Zilker Park. Barton Springs. 6th Street. Brushy Creek. Torchy's Tacos. Rudy's BBQ. Hula Hut. Carlos & Charlies on Lake Travis. The Iguana Grill, also on Lake Travis, home of the world's best salsa. Threadgill's Downtown. Waterloo Records. Whole Foods downtown. The Captial. Alamo Draft House. UT Campus. Fonda San Miguel.

I would go with Django over web2py. If you're looking for more of a microframework, I would look at Flask.

Having built enterprise applications in Rails and Django, I would recommend Django over Rails 99.999% of the time. Django is a very homogeneous framework; it's easy to use, learn, deploy, maintain and has built-in ways to do just about anything you'd need to do in a web application.

Rails is more like a piece of flypaper, that you trow a bunch of stuff (gems) at and pray they stick. You can't even upload a file in Rails without using a plugin.

Rails

As a former senior developer at Razorfish in Austin, I have only worked on 1, no that's not a typo, 1 Java project that wasn't an utter failure.

On the other hand, I've yet to work on a Django project that was delivered late or ran over budget and was easily 10x more maintainable.

I put my money on Python and Django in 2008, and haven't looked back.

I work on big insurance java project (initiated before 2003, 300 people in development, now more than 450 people). This is not a lot of little independent projects - it's ONE Project with more than 75 components! It's high quality software and we make crazy amount of user stories get every sprint out. Of course, we have errors, RTE's - we are all humans. I worked also for others companies and write PHP and Python/Django for years. Failure has nothing with language or framework or tools - it's always human fail. Sorry for my bad english.

I agree. A language or framework isn't a reason for failure, however, it's a pattern that I observed at Razorfish for a few years. Maybe the devs we had just weren't very good :)