Dive into Oracle ADF

Send me a mail
 Dive into Oracle ADF   Click to see the XML version of this web page.   (Updated: 2/3/2008; 9:17:12 PM.)
Tips and tricks from Steve Muench on Oracle ADF Framework and JDeveloper IDE

Search blog with Google:
 

Search BC4J JavaDoc:
 

May 2005
Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30 31        
Apr   Jun

Get Firefox!

Friday, May 20, 2005

From Dave Winer via Pandemia, you can go now to http://www.google.com/ig to start personalizing your own Google home page. It's like a Google portal. Interesting.
4:46:44 PM    



Oracle's Ted Farrell is interviewed by FTP Online where he talks about the importance of frameworks.


4:30:55 PM    


To keep things simple, all of the steps in the Building J2EE Applications with Oracle JHeadstart for ADF tutorial use the embedded J2EE server that JDeveloper 10.1.2 includes, which is none other than Oracle Containers for J2EE (OC4J). This means you can iteratively run and debug the J2EE application without going through any formal deployment step, which is a great thing for development effectiveness.

However, some users who have gone through the tutorial have asked the follow-on question, "How would I deploy the JHeadstart-generated, ADF-based web application to an external application server?" Great question. I've written up the few extra steps that you can follow here:

Steps to Deploy JHeadstart Tutorial Application


3:11:42 PM    


In the JDeveloper 10.1.2 release and earlier, when building a Swing client interface that is deployed in a 3-tier deployment, the thin client UI works over EJB remoting and works with a dedicated server-side EJB session bean facade that JDeveloper builds for you automatically when you indicate that you want your ADF application module deployed as an EJB session bean. The operative word in the previous sentence is the word dedicated. That is, 50 simultaneous users doing work need 50 instances of their server-side session bean. This is fundamentally different for how ADF-based web applications work, where 50 simultaneous users doing work might be able to be serviced by a server-side pool of only 10 application modules, depending on how much "think time" the users perform between doing actual work against the server.

In JDeveloper 10.1.3, we're introducing a cool, new feature for building much more scalable Swing business application clients. Stated most simply, it allows the Swing client to leverage the same server-side application module pooling infrastructure that today's web-based ADF applications enjoy for scalability. The client can perform changes to its local data (across multiple levels of master/detail information), and then send a whole batch of changes to the server in one round-trip. As is the case in the web scenario, the only persistent state needed across requests is a simple client cookie value that the client uses to "rendezvous" with its pending transaction state on subsequent requests over the life of a complex, multi-step business application transaction. From the point of the EJB container, the only persistent state is that cookie value. The rest of the pending transaction state management is identical to the state management features described in this ADF whitepaper on the subject, meaning that when running in a clustered server configuration there is an absolute minimum of information that the EJB container needs to synchronize across nodes.

We are looking for existing ADF Swing customers (using our JClient bindings) who might be willing to help us do some testing on this new feature, so if you are an existing ADF user building Swing clients and might be interested, send me an email.


12:12:22 PM    


© Copyright 2008 Steve Muench.