Is OpenGL the right choice for SL?

Now, let's get another little thing out of the way for just a moment, and that's OpenGL versus DirectX. Four or five years ago, OpenGL was the bees knees and DirectX was the red-headed stepchild. No longer. Pretty much every game released today is DirectX. There are a few notable exceptions, but they are the exceptions. OpenGL largely has itself to blame for this, for while it took ages for OpenGL to reach the 2.0 milestone, Microsoft has been radically increasing the functionality of DirectX. Game developers, knowing their largest market is the Windows platform, have gone where the money is, and Microsoft has done a lot to make DirectX very, very attractive to them, both from a financial standpoint and a performance standpoint. And while the idea of cross-platform gaming is a neat idea, I don't think too many game developers and producers lose too much sleep wondering how they're going to market to Mac OS X or Linux gamers; there are simply too few of them to matter. It's not nice to say that here, but it is the truth.

As for professional applications, typically the stronghold of OpenGL, even that has been fading. 3D Studio Max, one of the most popular 3D applications for the Windows platform, now strongly recommends you use the DirectX drivers versus OpenGL. Performance is usually better, and you get to make use of all the DirectX features implemented in hardware on gaming-derived video cards like nVidia's Quadro line. For a while, DirectX stability in these apps was dicey and OpenGL was rock solid, but even that has now been reverse. 3DS Max 7.5's OpenGL is noticeably slower and buggier than its DirectX, again a reflection of where developers are spending their development and Q&A time. Maya, Lightwave, and others are in similar positions. OpenGL won't go away, but it is slowly being marginalized. That's been going on for a while.

There's a lot of angst over what this will do the OSS platform since DirectX is solidly proprietary. Unfortunately, the answer isn't one you're going to like. Nobody, not Autodesk, not Microstation, not Adobe, not anybody in the large-scale commercial software space really gives two damns about the OSS market right now. And why should they? It represents an amazingly tiny fraction of their global market, yet would very likely consume a disproportionate amount of the overall support, troubleshooting, and Q&A resources available. That's not a winning combination for any company. Furthermore, most OSS adherents don't like paying for their software, and they certainly don't like paying $4,000-$10,000 for major OpenGL-based design packages. Thus, not only is the market tiny, it's unreceptive as well. Again, if the OSS market is wondering why no one is paying them attention, you don't have to think on it very hard to see why.

The good news here is that the OSS community can do whatever it damn well pleases with OpenGL. Don't want to follow the DirectX bandwagon? Write your own OpenGL app. Who knows? If it's done well, it might actually gain a foothold. But I sure wouldn't bet on it. I wish I could, because I'd love to see more competition in this arena, but I'm too much of a pragmatist to really have much faith in this area. Companies are going to keep using -- and paying for -- what works for them, and what works for them is what has always worked for them. Microstation shops generally stay Microstation shops. Ditto for Autodesk. Ditto ditto for Maya, Lightwave, 3DS Max, Photoshop, After Effects, and every other major commercial media app out there. OSS alternative may be available and free, but companies aren't going to convert to them just because they're free. They need to be better and free. They need to work well with all the entrenched applications. Without that, the TCO of a conversion far exceeds the cost of the commercial software, so nobody's going to do it except a few OSS diehards willing to put up with the tremendous risks, costs, and headaches associated with a platform move.

Would SL be more stable if it wasn't built on open GL?

It's a very interesting dicussion, and one I wasn't even aware of.

Microsoft has a lot riding on killing open GL. In some ways, Open GL is a competing operating system.

Would SL be more stable if it wasn't built on open GL?

Given that more people use the microsoft technologies, it no doubt gets far greater QA resources than Open GL.

Maybe a lot of the bugs and crashing are caused by not completely QA'd versions of OpenGL rather than something LL is doing wrong.

The new version of Windows being developed "emulates" OpenGL support THROUGH DirectX, but only up to version 1.4. This is Windows Vista, the oft-delayed and features-gutted replacement to Windows XP.

Current features seem to be:

1) Decreased control over your own computer;
2) Increased stranglehold on competetive products (MSIE is being re-integrated, OpenGL is being shafted);
3) A shinier, 3d-accelerated GUI that will slow down your computer,
4) Higher minimum specs.

I'll be sticking with XP for a while, I think.

LF

3) A hardware accelerated, scaleable interface - on lower end video cards, it does not look much different than XP, and it is completely configureable.

4) Windows XP had higher minimum specs than Windows 98 and ME. Mac OS X had higher specs than MacOS 9. What exactly is the point?

Furthermore, most OSS adherents don't like paying for their software

I wouldn't compare OpenGL to an "operating system" as was suggested above. It is merely an API standard to speciailized hardware.

1- DirectX has zero crossplatform compatibility.OpenGL can run on over 47 architectures (and has 3D accelleration on a good deal of them.)

2- DirectX is slower than OpenGL for similar apps, OpenGL also introduces less library bloat into your application in most situations.

3- OpenGL is more versatile,and gives the programmer more control over the final rendering output.This is extremely important for something like SL where standard 3D ground rules go out the window. (no BSP trees, no prerendering or prelighting)

4- OpenGL is a nonproprietry open standard; it's direction is headed by an architecture review board who is controlled by all the major players in the 3D industry; not just Microsoft.

5- OpenGL 2.0 brings OpenGL to, and beyond the latest DirectX feature sets.The current OpenGL 1.4 spec is roughly comparable to DirectX 9, with the exception of some Vertex/Pixel shaders which have to be handled through OpenGL extensions.