Choosing an API

I got a group of about 10 people together and we are going to start developing a game. We have been discussing API’s over the past week. JOGL, LWJGL, Xith, jME…



We hate to think that we would have to rewrite what jME and Xith might already do for us but we have come to some degree of a standstill due to extending either API. I’ve talked with the Xith guys and it seems the only way to manipulate opengl directly is by extending their API and often times actually modifying major Xith classes. I don’t like that approach.



The only reason we opted for Xith first was because we wanted to go the JOGL route and jME doesn’t support JOGL, yet, does it?



With that being said, we might be able to live with LWJGL until you get JOGL support in if I knew how hard/easy it is to add features into our game through jME.



A very bad example. Say we want to create our own Fog. I know that I would want to create Fog as a node to be added to the scene, but what classes would I need to extend/implement and are there any core jME classes that would need to be modified for rendering our new node?

I believe that jme allows you to directly access LWJGL without too much hassle, though I think the goal is to make that not necessary, unfortunately I’m not very experienced with jme yet so I don’t know the exact answers to all questions. I have had a lot of experience with JOGL, and I have to say that despite my original opinions, probably influenced heavily by the fact that jogl is a sun release, lwjgl is far better at least for gaming stuff, probably for everything except possibly when mixing awt components and mac os. There are some threads about JOGL and it was supported at one time, and there is at least one thread that implies supporting JOGL is planned again in the future now that JOGL has had an update or two again. I think the weakest area of jme right now is model loading/animation stuff, but I think there are a few nearly completed loaders with better animation around the corner (not the loader I built… that is probably stagnant again for at least a few more months).