[jdom-interest] Performance tests

[philip.nelson at omniresources.com]

> It's really rough trying to get consistent results for timing 
> with a Hot Spot
> JVM, and some of the timing runs are dramatically better than 
> others, but the
> above figures should be loosely representative for the most 
> recent released
> versions of the products (Beta 6, in JDOM's case).

This is good to finally do.  Please do post your code.  There was a very
significant performance enhancement that Jason did after beta 6, though I
don't remember the specifics of when it came into play.

I haven't compared head to head, but anecdotally, these results don't seem
correct. In building, it certainly is possible.  When you compared building
speed, which parser did you use? Did you compare a cold run to a repeated
run (a single run with a new jvm instance to mulitple runs in the same
instance)?  I do use serialization every day now, so you should tell us what
problems you are having there.