DirectX...for X

[Fixxer_Linux]

Another way could also making exclusive Linux titles, as it is made for consoles.
It's known that just one exclusive title can make sell a lot of consoles. Those titles are, for example of the Playstation, Gran Turismo or Resistance. For Xbox, it's Halo.

As it has been done as a wide project in movie making with elephant dream, couldn't some of the community concentrate on one single project that could be as good as a commercial one. But exclusively available for unix based systems ?
For example, we already have good FPS with "warsaw" or "nexuiz", or as simulators "danger from the deep" or "torcs".
As those games are well made, they are still far from being like the commercial ones you can see from Gran Turismo or UT3. Mainly not because of lack of programming power, but, imo, because of lack of artists (textures, level design, music, character design, scenarios).

It has been possible for elephant dream to achieve a professional quality movie. Couldn't it be possible for a game ??

[NeoBrain]

Another way could also making exclusive Linux titles, as it is made for consoles.
It's known that just one exclusive title can make sell a lot of consoles. Those titles are, for example of the Playstation, Gran Turismo or Resistance. For Xbox, it's Halo.

As it has been done as a wide project in movie making with elephant dream, couldn't some of the community concentrate on one single project that could be as good as a commercial one. But exclusively available for unix based systems ?
For example, we already have good FPS with "warsaw" or "nexuiz", or as simulators "danger from the deep" or "torcs".
As those games are well made, they are still far from being like the commercial ones you can see from Gran Turismo or UT3. Mainly not because of lack of programming power, but, imo, because of lack of artists (textures, level design, music, character design, scenarios).

It has been possible for elephant dream to achieve a professional quality movie. Couldn't it be possible for a game ??

http://yofrankie.org/

I don't want to say that this game isn't good, but it's not near as close to comercial games as they planned (there is only one level for example).

[curaga]

Well, with all open-source games, there would be the issue that some people would do a windows build anyway, and then the hotlines all around the web would sound "linux exclusive fails, windows binary provided by third party".

In fact I would really like to see some linux exclusives happening.

Maybe, if the game is really good, and the devs do all in their power to avoid a windows port, an open-source exclusive game should be possible. For starters, they should use only linux-specific api's and toolkits. Ignore all portability, and just for certainty's sake, insert several of

#ifdef WIN32
exit(1);
#endif

into most parts of the code. Then add checks for Linux. Geez, this is starting to sound like a Linux Digital Rights Management, in a positive way

[Wyatt]

But do we really want to be "That Guy?"

No, just use glib and pull an ffmpeg with, "build from trunk." At that point, you've already created a barrier to the latest code such that most users will be unable to overcome. (What's mingw, an emulator? C...com...piling? What?)

[curaga]

But if it finds its way to a repo, the repo will have a source package available. I've seen some in both Ubuntu and Debian that have been fetched from trunk.

Then there's only the compiling

[Wyatt]

But if it finds its way to a repo, the repo will have a source package available. I've seen some in both Ubuntu and Debian that have been fetched from trunk.

Then there's only the compiling

Debian provides source packages for their repos? If it's a trunk build it would seem more sensible to just provide a revision stamp to check out and leave it at that.

Well, in any case, a Linux binary still doesn't help windows users much...

[portets43]

I'm getting into game development and my idea was to make cross platform open source demos of my games, but release a closed source binary exclusively for linux with better textures, features, more levels, and overall more complete.

[BlackStar]

I'm getting into game development and my idea was to make cross platform open source demos of my games, but release a closed source binary exclusively for linux with better textures, features, more levels, and overall more complete.

Great idea, this way you can piss off both crowds! Linux-specific closed binaries, that's a first.

My advice: develop a great game and release it. Create it on Linux. Get it included in distros. Optionally cross-compile for Windows / Mac OS X. This is going to do a lot more for Linux than trying to create some artificial tie-in.

No single game is going to suddenly bring gamers to Linux. However the *availability* of good games is a selling point.

An interesting tidbit: my current project (a brain-computer interface with high-end 3d graphics, quad-buffer stereo, etc, written in C#) runs at 72fps on Linux vs 61fps on Windows. This is the same system, with a Radeon 4850 on catalyst 9.3 and vsync off. Same binaries, different runtimes (Mono vs .Net). I haven't been able to pinpoint the source of the difference yet, but it is there and it is significant; and if it persists, guess which operating system we are going to use for the final product.

[Porter]

The original intent of this thread has some merit, I think. The expanded question, if I understand it correctly, is something like:

"Why not build a DirectX "implementation" for X, that would allow commercial software developers to port their Windows applications to Linux and still make "native" DirectX API calls."

I get it... that would simplify the whole process of porting software to the Linux platform simpler for developers that have standardized on the DirectX API for games, 3D apps, etc, because they could create Linux binaries that link to the "linux" DirectX-clone implementation.

As several folks mentioned before, the problem is that it's much easier said than done. It would require an enormous amount of documentation about the inner workings of DirectX, and would require explicit driver support from the graphics card manufacturers, both of which would be very difficult to achieve.



On a different topic... I agree that not enough work is being done on creating MacOS binary compatibility (or at least some type of compatibility layer similar to WINE) for Linux. That would be HUGE for the Linux platform, even if for one reason only: PHOTOSHOP. If it were possible to run the Mac version of Photoshop "natively" under Gnome or KDE, it would be an absolute game-changer for the PC OS marketplace, and the rate of Linux adoption would probably skyrocket. It would have to be easier to do this for Mac software than for Windows stuff, I don't understand why so much development has been done on Wine but so little on Mac binary compat.

[portets43]

Great idea, this way you can piss off both crowds! Linux-specific closed binaries, that's a first.

My advice: develop a great game and release it. Create it on Linux. Get it included in distros. Optionally cross-compile for Windows / Mac OS X. This is going to do a lot more for Linux than trying to create some artificial tie-in.

No single game is going to suddenly bring gamers to Linux. However the *availability* of good games is a selling point.

An interesting tidbit: my current project (a brain-computer interface with high-end 3d graphics, quad-buffer stereo, etc, written in C#) runs at 72fps on Linux vs 61fps on Windows. This is the same system, with a Radeon 4850 on catalyst 9.3 and vsync off. Same binaries, different runtimes (Mono vs .Net). I haven't been able to pinpoint the source of the difference yet, but it is there and it is significant; and if it persists, guess which operating system we are going to use for the final product.

well, the demo's will be open source. and when i create a sequel(or if enough time passes by) i will open the source to the first version. and besides, i'm trying to attract people from other os's. people under wincrap don't care about open source generally