PDA

View Full Version : New NV extensions...



Nutty
01-23-2002, 11:48 AM
I was trying out these 27.00 drivers that seem to have fallen off the back of one of Nvidia's lorry's. The rumours about new extensions are true.. heres the ones that look new to me, from the extension string.

GL_NV_occlusion_query
GL_NV_vertex_array_range2
GL_NV_vertex_program1_1

Can't wait for the new specs for these to be published..

Humus
01-23-2002, 11:51 AM
Isn't NV_vertex_array_range2 quite old?

Nutty
01-23-2002, 11:53 AM
I dont know.. it's not in my current spec..
perhaps I should get a new one..

Ysaneya
01-23-2002, 11:55 AM
I don't think GL_NV_vertex_array_range2 is new, i'm already using it in my program, and i think you can find the specs in the opengl extension registry. If i remember well, it's a little modification of the "normal" VAR extension which doesn't cause an internal flush in some situations.

The GL_NV_occlusion_query sounds interesting. Is/will it be like the HP occlusion test ? I've always been willing to use it in my programs, but it's only present on very few cards, limiting its interest.

Y.

Nutty
01-23-2002, 11:57 AM
HP occlusion test is listed also..

heres the full list.

GL_ARB_imaging
GL_ARB_multisample
GL_ARB_multitexture
GL_ARB_texture_border_clamp
GL_ARB_texture_compression
GL_ARB_texture_cube_map
GL_ARB_texture_env_add
GL_ARB_texture_env_combine
GL_ARB_texture_env_dot3
GL_ARB_transpose_matrix
GL_S3_s3tc
GL_EXT_abgr
GL_EXT_bgra
GL_EXT_blend_color
GL_EXT_blend_minmax
GL_EXT_blend_subtract
GL_EXT_compiled_vertex_array
GL_EXT_draw_range_elements
GL_EXT_fog_coord
GL_EXT_multi_draw_arrays
GL_EXT_packed_pixels
GL_EXT_paletted_texture
GL_EXT_point_parameters
GL_EXT_rescale_normal
GL_EXT_secondary_color
GL_EXT_separate_specular_color
GL_EXT_shared_texture_palette
GL_EXT_stencil_wrap
GL_EXT_texture3D
GL_EXT_texture_compression_s3tc
GL_EXT_texture_edge_clamp
GL_EXT_texture_env_add
GL_EXT_texture_env_combine
GL_EXT_texture_env_dot3
GL_EXT_texture_cube_map
GL_EXT_texture_filter_anisotropic
GL_EXT_texture_lod
GL_EXT_texture_lod_bias
GL_EXT_texture_object
GL_EXT_vertex_array
GL_EXT_vertex_weighting
GL_HP_occlusion_test
GL_IBM_texture_mirrored_repeat
GL_KTX_buffer_region
GL_NV_blend_square
GL_NV_copy_depth_to_color
GL_NV_evaluators
GL_NV_fence
GL_NV_fog_distance
GL_NV_light_max_exponent
GL_NV_multisample_filter_hint
GL_NV_occlusion_query
GL_NV_packed_depth_stencil
GL_NV_register_combiners
GL_NV_register_combiners2
GL_NV_texgen_reflection
GL_NV_texture_compression_vtc
GL_NV_texture_env_combine4
GL_NV_texture_rectangle
GL_NV_texture_shader
GL_NV_texture_shader2
GL_NV_vertex_array_range
GL_NV_vertex_array_range2
GL_NV_vertex_program
GL_NV_vertex_program1_1
GL_SGIS_generate_mipmap
GL_SGIS_multitexture
GL_SGIS_texture_lod
GL_SGIX_depth_texture
GL_SGIX_shadow
GL_WIN_swap_hint
WGL_EXT_swap_control

SirKnight
01-23-2002, 04:35 PM
That new vertex program extenstion sounds pretty interesting to me. That and the GL_NV_occlusion_query extension sounds cool. Hmmm, i smell a geforce 4. http://www.opengl.org/discussion_boards/ubb/wink.gif

-SirKnight

Lev
01-23-2002, 07:52 PM
Nutty: maybe you'll get even a bigger list by activating NV25Emulate??

-Lev

PH
01-24-2002, 03:12 AM
Originally posted by Lev:
Nutty: maybe you'll get even a bigger list by activating NV25Emulate??


I tried that but the new drivers don't show anything new. Some time ago I saw NV_fragment_program in one of the drivers ( and NV_texture_shaders3 ), so I know what I'm looking forward to ( very soon I hope ).

Lev
01-24-2002, 03:40 AM
Anyone(NVIDIA) knows if EXT_multi_draw_arrays works with VAR?

-Lev

Lev
01-24-2002, 09:04 AM
I just looked through the nvoglnt.dll 27.10 and found NV_point_sprite and NV_texture_shader3 in addition to what Nutty listed, nothing with NV_fragment_... though

-Lev

PH
01-24-2002, 09:25 AM
You can find NV_fragment_program in drivers 21.81, 21.85 and 21.88 ( possibly others ). Introducing NV_point_sprite seems a bit strange, I thought ATI was working on such an extension ( I hope they can agree on this one extension ).

davepermen
01-24-2002, 10:29 AM
hm yeah.. nvidia and ati:
how about developing some extensions TOGETHER?! would be much more simple for everyone..

i dont think there can be much difference for ati pointsprites and nvidia pointsprites so why 2 exts?

hm fragment-program.. sounds like real textureshaders for me (lookups not prebuild in hardware but simple after doing some userdefined program (like ps1.4 + )

if this is gf4 then i'll go for one possibly (if its not too expensive..)

GPSnoopy
01-24-2002, 01:20 PM
I remember NVIDIA talking about the occlusion query when the GF3 came. (they said you would be able to query whether a triangle you specify would be occluded or not).
However I've never seen it coming. It's probably that extension. http://www.opengl.org/discussion_boards/ubb/smile.gif


Lots of new exciting extensions! But let OpenGL 2.0 clean this mess!

[This message has been edited by GPSnoopy (edited 01-24-2002).]

Humus
01-26-2002, 04:53 AM
Originally posted by davepermen:
hm yeah.. nvidia and ati:
how about developing some extensions TOGETHER?! would be much more simple for everyone..


Yeah, nobody has something to win on fragmenting OpenGL. GL_NV_vertex_program1_1 comes as a dissapointment for me, I was hoping nVidia would join up with ATi and Matrox on GL_EXT_vertex_shader, not going further in their own direction. This means I'm not sure I'll be able to support nVidia cards in my demo/game/whatever (http://hem.passagen.se/emiper/temp/specular.html) I'm working on. http://www.opengl.org/discussion_boards/ubb/frown.gif

OldMan
01-26-2002, 06:58 AM
That will only happen in OGL 2.0... until there they are more interested in tring to make other companies suffer.

Carmacksutra
01-26-2002, 07:12 AM
Have you seen this? http://www.pcrave.com/news/news.htm

among others:


GL_NV_draw_mesh
void glDrawMeshNV (GLenum mode, GLsizei count, GLenum type, GLsizei stride, const GLvoid *indicesTexCoord, const GLvoid *indicesNormal, const GLvoid *indicesVertex);


Did they implement multi-index-array?
(despite Matt's *strong-scepticism* he expressed few threads ago http://www.opengl.org/discussion_boards/ubb/wink.gif)

If it is, then it is quite poor in comparison to gl2.0's solution, i think.