OpenGL 3.0, GLSL 1.30 Released

[phoronix]

Phoronix: OpenGL 3.0, GLSL 1.30 Released

From SIGGRAPH 2008, one of the premiere computers graphics conferences, the Khronos Group has announced the release of the OpenGL 3.0 API specification and the GLSL 1.30 shading language specification. This is the first major update to this cross-platform 3D programming API since the OpenGL 2.1 release two years ago. In this article we have a bit of information on these OpenGL and GLSL updates and when we can expect to see the Linux graphics scene moving to this new standard.

http://www.phoronix.com/vr.php?view=12730

[Nille]

Hey WOW OpenGL 3 nice ! but WTF the OpenSource ATI Drivers Support only OpenGL 1.3 http://wiki.x.org/wiki/RadeonFeature and no GLSL :(

[Regenwald]

afaik comes ogl3 with gallium3d. in other words: it doesn't make sense to put lot of effort into the "old" driver model.

[wiscados]

The word "finally" comes to mind.

[bugmenot]

Yes, yes and yes!
That's what I was waiting for.

[some-guy]

Seems like many people are disappointed: http://www.opengl.org/discussion_boards/ubbthreads.php?ubb=showflat&Number=243267&fpart=1

[R3MF]

how long until it supports my DX10 hardware, and will that be before i have purchased DX11 hardware?
i posit this question pessimistically, due the growing irrelevance of OpenGL to me as i see it.

but i do hope.....................

[pierluc]

Will it be possible to create videos-games with very realistic graphics like directx10 videos-games with this new OpenGL?

[NeoBrain]

Will it be possible to create videos-games with very realistic graphics like directx10 videos-games with this new OpenGL?

In fact, many games that offer both DX9 and DX10 versions don't differ _that_ much in graphics quality, so it's actually possible to do the same "very realistic graphics" of DX10 and DX9. And as most (if not all) DX9 features are also implemented in Opengl 2.x, it's even possible to do these kinds of graphics in "old" Opengl versions.
What OpenGL 3 improves here, however, is the general API look and also it optimizes a few basic concepts as well as introducing a better shader language.
I.e. both DX10 and OGL3 don't make these graphics possible, but make it easier for games to achieve them.

[Melcar]

Wasn't OpenGL3 supposed to just be a "clean up" version, with the really good stuff coming in OpenGL3.1?

[R3MF]

yes, but some might say it is swiftly becoming an irrelevance given how late it is.

[Melcar]

Seems like many people are disappointed: http://www.opengl.org/discussion_boards/ubbthreads.php?ubb=showflat&Number=243267&fpart=1


Reading that depresses me :(.

[energyman]

Seems like many people are disappointed: http://www.opengl.org/discussion_boards/ubbthreads.php?ubb=showflat&Number=243267&fpart=1

people are always disappointed. If the new release would have broken all the apis for the sake of 'clean up' a lot more people would be disappointed.

Whatever you do, you will always see some whining. Just look at this forum and the whining bridgeman& co have to endure (and endure they do, with humour) ;)

[Melcar]

people are always disappointed. If the new release would have broken all the apis for the sake of 'clean up' a lot more people would be disappointed.

Whatever you do, you will always see some whining. Just look at this forum and the whining bridgeman& co have to endure (and endure they do, with humour) ;)


Agree. But then again, the "clean up" was expected.

[Moustacha]

Michael, I hope you're not shooting yourself in the foot about nVidia's conference. But you usually have been spot on the money in the past, so I'll have to wait and see what comes out regarding Linux from nVidia.

[dungeon]

I was not expected that all X devs goes to Intel and now even Ian Romaniac is. He will (for sure), step-by-step, aligning Mesa to work on "every" platform with Larrabee, but not... Who cares for busted & delayed X experiment:mad:? Or for OGL3? Ian's & Transgaming's extensions? HAHAHAHA!!!:D:mad::D

[Heiko]

people are always disappointed. If the new release would have broken all the apis for the sake of 'clean up' a lot more people would be disappointed.

Whatever you do, you will always see some whining. Just look at this forum and the whining bridgeman& co have to endure (and endure they do, with humour) ;)

No, in fact they have a good reason to be upset. A year ago a new api was promised, examples were shown in a newsletter and it would all be wonderfull: a modern, clean, easy to develop for api. An api that could be on par with DirectX 10. Now, a year later, after being quiet for almost a year (with no reasons given why OpenGL 3.0 was delayed) they come up with something that they should have called OpenGL 2.2.

