Python and Zope

[Paul Boddie]

"Andrew Murray" <mr.murray at mindspring.com> wrote in message news:<9nettc$ce1$1 at slb7.atl.mindspring.net>...
> Frankly I was pretty unimpressed with Zope. After one gets over the
> 'gee-whiz' factor, I've found that the learning curve behind Zope and the
> relative non-portability of the code you must write for it far outweigh the
> benefits (what are they again btw?) of using Zope.

I found that there were a lot of nice things about some of the
technologies - back in the days before Zope, the whole idea of
dispatching requests through an object "containment hierarchy", whilst
templating the whole lot in DTML, was rather exciting. When Zope came
out, it just seemed to raise the level of effort to do the same things
as one would with Bobo, and I had to study the vague documentation
about writing products and experiment with the framework for a long
time before I could get back to doing things as complicated as I had
been doing with Bobo.

Of course, things may be a lot different now, but Zope slipped into
irrelevance for me after lots of features were added without any
decent explanation of what they did. The mailing lists seemed to be
continuously increasing in "noise" too; a new feature would be added,
358 messages expressing wonder at the new feature's coolness would
then follow, and yet one would still be wondering where such a feature
would be useful. Secretly, I suppose the authors of those 358 messages
also had their doubts too.

For me, Webware has taken over the role of innovator in my area of
interest. However, as Mr Murray notes, it is especially important to
work out what it is you want to develop and what it is that Zope can
help you with. The latter investigation might take longer than you
expect - for a while, Zope was advocated as the "one size fits all"
Python Web application server, but the experiences of many seem to
suggest that Zope's "one size" doesn't fit all, and this observation
seems to be supported by the number of alternative frameworks for Web
development in Python:

  http://www.paul.boddie.net/Python/web_modules.html

(Although it does seem rather too fashionable at the moment to invent
one's own templating framework, different only from its nearest
competitor by the slightest of nuances.)

> If you've used mod_perl, mod_php or mod_python before, I'd suggest going
> that route.

It's probably good advice to stick with what you know, if that can let
you do what you want to do. Only if Zope and competitors can let you
do that thing much more easily is it worth it to learn those
technologies. I would also suggest investigating the deployment
options for your application and the different frameworks; for a
while, people really seemed to be having problems with Zope and
Apache, even though the main focus of Zope development seemed to
concentrate on more exciting areas (ZKitchenSink?), but if you can't
deploy your application with other, desirable technologies then the
whole affair seems somewhat pointless.

Paul