Stefano Mazzocchi wrote:

You describe the revolution from the client side. The other part of the (r)evolution takes place on the server side. There is a change in how enterprise applications are designed. Just let me throw in the latest buzzword: ESA. One of my customer's projects goes into this direction and we will live quite happy without Cocoon, even without *any* server side presentation framework at all (last week we finally threw out Struts.). Castor-XML and a couple of (very) simple servlets does everything we need to provide data for a Flash client.

Your question was: Is Cocoon obsolete? Personally, I don't think so. I think that we will just see a shift in what Cocoon is needed for over the next years. The question only is: Where is Cocoon shifting to?

a) Cocoon could become *the* client-neutral framework that can support all this different rich-clients (XUL, XAML, xforms, etc.) *transparently*. Of course this needs more work on our side to make this happen.

b) Not XUL, XAML or Flash are the least common denominators but Ajax. Ajax is the only cross-browser alternative. One of my customers wanted to implement a client based on Flex. This was 10 months ago and Ajax was a football club in Amsterdam. Over the last weeks his opinion has begun to change and I'm curios which technology we will finally choose.

Personally I think that alternative b) will become mainstream for many reasons, the cross-browser argument will only be the strongest. The others are:

- If alternative a) becomes mainstream, what will happen with the thousands of Struts, JSF, Cocoon, RoR developers? Usually they don't like to write rich-client apps. Of course ESA will take its toll and more people will have to write webservices. - This would also mean that managers would have to admit that they have bet on the wrong horse and large investments would become obsolete. They will make sure that this doesn't become obvious. - The variaty of rich-client alternatives will grow much more because of the success of mobile clients (PDAs, mobile phones, etc.) and finally the least common denomintor will be some kind of XHTML and Ajax. - The world will not agree on one rich client technology. M$ will do it's best that this will never happen ;-)

-- Reinhard Pötz Independent Consultant, Trainer & (IT)-Coach

{Software Engineering, Open Source, Web Applications, Apache Cocoon}

web(log): http://www.poetz.cc --------------------------------------------------------------------