[CDBI] Class::DBI vs DBIx::Class

Ryan Tate lists at ryantate.com
Wed Jan 18 23:57:43 GMT 2006


On 1/17/06, Matt S Trout <dbix-class at trout.me.uk> wrote:
> One of the issues with consolidated documentation is that DBIx::Class
> isn't "a module", it's a collection of classes and components that snap
> together to form as much (or as little) of an ORM as you need for your
> project.

Idea: The starting place for documentation, be it DBIx::Class or
DBIx::Class::Manual::Intro, makes an editorial assumption of the best
(or "best" if you're feeling relatively relative)  way to build an ORM
out of DBIx::Class for the typical ("typical") new user.

I would argue that the typical new user is someone who wants more
power and fance (fanciness) than vanilla CDBI (automatic joins!) but
is not interested in using the CDBI API, either out of desire for
fresh start or because he never learned it in the first place and went
straight to DBIx::Class. This paragraph is even more IMHO than the
preceding.

Having made some opinionated decisions, the perldoc then proceeds to
lay out the options briefly, gives lots of preliminary details and
examples of how to do things the Most Favored Way, then points the
user to another module for even more info on Most Favored Way.

Then under a subsequent =head1, the perldoc provides just a pointer
where the user can go read about the Less Favored Way(s).

Just an idea.

> Documentation should start to settle down after 0.05; most of the
> user-facing APIs now have at least skeleton reference documentation and
> there's some fantastic introductory stuff in the works.

Sounds cool!


> Would you mind either posting your e-mail to the DBIx::Class list (if you
> don't want to subscribe I'll happily approve the posting) or letting me
> forward it o
n? The documentation contributors would probably be interested
> to read it.

I'd be happy to let you forward it. Just note that

s/even right up there in the info/even right up there in the intro/

Cheers
rt




More information about the ClassDBI mailing list