Storm vs. SQLAlchemy

[Gustavo Niemeyer]

Hey Karl,

> SQLAlchemy and Storm seem to be really similar at the first look. Did  
> anyone take the time to compare the two? The only comparison I could  
> find was the discussion on reddit:
(...)

Nope.. there are no good comparisons so far.  Considering that they
take such a different approach, I'm interested in seeing people
talking about their experiences on real world projects, rather than
just a superficial bullet-point-like approach.

There are main obvious conceptual comparison points that are
noticeable, even then:

- Storm doesn't need the schema to be declared on the Python side.
  That's a side-effect of the fact that in Canonical we like to
  do schema management with SQL.

- Because of that and the way that Storm is put together, building
  backends is much simpler.

- It also means that Storm doesn't handle schema management.

- Storm's API is a lot more lightweight.  Most operations are done
  with a few methods in the Store, and the ResultSet.

These are important bottom line aspects.  Comparisons which are
too fine-grained will probably not be so valuable, since features
might be ported from one ORM to the other with certain ease.

-- 
Gustavo Niemeyer
http://niemeyer.net