OpenGL3 at SIGGRAPH

GLX

Newcomer
http://opengl3.org/

"OpenGL 3.0 Specification Overview
OpenGL hardware and driver plans - AMD, Intel, NVIDIA
Developer's perspective on OpenGL 3.0
The new Khronos Compute Working Group and how that affects OpenGL
and more..."

If there is an overview of a specification, does that mean there's finally a specification?
 
Well, the OpenGL guys seem to like Siggraph... Nothing wrong with Siggraph, in fact, I really loved it when I had the opportunity to be there in 2006, but if they missed the first Siggraph 2007 schedule it would have been nice if they didn't make us wait until Siggraph 2008. This last year there has been essentially zero new updates on the API and how it's progressing. From an outside perspective, it looks almost like they've just waited for Siggraph 2008 instead of say releasing the spec in the beginning of the year. In the mean time the API has been slowly but surely sliding into irrelevance ...
 
http://opengl3.org/

"OpenGL 3.0 Specification Overview
OpenGL hardware and driver plans - AMD, Intel, NVIDIA
Developer's perspective on OpenGL 3.0
The new Khronos Compute Working Group and how that affects OpenGL
and more..."

If there is an overview of a specification, does that mean there's finally a specification?
I really hope, its overdue for 6 months or so? :rolleyes:
If they say something like "We will be providing specs in Arial 12pt" im surely gonna be a tad dissapointed.
 
No-one really expected anything else, as far as I know anyway, right?
I mean, at least for all i've read and understood, 3.0 wasn't meant to be anything but a "rewrite" of the API with the next version supposed to bring the new features bringing OpenGL features to match "DX10 level" hardware

It was expected to be a 100% from the ground up rewrite.
 
No-one really expected anything else, as far as I know anyway, right?
I mean, at least for all i've read and understood, 3.0 wasn't meant to be anything but a "rewrite" of the API with the next version supposed to bring the new features bringing OpenGL features to match "DX10 level" hardware

3.0 was expected and promised to be a completely new API with new, simplified and straightforward programming model. For political reasons ARB decided to keep the old API.
 
In the mean time the API has been slowly but surely sliding into irrelevance ...

I think that is arguable given how "relevant" DX10 has been given that it isn't supported on XP and isn't supported on the consoles.

GL prior to 3.0 was supporting DX9 level features, which is still the API level target for just about all games. With GL3, DX10 level support is core in the API, and not only will you see GL3 support on XP from NVidia (I'm guessing here, because I have any insider info), other vendors might also support this as well.

So in this regard GL3 certainly has some serious potential market advantages over DX10, not only will be it supported on Macs, and Linux, but also probably have better support on windows than DX10 (because of XP)...
 
Last edited by a moderator:
As far as I understood some CAD companies were against the new APi because it would - ahh - break their precious badly written rendering code. That is what I call politics ;) The IHVs themselves seem to want the new API badly, because it means greatly simplified driver developement and better driver quality. But hey... what do you want from a consortium that has so many members? It is known that the larger the group the less work it can do...

In my opinion someone (hey Apple :) )should just take the lead and publish the alternative OS-independend API. With some luck it may become the new standart. Anyway, the current GL programming model (selectors and stuff) is not up to date for the modern world and badly needs a replacement.
 
Last edited by a moderator:
The problem is, the rest of the world (non-Microsoft) has no alternative. The OGL API is reasonably abstract and can be ported to almost every platform, DirectX is tightly woven to MS platform concepts.

If Microsoft was smart, they'd create an "OpenDirectX" divorced from Windows that could be ported to run on OSX, Linux, non-WinCE mobile platforms, etc. They'd get a lot of goodwill, and driven the nail in GL's coffin.
 
How do you get IHVs to work on a third API drivers? It's not like they can ditch OpenGL.
It's a lose lose situation, OpenGL is (unfortunately) not going to die anytime soon, at the same time we are stuck with an API that imho didn't make any sense 15 years ago and makes even less sense now.
 
It's a lose lose situation, OpenGL is (unfortunately) not going to die anytime soon, at the same time we are stuck with an API that imho didn't make any sense 15 years ago and makes even less sense now.

Could you elaborate? GL was clearly superior to DX till about DX9.0 both feature wise and API-wise. GLs stagnation started 5-6 years ago because ARBs incompetence.
 
Isn't it possible to reverse this unfortunate event or will most of the OGL dev community drop it if they can? There must be some drive to fix what's wrong and since most devs seem to openly show their displeasure It wouldn't be that far fetched to see the Khronos group going back. I don't work with this, so I don't know how it works in the industry. But being that OGL is important to many platforms It'd be weird if they didn't try and redo this to keep or gain some confidence again.
 
If Microsoft was smart, they'd create an "OpenDirectX" divorced from Windows that could be ported to run on OSX, Linux, non-WinCE mobile platforms, etc. They'd get a lot of goodwill, and driven the nail in GL's coffin.

But what would Microsoft gain on that? I think they (unfortunately) make perfect business sense now. They have no particular interest in killing OpenGL. Winning an API war doesn't bring any cash in. DirectX in itself is not generating any profits for Microsoft. It's all about selling copies of Windows and preserving the OS monopoly. DirectX makes Windows the only gaming OS for PC, and Linux and Mac is never going to gain any serious market share unless you can also play games. And with Xbox it's of course also about making sure the best games comes to their platform of choice. To me it seems like MS started getting it right about when they went into the console business. I'm not sure if that's because they got some hardware experience on their own, or simply because they put more effort into DirectX because of it.

Now if they were to release an OpenDirectX, then that would be inviting Linux and Mac into the gaming community and making these platforms a lot more attractive for a lot of people. I don't see the business sense of that from Microsoft's point of view, although I would applaude the move if they did. The best thing for Microsoft would be if DirectX kept widening the gap, while OpenGL stayed around and bogged everyone else down.

Instead, I think Apple and the Linux community must realize OpenGL is a burden at this point. Quite frankly, no one cares (or at least no one should care) about syntatic sugar and convenience functions (that belongs in D3DX/GLU/GLUT anyway) and the exact functions that you call to make things draw on the screen is irrelevant. All that is going to be stuffed behind your own abstraction anyway if your code is well written.

The best thing would be if there was an opensource innitiative to create that OpenDirectX that would bring native bindings into the Linux kernel and allow IHVs to write drivers for it. And then write a GL-to-DX wrapper and deprecate the OpenGL support. In the long term that would be a gain for the IHVs too. Only one API to support.
And perhaps also take a page out of Microsoft books, "embrace and extend", to bring extensions to the DirectX as well and let you do fancy stuff on Linux and Mac that the native interface in Windows can't do.

Quite frankly, I think that's the only way anyone's ever going to break Microsoft's monopoly and bring some competition to the OS market.

Isn't it possible to reverse this unfortunate event or will most of the OGL dev community drop it if they can? There must be some drive to fix what's wrong and since most devs seem to openly show their displeasure It wouldn't be that far fetched to see the Khronos group going back. I don't work with this, so I don't know how it works in the industry. But being that OGL is important to many platforms It'd be weird if they didn't try and redo this to keep or gain some confidence again.

If you take twice the time and deliver less than half of what was promised, I don't know what will give people back the confidence in it. And it's not the first time either. The API was also going to be rewritten for OpenGL 2.0. There were a lot of visionary talk and in the end we got a bunch of extensions promoted to core. People were disappointed then too. This time people really expected those visionary talks to finally come true. And all we got was a bunch of extensions promoted to core. No one's going to believe a word of what they say from now on. It's a lost cause at this point as far as I'm concerned.
 
Back
Top