Archived

This topic is now archived and is closed to further replies.

The new March ARB meeting notes are (finaly) uploaded.

This topic is 4944 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Check this out Sooo: Uber (super) buffers aren''t included in OpenGL 2.0. And all the low level fragment and vertex shading extensions (ARB_fragment_program,ARB_fragment_program_shadow,ARB_vertex_program) are not a part of the OpenGL 2.0 core. But they are still supported, right? Anyway a few new extensions were added (and an ARB floating point texture extension). Funny GLSL wasn''t mentioned.
"C lets you shoot yourself in the foot rather easily. C++ allows you to reuse the bullet!"

Share this post


Link to post
Share on other sites
ah good, been waiting for this

Errm, no mention of GLSL? Did you miss the largeish section "Marketing Working Group Review" which mentions the HLSL (GLSL by any other name) and from the OpenGL2.0 feature selction poll you get this :
quote:

OpenGL Shading Language (ARB_shading_language_100) + related API features (ARB_shader_objects, ARB_vertex_shader, ARB_fragment_shader).
Straw poll: YES, unanimously.





The low lvl vertex/fragment programs will still be supported by IHVs for a while I dare say, just dont expect them to get an update to bring them into line with new hardware, basicaly what you''ve got is what you''ve got.

Some nice stuff on the shortlist, i''m not surprised by the lack of SuperBuffers, its a very complex spec to knock out and the addition of PBO should offset its need in the short term at least, maybe even long term.
Standard floating point buffers is also a good call

I did find it intresting that the discussion on SuperBuffers mentions an ''OpenGL3.0'' spec, hehe, 2.0 isnt even out the door yet and they are already considering what will be required to make a 3.0 spec

Roll on the next meeting and SIGGRAPH04 :D


[Phantom Web | OpenGL Window Framework ]
"i wonder why i do that... type words which are nuffin like the word i wanted every now and badger"

Share this post


Link to post
Share on other sites
quote:
Original post by _the_phantom_
ah good, been waiting for this

Errm, no mention of GLSL? Did you miss the largeish section "Marketing Working Group Review" which mentions the HLSL (GLSL by any other name) and from the OpenGL2.0 feature selction poll you get this :
quote:

OpenGL Shading Language (ARB_shading_language_100) + related API features (ARB_shader_objects, ARB_vertex_shader, ARB_fragment_shader).
Straw poll: YES, unanimously.






I'm talking about some updates and specification changes. There are some things that SHOULD be changed.

quote:
Original post by _the_phantom_
The low lvl vertex/fragment programs will still be supported by IHVs for a while I dare say, just dont expect them to get an update to bring them into line with new hardware, basicaly what you've got is what you've got.



I don't need an update, I just want to make sure that my CG codepath is will be still supported.

quote:
Original post by _the_phantom_
Some nice stuff on the shortlist, i'm not surprised by the lack of SuperBuffers, its a very complex spec to knock out and the addition of PBO should offset its need in the short term at least, maybe even long term.



Still ARB are working too slow. They should work faster and a complex spec is no excuse.





edit: typo




"C lets you shoot yourself in the foot rather easily. C++ allows you to reuse the bullet!"



[edited by - CPPMaster Poppet on May 29, 2004 1:09:33 PM]

Share this post


Link to post
Share on other sites
quote:
Original post by CPPMaster Poppet
I'm talking about some updates and specification changes. There are some things that SHOULD be changed.



Such as?
tbh, I cant find anything i dont like about the spec or dont find logical, only think which might need changing off hand is the need to put .0 at the end of all floats, but thats pretty minor.
Although, given you're Cg comment later I'm guessing you are one of these 'omg, why cant we cast?' ppl, which tbh I dont see as an issue with how its done, maybe its logical in Cg to do C-style casts as its kinda C-for-graphics, however GLSL draws on sections of it for syntax but makes no really claims to be a C-language *shrugs*, out of intrest, do you C-cast in C++ or use C++'s casts?
The other think someone brought up was lack of inversematrix stuff, but i know for a fact thats being corrected (mentioned post ARB meeting).
Oh, and they also mention about sorting out how to handle MRTs and FP buffers as well in GLSL, which is nice

quote:

I don't need an update, I just want to make sure that my CG codepath is will be still supported.



Fair play, I just wouldnt count on anyone other than NV and ATI having them as they are a required part of the spec, but then all new cards are DX9 so I'd hope they would get a GLSL compiler in there instead.

quote:

Still ARB are working too slow. They should work faster and a complex spec is no excuse.



Dont agree with that at all, SuperBuffers is a MAJOR extension, heck they even mentioned an OGL3.0 in conjuntion with bring it in. So they should take the time and hammer out the spec and get all issues sorted before a beta release, because if something is done wrong we will be stuck with a MAJOR part of OpenGL which will bug us for ALONG time.
Also, its not going to be easy for the IHVs to get the damned thing written from my understanding, so they need to make sure everyone is on the same page. I'm not even sure how many cards can support some of the stuff they are talking about, such as render-to-VBO, in hardware atm (certainly the NV40 can, but beyond that i'm not sure, I would suspect the R300's can but dont quote me on that).


[edited by - _the_phantom_ on May 29, 2004 12:13:19 AM]

Share this post


Link to post
Share on other sites