I think the key is to support selected opens source game projects.
This games should be the leaders in their sector of the PC gameing industry.
Support the best 3D engine.
The best RTS engine.
Assure that it are big projects with enough developers.
And then resease this games with a haiku distribution.
Vestifex wrote:
Yeah, a port of the OGRE engine (http://www.ogre3d.org/) would be really nice and probably not very hard to do. It's licenced under LGPL. I think it's mostly a matter of compiling the required libs, among which I think zziplib and DevIL are the only ones I haven't seen on BeOS. For a start the SDL port already included in the engine can be used, and later on a Be API port could be done if needed. Quite a lot of projects are using this graphics engine, both commercial and open source. Have a look at the screenshots on the website..
Support the best 3D engine.
umccullough wrote:
Serpentor wrote:umccullough wrote:I didn't see the sarchasm tag...Not exactly sure what he is trying to say but I think it is something to the effect of "let’s make Haiku only games." (ie games you can only play on Haiku and not worry about letting big game companies easily port their games to Haiku). And yes, I think he was trying to be serious about it! I don’t know if I like the idea of not letting big game companies easily port their money makers to Haiku but I do like the idea of building a couple of excellent games for Haiku that you can NOT play on winblows or OS X!
Locking people from making cross-platform games for Haiku would very much put the final nail in the "Haiku == Games" coffin.
That would be a similar attitude as Microsoft’s DirectX… which doesn’t benefit the end-user, only Microsoft…
I tried keeping weight on the first sentence: lets make game development easier and cheaper.
Without considerable change in ability to support games, this really would be a wrong choice and wouldn’t benefit anyone. The last sentence would come as side-feature, or not at all. I’m thinking about not caring whether ppl can port games from/to haiku.
So we shouldn’t just port things. We should create much more advanced system than those which exists. One where one person could actually do vital game straight out of idea. This’d require considerable development on things like producing graphics, audio, the improvement of programming interfaces would be only small portion of making game development cheaper.
Open source projects are quite effective with regard to makeing development cheap.
What Haiku could do is create a link between open source projects and commercial interests.
The licence haiku uses is much more commercial friendly than that of linux or other open source OS.
So haiku could try to catch the image of the OS for both open source and commercial interests.
And in terms of marketing image is allmost everything.
Okay. Support for games, I think, is the epitomy of support. When you support hardware enough to run a game (one of the most intense operations any average user could undertake, unless you’re the average unix-user who’s into 3D modelling and has a renderfarm…), I’m sure you’ve covered many of the other bases. Shouldn’t it be within the capacity of this project to make Haiku capable of taking on whatever sort of applications are out there? Games aren’t the only multimedia application out there, but they are the ones that get the community excited.
I say that the project makes itself as well-rounded as possible: it should be forgiving for the *nix convert; it should be useful for the college student; it should be perfect for graphics work and video editing; it should be lightweight for the average user who complains about microsoft word slowing down too much; it should be easy to develop for, showered in great documentation; it should be filled with games – but not just chintzy games like breakout/tetris clones written in SDL, but real games that the world has never seen before that run better on slower hardware.
It shouldn’t be the case that Haiku == Games. It should be something that it can do without blinking an eye. 
None of this can be done with code alone, even though code is 70% of the battle (or whatever majority-resembling statistic you want to stick in there). So many lovely projects get neglected. Seems like Unreal is the only series with releases in linux. People must be convinced. A campaign must be launched to make sure that Haiku is known to the people inside who need to do the other part of the coding. (;
Yeah. These are all big words from someone who hasn’t contributed much so far, but perhaps I can contribute… big words. snickers
By the way, is there any way to make something binary compatable with Direct X – and can you write your own implementation of the Direct X standard without being bound to some EULA from Microsoft?
apoth wrote:
...It shouldn't be the case that Haiku == Games. It should be something that it can do without blinking an eye. :/
I think generally, this is how it works anyway. Once you provide the stuff that games need (SDL and the like, OpenGL, Sound output, 2d/3d accelerated video, alternate input devices, network support, and ample processor time) the rest of the OS pretty much just gets out of the way. The largest problem is generally accelerated 3d right?
SDL and OpenGL of course are not the only APIs that game developers use - so providing the ability to wite/port other game engines is also critical.
I believe it is important to not force game developers into a specific API. While this may seem like a good idea - I think it stifles innovation in game development, and prevents game developers from finding that new edge against their competitors.
It is also NOT the job of the OS developers to write the game engines (unless they’re building a console OS) - only to provide proper access to the hardware.
Quote:
By the way, is there any way to make something binary compatable with Direct X -- and can you write your own implementation of the Direct X standard without being bound to some EULA from Microsoft?
This is basically what Wine does… and a good port of wine (not one that relies on X11) would be a great addition to BeOS/Haiku for compatibility with Windows software and GAMES… but I think Wine has left a bad taste in some people’s mouth due to it’s Linux and X11 ties - preventing many from embracing it as a possibility.
I think Wine has plenty of potential, if you are able to toss out the "bad" parts and use the good parts. ReactOS for example, relies on much of the Wine code for their compatibility with Windows.
umccullough wrote:
I believe it is important to not force game developers into a specific API. While this may seem like a good idea - I think it stifles innovation in game development, and prevents game developers from finding that new edge against their competitors.It is also NOT the job of the OS developers to write the game engines (unless they’re building a console OS) - only to provide proper access to the hardware.
I completely agree. Perhaps where others succeed is in proving to the bigwigs in the market that we are worth investing some time (and ultimately, money) into. I feel like there’s this elitism in the Linux community, as if there’s some sort of rule against corporate cooperation. I think it’s essential to cooperate with certain people to enact change while preserving the nature of your own project.
You’re right, the developer must be given free reign in any given environment. All APIs should be supported.