[Chicago] Web Frameworks?

Cosmin Stejerean cstejerean at gmail.com
Fri Aug 17 20:56:42 CEST 2007

On 8/17/07, Massimo Di Pierro <mdipierro at cti.depaul.edu> wrote:
> Thank you for the suggestions to look into elixir.
> I think elixir is great and I will use it as soon as this issue with
> Tesla is fixed
> http://code.google.com/p/tesla-pylons-elixir/issues/detail?id=25
> I did not know that mod_wsgi was now stable. I will definitively
> check it out.
> Please do not misunderstand my previous comments:
> I am not talking about creating another python web framework and I
> think you people have created impressive python libraries. All I am
> saying is that today I find very hard to make the case to my students
> why they should use one python framework over another, and I find it
> even harder to make the case to businesses.

I don't think you should try to convince your students why they should use a
Python framework over any other Python framework. Most often this decision
comes down to a matter of taste. You should finds the best framework for
your needs and encourage students to explore the other options and pick the
one that feels most comfortable to them. Personally I'm often frustrated by
"frameworks" and often end up building something out of components that
suits my needs. Most of the existing frameworks are great when trying to
build an application quickly but I believe that any serious large scale
commercialy application will deviate significantly from the standard way of
doing things.

I think it will be harder to convince companies to adopt Python web
frameworks simply because AFAIK there are much fewer Python developers and
they are generally more expensive than the army of J2EE developers.

In order for the
> community to be able to make that case you must come together and
> concede something: choose one ORM, one template language, one
> framework. I only gave you my point of view about components to stir
> discussion, and my point of view is biased by my teaching needs.
> I could write my own "perfect" web framework but that is not my
> point. Too much choice is killing python web frameworks because the
> work of many good people is scattered and not focused.

I don't think there is one right way of doing web application development. I
like that frameworks like Pylons (i'm most familiar with Pylons, as far as I
know TurboGears allows the same) allow one to use SQLAlchemy or SQLObject as
ORMs and Kid, Genshi, Myghty, Mako (and probably more) for templating. If I
recall correctly someone was complaning earlier that the Django ORM is too
hard to use outside of Django otherwise I bet it could be integrated into
Pylons as well.

I haven't done much J2EE web development but as far as I can tell there are
many choices in that area as well with things like JSPs, JSF, Spring,
Struts, JDBC, Hibernate, EJB

I could be wrong but I don't think that the reason Java is popular for web
development is due to having just one way of doing things. I think it was
better commercial support, options for integrating with legacy systems and
availability of developers that helped Java become the de-facto enterprise
standard for web application development.

I think you should budget some time during the next meeting to talk
> to users, discuss the possibility to merge the existing frameworks,
> simplify them, agree on a small set of stable APIs, give some
> guarantees to the users about the stability of the APIs and document
> them or businesses will not invest in it.

I think this kind of effort will lead to Sun style over-engineering trying
to make something that will satisfy everyone and likely not ever be done. I
don't know about other folks but I care more about finding the right tool
for the job than building frameworks for the sake of building frameworks. I
think most of the current frameworks emerged out of the effort of a couple
of individuals to build something that would work for them.

So far only Django has been able to gather enough attention but it is
> still too difficult to use when compared with Rails.
I think there are enough people in the area to come out with an
> excellent product (but not 4 or 5) and an excellent marketing machine
> for it.

What I think is missing is a strong business arguement for using something
like Python for web application development. I've seen a couple of rather
unconvincing articles on the topic but no concrete evidence. Can a team of
experienced Python developers can write something better, faster or cheaper
than a team of experienced Java developers? (By better I mean one or more
of: less bugs, easier to maintain, or faster to add new features).

> _______________________________________________
> Chicago mailing list
> Chicago at python.org
> http://mail.python.org/mailman/listinfo/chicago
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/chicago/attachments/20070817/23a343af/attachment.html 

More information about the Chicago mailing list