Comments on: Why OpenGL 3.0 is Important http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/ The informal yet informational sub-site Thu, 11 Mar 2010 04:55:54 +0000 http://wordpress.org/?v=2.9.2 hourly 1 By: OpenGL 3.0: Well, that explains it. | Scriptionary Blog http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-441 OpenGL 3.0: Well, that explains it. | Scriptionary Blog Fri, 22 Aug 2008 19:03:43 +0000 http://blog.scriptionary.com/?p=21#comment-441 [...] a little history (May, 2008): In may I wrote a blog entry about the then upcoming OpenGL 3.0 based on information released by the Khronos ARB. The entry was [...] [...] a little history (May, 2008): In may I wrote a blog entry about the then upcoming OpenGL 3.0 based on information released by the Khronos ARB. The entry was [...]

]]> By: Neonlabs » Blog Archive » OpenGL 3.0 Spec Released Today http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-421 Neonlabs » Blog Archive » OpenGL 3.0 Spec Released Today Thu, 21 Aug 2008 05:49:23 +0000 http://blog.scriptionary.com/?p=21#comment-421 [...] and it was a complete disappointment (worse than version 2.0). Not even one feature that was promised nearly two years ago is in the spec for version 3.0. The Khronos Group has failed to provide a [...] [...] and it was a complete disappointment (worse than version 2.0). Not even one feature that was promised nearly two years ago is in the spec for version 3.0. The Khronos Group has failed to provide a [...]

]]> By: OpenGL 3.0 uccide OpenGL e le speranze degli utenti Mac/Linux « L’importante è avere un titolo http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-396 OpenGL 3.0 uccide OpenGL e le speranze degli utenti Mac/Linux « L’importante è avere un titolo Sun, 17 Aug 2008 20:30:35 +0000 http://blog.scriptionary.com/?p=21#comment-396 [...] aveva promesso, ad esempio, un’API orientata agli oggetti, seppur ancora basata su C. Un rifacimento quasi [...] [...] aveva promesso, ad esempio, un’API orientata agli oggetti, seppur ancora basata su C. Un rifacimento quasi [...]

]]> By: Free Culture News » Blog Archive » OpenGL 3.0 Released http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-372 Free Culture News » Blog Archive » OpenGL 3.0 Released Thu, 14 Aug 2008 14:24:36 +0000 http://blog.scriptionary.com/?p=21#comment-372 [...] Scriptionary [...] [...] Scriptionary [...]

]]> By: OpenGL 3.0 publicat de Khronos Group « Fly on the Windscreen http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-369 OpenGL 3.0 publicat de Khronos Group « Fly on the Windscreen Wed, 13 Aug 2008 07:39:10 +0000 http://blog.scriptionary.com/?p=21#comment-369 [...] multe despre de ce e important OpenGL 3.0 pe Scriptionary. Publicat [...] [...] multe despre de ce e important OpenGL 3.0 pe Scriptionary. Publicat [...]

]]> By: martin b.r. » Blog Archive » OstionGL 3.0 http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-368 martin b.r. » Blog Archive » OstionGL 3.0 Tue, 12 Aug 2008 13:44:45 +0000 http://blog.scriptionary.com/?p=21#comment-368 [...] las esperadas espeficicaciones de OpenGL 3. Nada de lo prometido, adios a la principal novedad, los objects y a la reorganizacion general. En muchos foros ya felicitan a Microsoft por ganar la guerra de las [...] [...] las esperadas espeficicaciones de OpenGL 3. Nada de lo prometido, adios a la principal novedad, los objects y a la reorganizacion general. En muchos foros ya felicitan a Microsoft por ganar la guerra de las [...]

]]> By: OpenGL 3.0 - 1 hour after | Scriptionary Blog http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-366 OpenGL 3.0 - 1 hour after | Scriptionary Blog Mon, 11 Aug 2008 14:57:08 +0000 http://blog.scriptionary.com/?p=21#comment-366 [...] reading the spec and looking desperately for the promised object model, I felt quite like a douche bag taking the newsletters seriously and writing about them so [...] [...] reading the spec and looking desperately for the promised object model, I felt quite like a douche bag taking the newsletters seriously and writing about them so [...]

]]> By: ¿Para cuando OpenGL 3? « culleredo http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-72 ¿Para cuando OpenGL 3? « culleredo Sun, 25 May 2008 14:27:04 +0000 http://blog.scriptionary.com/?p=21#comment-72 [...] pensado para competir con DirectX 10 y no quiero meterme en el apartado técnico. Desarrollar con OpenGL 3 sera mas beneficioso para cualquier compañía que no este atada con contratos a sistemas cerrados, [...] [...] pensado para competir con DirectX 10 y no quiero meterme en el apartado técnico. Desarrollar con OpenGL 3 sera mas beneficioso para cualquier compañía que no este atada con contratos a sistemas cerrados, [...]