Is it al that bad? Well... there are some positive points. There is a new depricated model that shows almost any fixed pipeline function to be depricated. This means, all depricated functionality being removed in the next version of OpenGL. It also means that it will take until the next version of OpenGL before it really gets easier for driver developers and application developers to develop OpenGL applications. Khronos members told in the opengl forums that the object oriented api is still being worked on, but the problem is that nobody trusts Khronos anymore after a year delay, a year of silence and a totally different api then promised.

Another thing that is different then promised is that you need DirectX 10 capable hardware for OpenGL 3.0 (initial plans were based on minimum R300/NV30). But probably the most important feature: geometry shaders still need an extension. So it is the question if Ati, Nvidia and Intel will support this extension.

Anyway: if you use the forward-compatible profile (profiles are also new), which doesn't use any depricated functions, you have a cleaned up api. But it would be nice if Khronos released a spec without all the depricated stuff in it (which takes up about 2/3 of the spec).

[remm]

I read the reactions of a bunch of DX fanboys on opengl.org. The bottom line however, is that if they wish to sell their games to me, they better not use DX 10+ (should stick with OpenGL or DX 9), due to the Vista+ only support. Some game developers understand this perfectly well.

[NeoBrain]

I think one problem with OpenGL is that it actually was never made for programming games with it. It was a tool to create powerful CAD applications, which didn't even need to be hardware accelerated. Because of its nature, OpenGL now keeps the support for the workstation business and tries to add support for games, but these two things actually have very different needs and thus, creating a "clean" API is very hard as there will always be areas which never get touched in either games or CAD programs.
Ever wondered why D3D is superior to OpenGL today? It's only supposed to be used for games, and as such the API is optimized for that use.
What we would need is a complete restart (yeah, I'm one of these guys who say starting from scratch is always a good idea ;-) ) with a new API that is tied to programming games only and let the OGL ARB do whatever they want to go back to workstation-only. The only problem with this approach is that it would probably damn hard to get the support from major companies like ATI or NVIDIA for that (however, that reason has never prevented some OSS developers to implement their ideas).

[rbmorse]

Fork! Fork! Fork! Fork!

[MetalheadGautham]

OMG! Boom goes plans to get a Radeon HD4850. How long before we see hardware support for OpenGL 3.0 ? I guess its direct competition is going to be DirectX 11.0, since OpenGL 2.1 takes care of DirectX 10.

Is it going to enter the kernel soon ? When ?

[energyman]

who programs games on bare ogl+glut anyway? sdl and friends are there for a reason.

A fork would be stupid - it would kill ogl and the fork.

[numasan]

I've also been waiting long for this announcement, but I'm not a developer so I'm sad to see the reactions. I do work with OpenGL apps as a 3D/compositing artist, and it's only been lately that 3D DCC apps have included OGL 2(.1) functions, so I guess OGL 3(.1) functions are a way off still.

About games, id Software still uses OpenGL for "Rage" the new title they are working on. It is IMO the best looking realtime graphics I've seen yet, so I don't understand those of you who think DX10 is superior.

[deanjo]

OMG! Boom goes plans to get a Radeon HD4850. How long before we see hardware support for OpenGL 3.0 ? I guess its direct competition is going to be DirectX 11.0, since OpenGL 2.1 takes care of DirectX 10.

Is it going to enter the kernel soon ? When ?

The hardware is already out, has been for a while. a DX10 card has no issues running OGL3.

[flami]

I've also been waiting long for this announcement, but I'm not a developer so I'm sad to see the reactions. I do work with OpenGL apps as a 3D/compositing artist, and it's only been lately that 3D DCC apps have included OGL 2(.1) functions, so I guess OGL 3(.1) functions are a way off still.

About games, id Software still uses OpenGL for "Rage" the new title they are working on. It is IMO the best looking realtime graphics I've seen yet, so I don't understand those of you who think DX10 is superior.

Its not the features that are the problem , its that they promised a modern API , as in easily to maintain , 1 way not 5 to get 1 result, object based etc.
But this didnt happen, nothing got deleted , so theres still 5 ways to do the same, its not object based, and because of that its hard to maintain your code. ( and drivers devs will have fun implementing 5 ways to do the same stuff to suport all the people that use them) .

D3D changed its API several times ( and lost backwards compatibiltiy ), to meet the requirements of modern APIs. ARB promised a similar aproach for OGL3, but the API still is similar to the one used in the 90s . This is the main reason people say D3D is superior to OGL.

