Java EE 6 And Rethinking - JUG HH Afterglow

I really enjoyed the session at Lehmann's (cool book store with impressive amount of not only Java books).  It was organized by JUG in Hamburg. We had interesting, late night, discussions - and finally some consensus:

  1. DAO is not completely dead - only in the majority of all cases :-).
  2. EJB 3.1 are lighter than expected
  3. Wicket is cool and JSF just works - especially with EJB 3.1.
  4. OSGI inside the applicationserver could solve some interesting problems in HA and highly dynamic environments.
  5. Layering is not always good.
  6. WAR deployment of no-interface beans is interesting for small web apps with databinding...
  7. Async methods are interesting - especially in case they can be monitored and managed by the appserver.
  8. Singletons are interesting for pre-initialization etc.
  9. ...and other stuff - it was finally to late.
This time I used GF v3 prelude b24 with Netbeans 6.5 development build (200809130201) to show some new stuff. It worked surprisingly well...Thanks to the organizers!


Regarding point 7:
You mentioned that Async methods can be taken into account as an alternative to JMS in some circumstances. I fully agree with that and want to give just a little remark:

I can imagine that it would be more "light weight", i.e. will use less resources as JMS and so can be faster and/or more compact. That is because of the implementation on the basis of the Future mechanism and such stuff is running only in memory whereas JMS is using Databases (at least I do not know of another way).

Okay, the Async feature will certainly not support recoverable transactions or persistent states which will be recovered after a server crash-and-restart, hence will not guarantee the processing of the method, but many asynchronous tasks don't need such "recoverablility".

Posted by Ronald on September 18, 2008 at 11:37 AM CEST #

Post a Comment:
  • HTML Syntax: NOT allowed
...the last 150 posts
...the last 10 comments