how does performance of jpct compare to the performance of xith and jme?

[Toby]

and also is it easier to use?
thanks!

[Kearnan]

This is just my opinion.  Xith and JME are game engines while JPCT is a set of tools which make OpenGL (LWJGL in this case) easier to use for someone who would like to create their own game engine.  I cannot see where performance would really be debatable in this case.  We're not comparing apples to apples.  On one hand it's hoping someone else's methods are what you need and on the other it's creating your own methods without the dull work.

OpenGL (LWJGL) is easy to work with.  Creating your own model/texture load methods and collision detection is pretty much a pain I've dealt with and would rather not.  JPCT makes both of those easy without interfering with what I want to do with LWJGL.

Easier to use?  Yes, it's far easier to use. But no bells n whistles.

[EgonOlsen]

We already had a similar discussion here: http://www.jpct.net/forum2/index.php/topic,809.0.html (and if the search-function of this foprum wouldn't suck, one would be able to find it more easy... )
..without coming to a final conclusion. I would say that you can always create a test case where jPCT loses and XXX wins and you can create the opposite situation too. It depends on what you do and how.