My showdown: Ruby/Rails, Perl/Catalyst, Python/Turbogears

My personal showdown:

Using Ruby on Rails:
I was pretty productive right away. Documentation though for the Rails
APIs and stuff is not as good as I’m use to in Perl (particulariy CPAN).
No being critical, just saying the Rails documentation left me hanging
more than once and guessing as to what I was doing wrong.

Production deployment is a bit much. I’ve found using Litespeed and
LSAPI was the easiest to get to production than the Apache + mod_proxy +
mongrel cluster solution.

Using Turbogears (Python)
Pretty productive here as well right off the bat. Documentation is
sparse at best and this hampers you when you want to step out of the
box.

Production is pretty easy, using CherryPy and Apache is fairly easy to
get going. Turbogears seems to be in perma-beta though. Also its
confusing in the sense of which template system you want to use, which
ORM to use (SQLAlchemy, SQLOBject, etc.) I wish they would just pick
one and stick with it. Rails uses ActiveRecord and thats good enough -
I can focus on just that.

Perl Catalyst
Runs fast. Not the easiest to get setup and running though. Perl if
written correctly can be readable, so I’d say its productive.
Documentation is somewhat sparse but improving. Perl never ceases to
amaze me in its execution speed, even poorly written Perl seems to
scream :wink:

Haven’t put it to production yet.

Anyone else messed around with these?

In the end I think I can get to production faster in Rails, but if I was
to have a site that had a lot of traffic I think I might feel safer with
Perl. CPAN is an immense resource and RubyForge isn’t there yet, nor is
the documentation. But there is a lot of working going on.