Consoles really aren’t going to happen… until they do. But before then there really is no point considering them in your game designs or business plans. Concentrate on using what technology there is to get games out there on actual computers that currently exist.
Right now we’re seriously missing a DirectX binding, and that’s the missing technology link that’ll open up the doors to everything else… XBox, portals, PCs of all flavours, abstracted rendering layers, etc.
Cas 
[quote]Right now we’re seriously missing a DirectX binding, and that’s the missing technology link that’ll open up the doors to everything else… XBox, portals, PCs of all flavours, abstracted rendering layers, etc.
[/quote]
In the 2D world, Java 2D is really promising if you consider Java 6 and the fact that this release will be omnipresent in the future (within 5 years maybe?)
In the 3D world, we still have Java 3D. Why not seriously considering supporting its dev and maintenance to make it up to date with the latest DX features (because that’s an actual problem)? Its architecture might cause performance issues compared to Xith3D for example but why not focusing on those weaknesses?
DirectX programmers that do 2D really don’t want to learn Java2D, it’s a different beast, and you have a lot less freedom.
DirectX programmers that do 3D really don’t want to learn Java3D, because they either have their own engine, or don’t want to have to wait before bugs gets fixed (the fact that’s opensource doesn’t mean you can just fix something the way you like best), or simply because it’s way too highlevel.
If we want to attract DirectX programmers, we need a DirectX binding, not another API they don’t (want to) know about and makes them feel restricted.
How hard would it be for someone to do this binding? I browsed the DirectX documentation and it is very well written with lots of examples and tutorials. Microsoft Visual Studio is even free now and all the libs are freely accessible from the internet. For someone that has already made something similar it should be relatively easy i suppose.
On Java3D i like this api a lot since it was the frst high-level graphics api i used and it could be improved a lot by making it more flexible. That is we can choose whatever graphics device, input device, sound device or whatever we want to use. The problem with improving Java3D directly is that this effort would only benefit Java3D and could not be used by other engines like Xith or jME.
The binding itself is really not that hard to do. LWJGL could be relatively easily modified to cope and add it as an “optional module” like we do for DevIL and FMOD. Somebody just needs the time and inclination to do it. Maybe we should post an advert?
Cas
Wouldn’t a lot of the current display code require changing to cope with creating either an OpenGL display or a D3D display?
I thought D3D made heavy use of reference counted objects too, which doesn’t sound nearly as easy as binding to some static C functions.
I hope someone has the time and inclination because this is something that would help the entire community. All those guys that are working on Java game engines consider the benefits of a DirectX binding that could be used by everyone. Unfortunately the only thing i can do is make threads like this here and in other java related sites and hope someone who can do this binding will see my point of view. Posting for help on a non-java related site would be useless i think.
The number of volunteers is surprising, it has to be said… (that is, zero). I sort of expected there’d be some more enthusiastic types here who would jump at it but no-one’s biting.
Cas
I’m kind of too busy with my day job and another project on the side, so at the moment I don’t have the time to both learn D3D and to get the binding started.
When I’ll have some more free time on my hands (hopefully somewhere this summer or fall), I might give it a go.
The way I see is that for any game maker out there, there already exists a 3D binding that “does the job” (LWJGL or JOGL), so there is no incentive for anyone to make a D3D/DX binding.
In anycase, the next OGL release would be Object based rather than ID based like we have now. So both wrappers are going to have to change to accompany this. This is a good thing for a DX binding because it bridges DX and OGL ever closer in terms of API and then someone perhaps might be inclined to make a DX binding
DP
My opinion is nobody on this Java forum is interested, because everybody that is interested in DirectX, doesn’t use java (and therefore doesn’t use this forum).
To get some DirectX coders here, we need a DirectX binding first…
My C skills are lacking bigtime, so I can’t volunteer for it either.
[quote]My opinion is nobody on this Java forum is interested, because everybody that is interested in DirectX, doesn’t use java (and therefore doesn’t use this forum).
[/quote]
Well…
[quote]I for one will feel much safer knowing there is a DX binding available for Java should I need to unlock that 35% of Windows machines (actual statistic I have measured myself over 3 years) that don’t have OpenGL.
[/quote]
I think we’d have at least one forum member interested, and with good reasons.
35% of windows users is a lot of potential left on the table without a DX binding, a large figure which might even rise in the near future…
I’d say anyone remotely interested in using java for commercial games will be interested.
I’m also interested, but currently too busy to work on that subject 
Java needs a DX binding to attract developpers from outside our community, once they’re in, there are many chance everybody will benefit from their contributions.
Lilian