Slashdot Mirror

2.) DirectX is a MICROSOFT ONLY format. It will never, ever, be in any linux distro except in emulation form. And for second, why should it be? OpenGL is fine and great, and with 2.0 coming out you can stuff DirectX where the sun don't shine.

At its very core, DirectX is just a set of APIs. Yes, it's a Microsoft API, but the exposed interfaces are well documented, and ignoring any possible legal issues, it is entirely possible to write a DirectX implementation on another platform. Okay, some of you may disagree on whether or not DirectX is well documented, but it's documented well enough for emulation purposes.

There are wrappers available that translate Direct3D calls into OpenGL calls (similar to Glide wrappers from the 3dfx days), and I don't see any technical problems with removing the OpenGL layer and having the new Direct3D implementation call the graphics card directly. However, and correct me if I'm wrong, I think Linux 3D graphics drivers are currently all proprietary, so nVidia and ATI would have to provide the Direct3D layer.

Still, even with an emulation layer, why SHOULDN'T DirectX run on Linux? Ignore legal issues and Microsoft's desires. Believe it or not, there are some developers who've only used DirectX and not OpenGL+SDL. It's worth having DirectX on Linux even if only a tiny fraction of those developers decide to port to Linux. That fraction may grow, and after familiarizing themselves with Linux they may switch to other APIs that are better supported on Linux, such as OpenGL and SDL.

At least, if this will do the bulk of the DirectX -> OpenGL heavy lifting, the remainder can probably be worked around enough to get a functional game. Then the implementor can work on the coverted stuff more or less at leisure.

will it help people port to other platforms? doubtful, as they're probably using other ms stuff if they're using DirectX.

will it help people use more advanced features on other platforms? no, since they're just using underlying opengl (and extensions) anyway, which they could do in the first place.

is it more performant? no way - it's another layer of indirection, so it's at least an additional pointer dereference, and extra stuff on the stack.

so i'm left thinking this is a solution in search of a problem. if you want portability, you write to opengl. if you want extensions, you use the _portable_ extension mechanism that opengl already provides. check out nvidia's directx vs opengl extension comparison some time - guess which one has better & more support? hint, it doesn't start with direct..

so, again, why would any sane developer write to this?

(and yes, i read the faq.)

1. Which operating systems the open sourced version at Sourceforge supports as of now?

Stephane Denis: Actually, the sources are designed for Win32 but they will be compatible for Linux and MacOS soon.
The ENSEIRB (a french engineer school in Bordeaux) are currently porting it for Linux and will probably speed up a lot the developement of the interface.

2. If the wrapper only supports Windows and BeOS, how easy/difficult a port to Linux or MacOS would be?

Stephane Denis: Personally, I continue the win32 emulation/wrapper to validate the wrapper compatibility and then I will do the MacOS version. For Linux, there is already a crew on it working. For the BeOS version, well I probably adapt it later.

A test D3D-2-GL program loading under WindowsXP. Click for larger version.

3. How far down the line the wrapper is? What has to be done yet?

Stephane Denis: Actually the Direct3d 8.0 part (only immediate mode) seems to works more or less, but not sufficiently enough to support really complex programs for the moment. But I expect that more people will look on other parts too, like DirectInput and DirectSound (since they are completly separated modules).

4. How fast/slow the implementation is when compared to a "native" GL app?

Stephane Denis: The speed or efficiency of the wrapper depends mainly of the OpenGL extensions supported. Since most of the time is consumed on rendering and not on API calls, this should be as fast as the original Direct3D code.

But you know, the goal is to get the DirectX API available on non-Win32 platform in order developers be able to create or port actual Direct3D 8 code. For Win32, some video cards like 3DLabs Wildcat or professional SGI video cards do not support Direct3d or Directdraw at all, but they have an excellent Opengl implementation. So the wrapper would be especially useful for theses adapters. A solution already exists, but we want to add Direct3D8 support.

I truly hope that more developers will join this interesting project now so we get things going in a faster pace.

A Direct3D application running in GL mode with the help of the wrapper.

Also, what about a wrapper against D3D? There is currently a nice integration with OpenGL, when SDL needs to use 3D acceleration, but under Windows (where a lot of SDL games already exist) most of the coders are using D3D. Also, based on the fact that D3D 8 is not as bloated as previous versions, a wrapper for other OSes could be easily achieved (IMHO).

Last question, again on the spirit of 'SDL Vs DirectX', are there any plans for more support for more input devices, like force feedback wheels etc.

Take care and thank you for all the hard work you have put on SDL the last few years. (and also greatly thank you for the BeOS port! :)