[jdom-interest] Java 5 planning

Jason Hunter jhunter at servlets.com
Sat Mar 8 00:43:56 PST 2008

OK, so let me summarize what I've heard over the last 4 days about my 
question on the advantages people see in a jump to Java 5 dependency.

I asked for "big ticket" advantages that would provide benefits 
sufficient to justify the cost in making users have to recompile and 
possibly rewrite some of their code:

* Myself, I'd like methods to have more specific return types.

* Rolf primarily wants to remove compiler warnings.  Secondarily, he'd 
like more specific return types.

(This list doesn't seem very impressive, does it?  Are there more?)

We got some other "while we're at it" advantages:

* Victor would like to clean up the scaffolding classes that wouldn't be 
needed with Java 5.

And lastly we had a "cool stuff but probably wouldn't require Java 5" 
feature request:

* Michael Kay would like to see some XPath 2.0 support.

(I expect we could add this in a backwards compatible way with new 
classes/interfaces like an XPath2.java if we found the right underlying 
support libraries.)

Are there other advantages people want added to the list?

Do people think the "big ticket" advantages are worth making people, 
probably 100,000 people, have to alter their code to get them?

Do we risk new users going with other XML object models if the other 
projects are Java 5 clean on compiles and we're not?


More information about the jdom-interest mailing list