Direct X conversion LIBs a possibliy?
 

Hey! I was having an idea to make a lib that would replace the real direct x libs and translate it to SDL, and OpenGL so we could offer this to companies and there wouldn't be any reason not to make a linux version because it wouldn't take any effort at all. Does this sounds crazy? lol

DarkNebula, there was some attemts... But it is better to implement small part of DX for the game to port. DX is really hard and new versions will be available faster then people implement something for linux...

We pretty much have this already. It's called Cedega or wine or any of that emulation stuff. Just remember, emulation is what will kill Linux as it will no longer be Linux it will just be a shell that windows rides on top of.

I don't mean emulation. I mean from the implemation of the source code itself. Is anyone a programmer and knows what I'm talking about? Replace the Direct X libs they link with, with a new kind that will "translate" the code into something that will work for Linux. And it will output a linux binary, not a Windows exe. And it would be a REAL linux program using OpenGL, a XLib. Well I think it would take a while, but a group of people might be able to keep up.

That's exactly one of the things that Transgaming is doing. They have a tool like that that allows programers to code in one API set that can easily be compiled for as many platforms as possible (PC Linux, PC Windoze, PS2, EcchsBox, CameCube, etc.).

Keep in mind that wine stands for; Wine Is Not Emuation. What it does is intercept DirectEcchs calls and translates them for Linux. Having a compiler do the same is just an extra step... A unified API is a good goal. A combined SDL and OpenGL suite may be the answer,... but most coders are indocrinated in DirectEcchs for M$ C++.

The problem is that there isn't a real demand for it. The developers have people with limited skill sets,... C++ and DirectEcchs. They don't bother hiring the smarter coders who are interested in doing greater things. They have vendor lock-in in their programming packages, and don't want to switch. Switching costs money, or doesn't mean a return on money they already spent,...

They don't like the idea of "wasting time" learning new things that have limited application to getting as much money as quickly as possible. Developers want to get a contract, get funding and push a game out the door as soon as possible. Learning new things is not conducive to that. That's why we get demand for very certain skill sets in programmers. That's why we get re-hashes of old titles made with pretty graphics. As the graphics engine improves, the programmers are simply supplying better meshes to plug into it and simple AI routines to keep things interesting. Good game play is a thing of the past...

Well, if you ask me... those are pretty crappy programmers! I'm a programmer, and I learn all that I can, Direct X (eww) OpenGL, SDL, FMOD, OpenAL, and more libs; so I can be the best programmer I can be and use the right lib for the right situation. I guess these companies should invest more time and work into their projects. No wonder we have so many... horrible games out these days. I haven't even had a formal education yet... I guess that kind of thing ruins it. lol Good post by the way. Makes sense to me. Any other comments?

Its not necessarily up to the programmers. Remember, its the suits that hold their leashes,... and the suits don't want them doing what they consider useless pursuits.

I actually ran into hostility from the developers of one game (X2 - The Threat) for suggesting a Linux port or cooperation with the people at Transgaming to get things working in Linux for their game. Their response was something on the order of "if our game is not multiplayer, you have to PROOVE to us that it is financially worth while to do that. Otherwise we will simply do what is profitable,... ONLY." I had people actually asking me for studies proving that mod-able games sell better and that multi-platform games get more return on investment than single platform games... all this as just a potential CUSTOMER asking for a little cooperation in getting their game to run in Linux, somehow or some way. Some people are VERY entrenched and it's tough to budge them. I really don't know if someone at M$ got to them with some financial incentives (free development tools etc.), but I'm not one who is usually inclined towards conspiracy theories.

Now, If I was with a major development house, and was waving money at them,... I'd bet their attitude would change real quick. It's sad, but true. And even one or two Linux fans within their development staff is not enough to win them over.

BioWare, for example, has hinted that it is very dependant on the publisher with regards to what platform they will release for. That's why Knights of the Old Republic, which was released by LucasArts, with an exclusive console release on Xbox, did not get ported to Linux,... M$ had their say with LucasArts. That was in spite of the fact that the Neverwinter Nights engine on which KotOR was based had already been ported (so the HARD work was already done)...

In terms of Linux versions, the order of multiplatform friendliness goes something like this:
ID
Epic
small, independant houses
Maxis
Atari (which partners up a lot)

You mean like a tool XYZ that when ran in windows with (ie:)

"XYZ gamesrc --target=linux\gamesrc" --what-so-ever-options-for-common-need-of-adding-them

would produce complete sources for the game for linux in linux - directory?
I have no idea on DX coding, so I just use an example of the work in C/C++.

If the source on windows with DirectX is presented as C and Linux as C++,
then it would probably replace (remember, this is *not* DX...)

"printf(what-so-ever);"
with
"cout >> what-so-ever";"

Did I get the picture?
That is actually a pretty good idea. And after a few thoughts, it wouldn't even be so difficult to do. Assuming that DX and OpenGL/SDL both have functions that can be used. I mean, if DirectX has a function like "run_like_hell_cause_your_house_is_on_fire and SDL/OpenGL has only function call_911_if_your_house_burns, then it will be a problem. If you run like hell you will save yourself, but without 911 you *surely* don't have a house anymore..
************************************************************
*hmm. ...Did I just spend 15 minutes of my work-time producing a long post that even I don't *understand anymore... Gee.
************************************************************

But, as said, the idea of a "source-converter" between win and Lin is good...

Yes, you have it almost exactly. I thought it would be a good idea, and that companies would pick it up because it woudn't take any time. (If there wern't any bugs). Thanks for getting it lol.

Hmm, not bad, not bad. As for the bugs are conserned, I assume that they'd be fixed in notime due the "opensource" - nature of our program. :P
If we could only find those DirectX - guru's needed. I'd take part on the job without a doubt too. (With my incomplete knowledge of C++...) :P

(By the way, since I'm a Finnish guy and my english is *quite* bad: Does "participate" stand for "take part of something"?)

Yes it does. lol And it would be cool to start such a project. I'm pretty good at C++, four years worth. I actually learned Direct X, and then said.. "There must be a better way!" So I picked up a OpenGL book and never put it down :D I wasn't thinking about starting this as an actual project, but it many are interested; maybe it would actually be a good idea!

Well, the idea is good indeed.
Well I am interested on that. Not only for my personal "learning by doing" - attitude, but it would probably be what we need for Linux to be able of hitting the Jackpot as a "Gamer's choice". For both, the game developers (since there are ppl NOT useing win at all) and the Linux community (Games, Yippee!!!). If you get a sudden inspiration on starting the project, and get a few more DX/OGL - coders, count me in too... If (or when..) my C++ skills aren't enough to be a part on the "hard" coding, then I could be one of the beta-testers to check how it works on different distros.

Seems good. I think I will wait until summer if I was to start such a project. I would first post on GameDev to get some more feedback from the community and see if they have any suggestions or if they want to join the project. Do you have a IM program?

IM program? does that stand for Instant Messenger? If so, then yeah, I do. With adress of thautane@mbnet.fi
I might also try to do some "digging" aroung for some ideas/etc on the project. ... .. . And maby start learning OpenGL... ;)

The please contact me on AIM as DarkNebula45, or on MSN at darknebula@comcast.net