/-+-- Module-1.1
                |
                +-- Module-1.3
               

And have

        use Module;     # get's v1.3
        use Module-1.1; # get's v1.1


ignoring legacy problems.



 
> Does PAR help?
> http://search.cpan.org/author/AUTRIJUS/PAR/

No.

Well, yes but I'd rather not distribute a huge quantity of Module with
an app. I still like to be able to install stuff like CPAN normally -
handling dependencies etc etc.

I'm talking more about installing something like, say, a webmail app -
I'd like to be able to do something like

nextpan> install foomail

and have it install in the right place. That might be turning CPAN into
YA Package Management system though. 


 
> > Better integration with various packaging systems:
> 
> CPANPLUS 0.50 is attempting to do this.

See, it's happening already! <grin>


> Of course, the real question is what is the problem you're trying to
> solve?

Not sure to be honest. Making sure that CPAN doesn't rest on its laurels
I suppose. Although If It Ain't Broke then Don't Fix It. But Every cloud
has a silver lining. On the other hand he who hesitates is lost and life
is very short and there's no time For fussing and fighting my friend


Or something.

Simon



Reply via email to