This is the big disapointment, its not about features.
( imagine you are promised a nice sports car and instead you get a raceboat , its nice but its not what you wanted )

[hochglanz]

... The only problem with this approach is that it would probably damn hard to get the support from major companies like ATI or NVIDIA for that (however, that reason has never prevented some OSS developers to implement their ideas).

I don't think that hardware support is too hard, because no vendor want to be dependend on only one player, which dictates the market and therefore everything.

Well, it was a rather bad day! Hope that throught OpenCL maybe also OpenGL will make a leap ahead, someday...

[Moustacha]

Fork! Fork! Fork! Fork!

Spork! Spork! Spork! Spork!

[Svartalf]

Seems like many people are disappointed: http://www.opengl.org/discussion_boards/ubbthreads.php?ubb=showflat&Number=243267&fpart=1

I'm a bit disappointed, if you want to know the truth of the matter- but I don't come away with the idea that "we're behind the times" and have been so since D3D 8.0 (Which is BOGUS. And it's not "easier" to code for D3D versus OpenGL- it's DIFFERENT and it makes some things easier at the expense of allowing sloppy code to go down.). Most of the eye candy potential is there with OpenGL 2.1 in the first place- all the PS3 titles sort of prove this one out already.

I would rather have liked to get what they tried to give us with Longs Peak originally- but most of the cruft goes bye-bye with OpenGL ES 2.0 and that's relevant for a lot of platforms (Wii, PS3, Pandora, other non-desktop solutions...) and can be made available via a thin abstraction shim or a native API edge on pretty much any platform with OpenGL 2.0 available to it.

[edit]

Perhaps we ought to be working at getting ES to be more the primary API and then force CAD-type apps down the ES 1.X leg which won't break anything for them and then do everything else in the ES 2.X leg for games, etc. ES 2.0 is the nice, lean-n-clean API, for the most part, that people want out of OpenGL (one good way of doing things, etc...) and is more akin to what we need in Linux. Honest. There's really two APIs here and they can both be supported and be cross-platform.

[Svartalf]

In fact, many games that offer both DX9 and DX10 versions don't differ _that_ much in graphics quality, so it's actually possible to do the same "very realistic graphics" of DX10 and DX9. And as most (if not all) DX9 features are also implemented in Opengl 2.x, it's even possible to do these kinds of graphics in "old" Opengl versions.
What OpenGL 3 improves here, however, is the general API look and also it optimizes a few basic concepts as well as introducing a better shader language.
I.e. both DX10 and OGL3 don't make these graphics possible, but make it easier for games to achieve them.

Folks...Read and re-read this over and over until it soaks in.

This is what's going on here. And the people doing the strongest bitching actually are in the "a poor artist blames his tools" category of games and application developers. It's not that difficult to "find the fast path" and a bunch of other things they were griping about in that thread. Oftentimes D3D produces opaque code that's hard to follow- in all honesty, some of this stuff they do, I can't envision people thinking it was "easier" than OpenGL was in the long run.

[Svartalf]

D3D changed its API several times ( and lost backwards compatibiltiy ), to meet the requirements of modern APIs. ARB promised a similar aproach for OGL3, but the API still is similar to the one used in the 90s . This is the main reason people say D3D is superior to OGL.

Heh... Breaking at every rev being superior? Niiice.

Anyone that tells you that it changes every time you turn around and come up with the next version, but is "similar" and has some smattering of backwards compatibility/emulation, but not always and not consistent when it did provide something- and this person calls it "superior" is selling something.

API stability is important for even games. Tracking D3D is "fun" in the masochistic sense of things and people use it because it's the "easy" way to do things under Windows because MS made it that way.

D3D is only "superior" because it followed the whims of MS and gave the devs precisely what they asked for. That's nice in a way, but it leads to all sorts of twitchy code and games that plain flat won't run 1-2 years after their sale. Now, that might be good for causing churn for games, but it does NOTHING for the poor bastards writing CAD and other serious 3D tools- which have tended to avoid D3D even when it's "superior" on Windows for some odd reason...wonder why?

[flami]

Just summed up what the people in that nice huge threat in the OLG forums are bi***ing about .

anyway its the others fault ^^

[R3MF]

Folks...Read and re-read this over and over until it soaks in.