]]> By: Aras Pranckevičius http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-68 Aras Pranckevičius Mon, 19 May 2008 18:06:03 +0000 http://blog.scriptionary.com/?p=21#comment-68 I don't mind D3D changing the API with each major revision. API itself is nothing; it's really easy to rewrite "the same stuff" to another API. Underlying hardware is still the same, so if the API is not Completely Stupid, it's not a big deal. What I <i>do</i> mind though, is whether API and everything related to it (drivers) <i>work</i>. OpenGL fails miserably in this regard, with it's approach of "let's let each IHV make their own bugs in the whole stack!" it's just horrible. FBOs? GLSL? Argh! Another thing that I don't mind that much is all the legacy APIs/extensions. There's probably ten ways to submit vertex data, while in fact there should be vertex buffers only. And even then, VBOs in GL don't match up what was available in D3D for years (updating a portion of VBO without additional memory copy, anyone?). So in my view GL3 does not bring anything new to the table. Ok, maybe it will be simpler for IHVs to implement, which would take care of some bugs in the drivers. But what they should really do is force a common stack upon everyone, and make only the lowest-level driver be a responsibility of IHV. I don’t mind D3D changing the API with each major revision. API itself is nothing; it’s really easy to rewrite “the same stuff” to another API. Underlying hardware is still the same, so if the API is not Completely Stupid, it’s not a big deal.

What I do mind though, is whether API and everything related to it (drivers) work. OpenGL fails miserably in this regard, with it’s approach of “let’s let each IHV make their own bugs in the whole stack!” it’s just horrible. FBOs? GLSL? Argh!

Another thing that I don’t mind that much is all the legacy APIs/extensions. There’s probably ten ways to submit vertex data, while in fact there should be vertex buffers only. And even then, VBOs in GL don’t match up what was available in D3D for years (updating a portion of VBO without additional memory copy, anyone?).

So in my view GL3 does not bring anything new to the table. Ok, maybe it will be simpler for IHVs to implement, which would take care of some bugs in the drivers. But what they should really do is force a common stack upon everyone, and make only the lowest-level driver be a responsibility of IHV.

]]> By: Dave http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-67 Dave Mon, 19 May 2008 09:38:10 +0000 http://blog.scriptionary.com/?p=21#comment-67 I think exactly the same to #Korval. Every version of DirectX has the great disvantage of relying on a binary format for method calling and structs memory layout that is completely incompatible between major versions. OpenGL has a smarter design in the sense of using raw C functions (instead of virtual C++ like methods) and almost no C structs. But it is released with a functional equivalence to DX almost 2 years before. Otherways, using non-standar extensions is a nightmare. Current OpenGL version has a lot of inconsistences, and a lot of legacy unusefull methods. But the modern design of DX is not the only advantage over OGL, but a consistent DirectX SDK with a compact documentation and a pletoria of examples. I think exactly the same to #Korval.
Every version of DirectX has the great disvantage of relying on a binary format for method calling and structs memory layout that is completely incompatible between major versions.
OpenGL has a smarter design in the sense of using raw C functions (instead of virtual C++ like methods) and almost no C structs.
But it is released with a functional equivalence to DX almost 2 years before. Otherways, using non-standar extensions is a nightmare.
Current OpenGL version has a lot of inconsistences, and a lot of legacy unusefull methods.

But the modern design of DX is not the only advantage over OGL, but a consistent DirectX SDK with a compact documentation and a pletoria of examples.

]]> By: Korval http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-60 Korval Fri, 16 May 2008 18:01:51 +0000 http://blog.scriptionary.com/?p=21#comment-60 Nobody really says that GL 3.0 isn't important. But the fact is that it's too late. There was a pretty good window of opportunity for a new and improved GL API to swoop in and garner attention among PC game developers. That time has passed. Vista adoption is up, which means that D3D10 usage will be up. The break between Vista/XP and D3D10/9 (ie: if you use XP, you can't have D3D10 features) was a good opportunity that the ARB has squandered. Furthermore, if the GL ARB couldn't get this done last year, why should we believe them this year? The fact is that the ARB has failed, monstrously. They did it before back when they were working on render-to-texture. They spent 2 years on something that shouldn't have taken more than 6. The younger GL programmers may not remember the days before VBO, but it took an awfully long time to get VBOs as well; until then, we had to rely on a bunch of vendor-specific extensions. I love the ideas behind GL 3.0; I even wrote my GL 2.1-based rendering abstraction layer based on those ideas. But the fact is that it's very, very late. And even that wouldn't be so much of a problem if they could just tell us WHY it's late. But they won't. Which means it's either IP or some stupid argument about something that matters very little. And if it's the latter, then that means that the ARB is so worried about making a mistake that they won't actually produce anything. Which makes them ineffective. For 1 year, the ARB was good about updating us on their progress. And then they stopped. That shows a blatant lack of respect for us the potential users of OpenGL. So why should we respect them back? Nobody really says that GL 3.0 isn’t important. But the fact is that it’s too late. There was a pretty good window of opportunity for a new and improved GL API to swoop in and garner attention among PC game developers. That time has passed. Vista adoption is up, which means that D3D10 usage will be up. The break between Vista/XP and D3D10/9 (ie: if you use XP, you can’t have D3D10 features) was a good opportunity that the ARB has squandered.

