another boring SOA-talk. from technology side, SOA is just another buzzword with old concept: communicating through interfaces.
to me SOA is rather a link/understanding-concept between IT and business. so you will only discover the so called SOA benefits if you emphasize business view (SOA-governance, business reorganziation, process-driven view, etc.), and that is a very very difficult thing to achieve, especially if the company has got many employees, which are mostly anyway annoyed with it-systems and thus aren't really willing to believe in SOA-jesus coming soon and walking over
water.
two objections (there could be much more) to mr monson watching j2ee decease:
what enterprise level webapplication does mr monson develop if there are no Servlets/Portlets or similar?
which management mr monson wants do convince, to chuck all J2EE-expenses out the window of the last 10 years and invest again for 'another' promising technology?
as a platform both jse and jee will definetely survive, accompanied with alternative apis (like great spring, apache commons,
testing frameworks etc....). ejb 2.x are crap, lets see what ejb3 is going to be (you still need an appserver).
anyway i think we should put more effort into making systems better maintainable, getting analysis better, improve customer-developer communication etc..
jumping on every new technology release train is just money waste, technologies cycles are just too short for that. and i think with current technologies and tools it is possible to build good apps.
[ July 12, 2006: Message edited by: manuel aldana ]