OK. A few more comments. First, these are the strengths I see in Cocoon:
1. MVC2X. Although the term didn't originate here it certainly was
meant to apply to Cocoon. The separation of concerns is IMO
Cocoon's greatest strength and is the greatest weakness of many
other technologies. Unfortunately, we provide several features
that let you violate this.
2. XML - There is a lot of debate about the overuse of XML, but I
have found that it is a great way to take data obtained from
various sources for various purposes and to make it available to
web authors.
3. Flexibility - Source resolvers are a godsend. We are able to
things with Cocoon that would be almost impossible with other tools.
4. Capability - It comes with practically everything but the kitchen
sink.
And the weaknesses:
1. The user base in the U.S. is minisule. Selling management on
using Cocoon is practically impossible. Managers all seem to
follow the lemming syndrome - they force technology no matter how
bad it is because everyone else is using it. Unfortunately, there
is a logical rational for this - it is cheaper to hire engineers
and web developers for standard tools than it is to find ones who
know an offbeat technology. This leads right into
2. There are few, if any, third party tools available. Although you
can get XMLSpy or Oxygen are they really going to be a web
developer's best friend?
3. The learning curve. Way too much stuff to know. Because it
includes so much stuff it is often difficult to know the right way
to do something.
Now while I agree with a lot of what you are saying I'm not really sure
it completely addresses all the issues. For example, while allowing
folks to use any container they want might be a good thing it can also
make it more difficult if that just becomes one more thing they have to
learn about. Frankly, I'd just say to heck with it and use Spring. At
this point I don't think we have to worry about radical changes or the
community dying. Of course, it then makes more sense to leverage Spring
WebFlow as I have in mind. And since Spring is such a popular framework
it makes it much more palatable for manager types - especially those who
think EJBs suck and Spring/Hibernate is the answer.
By the way - I'm not interested in Cocoon being sexy. I'm interested in
seeing its adoption rate increase, which is a very different thing.
As to your sitemap changes, I think what you have in mind is right on.
It wouldn't bother me if in 3.0 we have only javaflow and flowscript for
those who want to "roll their own" and webflow for those who prefer a
stateful flow.
As for your comments on Ajax, I don't know if you read it but you should
take a look at this entry from Carsten's blog
http://www.osoco.org/archives/2005/10/index.html.
Ralph
