OpenGL 3.0

[Luuk van Venrooij]

Has anyone taken a look at the new OGL 3.0 specs released several days ago? None of the promised stuff is in there. It`s more like OGL 2.2. Seems that this could be the end of OGL when compared to D3D. Personally I`m gonna stick with OGL since I use it for my engine.

Some links with stuff and other discussions:

http://www.khronos.org/news/press/re...enerations_of/
http://opengl.org/registry/doc/glspec30.20080811.pdf
http://www.gamedev.net/community/for...opic_id=504547

[pstudio]

Hmm, it sounds like I should focus on learning D3D then instead of OGL.

[waran]

Seems that this could be the end of OGL when compared to D3D.

Oh. When did Microsoft release D3D10 for Windows < Vista and Linux?
Must have missed this event somehow.

[chronozphere]

Oh. When did Microsoft release D3D10 for Windows < Vista and Linux?
Must have missed this event somehow.

There is no such thing as D3D10 for Vista. :lol:
He is talking about the API-war on the windows platform, wich is considered the best gaming platform.

Pitty, I was about to learn OpenGL (I still plan to learn it). IMHO OpenGL is alot more easy to use than D3D, and has a cleaner API. After reading some of those threads it became clear that OpenGL really needed this "great leap forward". For example, loads of OGL extensions should be promoted to core, lots of people wanted an Object Oriented API and Geometry shaders and other next-gen functionality. Maybe the most important thing that people were hoping for was the big cleanup, that would make driver implementation alot easier. It's a shame that none of these "promised" things got there, and it turned out te be another "small step". :cry:

Unlike many other people, i don't think it's time to say: "OpenGL is dead". It's just a big disappointment. I still want to learn OpenGL, and i really hope that they will significantly improve OpenGL during the next 12 month's. It's worth it!

[waran]

OpenGL does support geometry shaders (at least nVidia published
an extension on it - ATi/AMD seems to think this is unnessecary).

Thing is that OpenGL currently is out of alternatives since Microsoft kicked themselves
out of the great competition. Compared to the only available Direct3D on decent
Windows platforms (9) OpenGL 2.2 (aka "3.0") does outstanding well.

I can only speak for myself but Im not going to buy Vista for playing D3D10-Games;
and there's no other reason to buy it. Also hope there are other people with a spine
around so there is some boycott effect on Vista or Direct3D-Games so publishers
as well as Microsoft get the right signals. Its yours to choose what happens in the market.

[Luuk van Venrooij]

I just studied the OGL specs some more and geometry shaders is one of the extensions that are added to the core so that is nice. There is also a new version of Vertex Buffer Objects that is suposed to be simpler then the current workings.

They also introduced a deprication model which says which functionality will be dropped in future versions. For example displaylists, Vertex arrays and inmediate mode will disapear in future versions. Personaly I think this is a bad system. The ARB has been discussing OGL 3 for 3 years now and this model will only slow them down even more. This since they have to start discussing even more for future OGL versions. Are they going to discuss the OGL 4 specs for 6 years now because they need to take in account the stuff they want to drop in OGL 5?? :? Only time will tell.

Personally I hope the reaction of the community to OGL 3.0 will "RE" trigger the design cycle of the OGL api and that the next version will have all the new features and redesign they promised for OGL 3.0.

[JernejL]

Just because they failed now, do not forget that opengl is still the api of choice for cross-platform gaming, and just because things weren't deprecated already it doesn't mean you can't use better ways that are already availble..

[pstudio]

Just because they failed now, do not forget that opengl is still the api of choice for cross-platform gaming, and just because things weren't deprecated already it doesn't mean you can't use better ways that are already availble..

It's the choice because there aren't really anything else to choose, when we're talking 3D API. Besides, If I would publish a game, I want people to play it. Most gamers use Windows, and if DirectX is better than OGL on that platform, I would choose that.

And the deprication model sucks imo. Sure it says you shouldn't use that feature because it's old, but whe has that last stopped a developer?

I'm not the most suited to judge this new specification, but when so many OGL developers are unsatisfied and look towards D3D, something must be terribly wrong.

But ofcourse, OGL has worked so far, and it will probably work for a lot of people still. D3D just sounds like the API of choice, if you want the best right now.

[WILL]

So is there a hope of seeing a newly translated set of an OpenGL 3.0 header for Object Pascal users?

[chronozphere]

OpenGL does support geometry shaders (at least nVidia published
an extension on it - ATi/AMD seems to think this is unnessecary).

Ah okay... I didn't know that. But i those threads, people still mention "Geometry Shaders" now and then. I think they expected it to be core functionality that worked on most of the proper gaming systems. So far, implementing OGL geo-shaders is not worth-while because quite alot of people can use them (ATI + Intel chips).

They also introduced a deprication model which says which functionality will be dropped in future versions. For example displaylists, Vertex arrays and inmediate mode will disapear in future versions.

Does that mean that VBO's are the only option when it comes to rendering? :? Are the considered reliable these days (good driver implementation)?

Compared to the only available Direct3D on decent
Windows platforms (9) OpenGL 2.2 (aka "3.0") does outstanding well.

Nice! That sound really positive. I'm tired of reading all those "It sucks, They failed, OGL = dead" messages.