Furthermore, if the GL ARB couldn’t get this done last year, why should we believe them this year? The fact is that the ARB has failed, monstrously. They did it before back when they were working on render-to-texture. They spent 2 years on something that shouldn’t have taken more than 6. The younger GL programmers may not remember the days before VBO, but it took an awfully long time to get VBOs as well; until then, we had to rely on a bunch of vendor-specific extensions.

I love the ideas behind GL 3.0; I even wrote my GL 2.1-based rendering abstraction layer based on those ideas. But the fact is that it’s very, very late. And even that wouldn’t be so much of a problem if they could just tell us WHY it’s late. But they won’t. Which means it’s either IP or some stupid argument about something that matters very little. And if it’s the latter, then that means that the ARB is so worried about making a mistake that they won’t actually produce anything. Which makes them ineffective.

For 1 year, the ARB was good about updating us on their progress. And then they stopped. That shows a blatant lack of respect for us the potential users of OpenGL. So why should we respect them back?

]]> By: fldz http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-59 fldz Fri, 16 May 2008 17:56:56 +0000 http://blog.scriptionary.com/?p=21#comment-59 I saw the word "class" somewhere in the specs or proposals, etc. (I must've passed out or put up a mental-block, so.. I wouldn't know where in the specs) They would never go down that road even if they could "support" OOP, would they? "does not support the OOP (Object Oriented Programming) paradigm but a way around this is being implemented." way around what, you mean speed? Flexibility perhaps? Imagination? -a geek stuck in the 80's I saw the word “class” somewhere in the specs or proposals, etc. (I must’ve passed out or put up a mental-block, so.. I wouldn’t know where in the specs)

They would never go down that road even if they could “support” OOP, would they?

“does not support the OOP (Object Oriented Programming) paradigm but a way around this is being implemented.”

way around what, you mean speed? Flexibility perhaps? Imagination?

-a geek stuck in the 80’s

]]> By: E. Luten http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-57 E. Luten Fri, 16 May 2008 17:00:43 +0000 http://blog.scriptionary.com/?p=21#comment-57 <strong>Lionel</strong>, you're right but keep your eyes open during the next few months, I've got a <em>good</em> feeling that SIG08 will be the great turning point. Lionel, you’re right but keep your eyes open during the next few months, I’ve got a good feeling that SIG08 will be the great turning point.

]]> By: Lionel Barret http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-56 Lionel Barret Fri, 16 May 2008 16:54:11 +0000 http://blog.scriptionary.com/?p=21#comment-56 @E. Luten : I really hope it will be but in the past, Khronos/ARB have been so slow that I am not so sure. @E. Luten : I really hope it will be but in the past, Khronos/ARB have been so slow that I am not so sure.

]]> By: The Geeks Of 3D » Why OpenGL 3.0 is Important http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-55 The Geeks Of 3D » Why OpenGL 3.0 is Important Fri, 16 May 2008 16:53:14 +0000 http://blog.scriptionary.com/?p=21#comment-55 [...] post will attempt to take you through the pre-published materials on the OpenGL API, version 3.0 and [...] [...] post will attempt to take you through the pre-published materials on the OpenGL API, version 3.0 and [...]

]]> By: E. Luten http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-54 E. Luten Fri, 16 May 2008 16:36:08 +0000 http://blog.scriptionary.com/?p=21#comment-54 <strong>@Lionel:</strong> An update on OpenGL 3.0 will be made at the OpenGL BOF at SIGGRAPH 2008 this August. Judging that there's a year in between the last 3.0 announcement at SIGGRAPH 2007, this will probably be <em>it</em>. @Lionel: An update on OpenGL 3.0 will be made at the OpenGL BOF at SIGGRAPH 2008 this August. Judging that there’s a year in between the last 3.0 announcement at SIGGRAPH 2007, this will probably be it.

]]> By: nerdd.net | news and opinion http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-53 nerdd.net | news and opinion Fri, 16 May 2008 08:14:31 +0000 http://blog.scriptionary.com/?p=21#comment-53 <strong>Why OpenGL 3.0 is Important | nerdd.net...</strong> \r\nWhen discussing OpenGL 3.0, one question of them keeps popping up: Why do we need OpenGL 3.0? Th... Why OpenGL 3.0 is Important | nerdd.net…

\r\nWhen discussing OpenGL 3.0, one question of them keeps popping up: Why do we need OpenGL 3.0? Th…

]]> By: Lionel Barret http://blog.scriptionary.com/2008/05/15/why-opengl-30-is-important/comment-page-1/#comment-52 Lionel Barret Fri, 16 May 2008 07:31:52 +0000 http://blog.scriptionary.com/?p=21#comment-52 Sadly, the release date seems not very close. I don't follow closely the advancement (I should tho). Any info on that ? Sadly, the release date seems not very close.

I don’t follow closely the advancement (I should tho). Any info on that ?

]]>