This is what's going on here. And the people doing the strongest bitching actually are in the "a poor artist blames his tools" category of games and application developers. It's not that difficult to "find the fast path" and a bunch of other things they were griping about in that thread. Oftentimes D3D produces opaque code that's hard to follow- in all honesty, some of this stuff they do, I can't envision people thinking it was "easier" than OpenGL was in the long run.

i spend a boat load on computer hardware because that's what i like to do, if OpenGL 3.0 do not use that hardware (my 9800GX2) properly then i have a diminished interest in OpenGL 3.0. it is that simple.

[drag]

Remember there is a difference between what DirectX does and what OpenGL does, right? I mean DirectX does sound, input, 2D graphics, and all sorts of stuff like that.

In Linux-land we already have all that stuff with SDL, and it's a much cleaner and easier to program for then DirectX. If you want easy to program for stuff you do SDL. If you want easy to program for stuff, but want to get all twitchy with high-end graphics and such then you go OGL for 3D and SDL for everything else.


And I can understand why they don't wanted to 'objects' in a very general sense. I don't know much about OpenGL, but I know about programming stuff in general.

OpenGL is C-level stuff, right? Very low level.

Well... C can be every bit as object oriented as anything else. I am not saying that it's pleasent, but you have to realize that C++ is written in C, so there is absolutely nothing that you can do in C++ that you can't do in C. It does everything.. if you want object oriented design you program in a object oriented manner.

Of course your losing a lot of prepared work that other people have put into C++. Not saying it's wonderful and a good choice to stay with C, just that it doesn't restrict you from doing anything.

For example the Linux kernel is C and it follows a strong object oriented design with all sorts of massive amounts of code re-use and abstractions. It's much more dynamic then any other sort of large C++ project.

So if you want to use objects and you don't want to write all that crap yourself then you don't program in OpenGL. You program in software that is built on OpenGL.

What sort of game designer actually programs in a lot of Direct3D or OpenGL anyways? Maybe a little bit of hacking for special effects, but he people that program in OpenGL are not going to be game makers, they are gaming engine makers.

It's like if I want to make a applet for setting screen resolutions or something like that I am not going to go around hacking with developing new kernel syscalls or anything like that.. No, I'll just use python and gtk for the GUI and use X drivers to provide the functionality. If I want to always program using low-level interfaces I'll never get anything done.

-------------------------


Plus with the advent of CPU-like GPU with the programming models they will bring and the sort of flexibility that things like Gallium3D brings...

Do you really think that OpenGL is going to be the only path to the hardware?

I mean, if you can program in regular C, C++, or Python and have your code compile with LLVM and target both the GPU and CPU based on which will be optimal for your program.. Do you really want to be farting around with GLSL shader languages to make your application go fast?

So with Gallium3D (if it ever gets released) then you can develop all sorts of new ways to interact and program for the card that go far beyond anything you'd ever want to stick in OpenGL. It should be very API agnostic. The possibilities for different API models are quite high and I doubt anybody wants to see everything in the world stuffed into OpenGL.

----------------------------


Like I said I don't know much about OpenGL or really anything much to do with graphics.

But unless there is something fundamentally important you can't do in OpenGL that you can do in DirectX, then it may be just a better idea to leave OGL as low-level and flat as possible and let other people build new stuff from it that is easier to program for.

[Redeeman]

Ever wondered why D3D is superior to OpenGL today? It's only supposed to be used for games, and as such the API is optimized for that use.
What we would need is a complete restart (yeah, I'm one of these guys who say starting from scratch is always a good idea ;-) ) with a new API that is tied to programming games only and let the OGL ARB do whatever they want to go back to workstation-only. The only problem with this approach is that it would probably damn hard to get the support from major companies like ATI or NVIDIA for that (however, that reason has never prevented some OSS developers to implement their ideas).
except that d3d isnt superior, just because the propaganda says so, doesent make it true.

[Dragoran]

Several NVIDIA representatives have invited us to NVISION, so we do believe something significant could happen for Linux there, but we have yet to learn whether it pertains to an open-source strategy or Big Bang II on Linux or NVIDIA PR just trying to garner some extra attention. Though because of this, we have decided not to attend.

LINUX Genius Bar at the exhibition: Spend 1:1 time with NVIDIA’s LINUX driver experts on the exhibition floor

http://www.nvision2008.com/Professionals/index.cfm

[NeoBrain]

http://www.nvision2008.com/Professionals/index.cfm

That would give Michael a chance to do an interview with the NVIDIA Linux driver guys, however I think he'd just get the answer "no we aren't interested in OSS drivers" again, so I perfectly understand his opinion concerning the nvision.

[Regenwald]

http://www.nvision2008.com/Professionals/index.cfm

hm, i don't hope that this is all.