First taste of Pylons: Rocky

For the last week or so I’ve been messing a little bit around with Pylons during my limited free time in order to be of at least of some help for Martin with his new project. I’m not really sure how I should feel about this framework, though.

---

A few points on this:

• It’s probably the same with everything: It takes a lot of time getting used to something different, so please bear with me here. I don’t know all that much yet about Pylons, so these are just my first impressions.
• I like the idea of re-using components where possible. Being able to use SQLAlchemy or any other DB-layer or template engine wherever I feel like it, is definitely nice.
• … but there should be some templates for paster bundled with Pylons that already do the basic configuration stuff for often-used components like SQLAlchemy.
• The debugging output is absolutely great. From what I’ve seen so far, Pylons creates a separate debugging page (with its own URL) whenever an exception stays unhandled, which should make debugging broken POST or AJAX calls way easier.
• The documentation is completely fragmented. You have (1) the docs section on pylonshq.com, then you have the WIKI (which is actually a superset of the documentation) and then you have the documentation for every component you want to use on its respective project site.
• Pylons has quite a low level feeling to it. While Django seems to be at the 6th floor above WSGI, Pylons seems to be on the first half-floor. Definitely one of the strongest points of Pylons in my opinion (but as I demonstrated in Vienna: I like explicit ;-)).
• I’m not yet sure, whether I like the deployment of projects or not. I’m simply not really a fan of setuptools.

Pylons so far looks really interesting, but given the lack of a more complete set of paster templates, it takes quite some time to get a project under way if you’re using for example SQLAlchemy and other non-core components.