Jesse Vincent <jesse@fsck.com> writes:
> In 2011, I consider teaching new Perl programmers that they should
> prefer to build their own functionality when there are many better
> options on CPAN to be actively harmful to our users and to our
> community.
You hit the nail on the head. There are many better options on CPAN, and
the problem is that there are many better options on CPAN.
When I search for 'Class' on CPAN, I get more than 5000 hits. It will
take more time to sort these out than to write a moderate application.
So I'll try a couple of them, and if I don't find what I'm looking for
I'll write my own. Maybe I'll put it on CPAN, $hits{Class}++.
Want to use XML? Again, over 5000 hits. XML::Bare, XML::Easy,
XML::Trivial, XML::Simple, XML::Light, XML::MyXML, XML::Quick.
Same story.
This is where TIMTOWTDI backfires. And Perl loses.
OO frameworks, XML data handling, and much more, were experimental toys
15 years ago, now they are core parts of application development.
So let's decide on a good OO framework (XML, ...) and tell everone to
use this since it is the best. And include it in core Perl so that it is
guaranteed to be installed and maintained everywhere. And then point
people to CPAN for more -- and given the requirements, better -- options.
-- Johan
Thread Previous
|
Thread Next