xref: /third_party/openGLES/extensions/ARB/ARB_vertex_shader.txt

Name

    ARB_vertex_shader

Name Strings

    GL_ARB_vertex_shader

Contributors

    Kurt Akeley
    Allen Akin
    Dave Baldwin
    Pat Brown
    Matt Craighead
    Cass Everitt
    Evan Hart
    Phil Huxley
    Dale Kirkland
    John Kessenich
    Steve Koren
    Jon Leech
    Bill Licea-Kane
    Benjamin Lipchak
    Barthold Lichtenbelt
    Kent Lin
    Jeremy Morris
    Teri Morrison
    Glenn Ortner
    Randi Rost
    Jeremy Sandmel
    Folker Schamel
    Eskil Steenberg

    The ARB_vertex_program working group members. Several concepts and
    chunks of text are copied from the ARB_vertex_program specification.

Contact

    Barthold Lichtenbelt, 3Dlabs, Inc. (barthold 'at' 3dlabs.com)
    Randi Rost, 3Dlabs, Inc. (rost 'at' 3dlabs.com)

Notice

    Copyright (c) 2003-2013 The Khronos Group Inc. Copyright terms at
        http://www.khronos.org/registry/speccopyright.html

Specification Update Policy

    Khronos-approved extension specifications are updated in response to
    issues and bugs prioritized by the Khronos OpenGL Working Group. For
    extensions which have been promoted to a core Specification, fixes will
    first appear in the latest version of that core Specification, and will
    eventually be backported to the extension document. This policy is
    described in more detail at
        https://www.khronos.org/registry/OpenGL/docs/update_policy.php

IP Status

    As described in the Contributor License, which can be found at
    http://developer.3dlabs.com/openGL2/specs/3dlabs_contributor.pdf .

Status

    Complete. Approved by the ARB on June 11, 2003.
    Updated revision 0.81 approved by the ARB on June 17, 2004.

Version

    Last Modified Date: December 14, 2009
    Author Revision: 0.83

Number

    ARB Extension #31

Dependencies

    OpenGL 1.0 is required.

    This extension is written against version 1.10 of the OpenGL Shading
    Language Specification.

    The extension is written against the OpenGL 1.4 Specification.

    The ARB_shader_objects extension is required.

Overview

    This extension adds programmable vertex level processing to OpenGL. The
    application can write vertex shaders in a high level language as defined
    in the OpenGL Shading Language specification. The language itself is not
    discussed here. A vertex shader replaces the transformation, texture
    coordinate generation and lighting parts of OpenGL, and it also adds
    texture access at the vertex level. Furthermore, management of vertex
    shader objects and loading generic attributes are discussed. A vertex
    shader object, attached to a program object, can be compiled and linked
    to produce an executable that runs on the vertex processor in OpenGL.
    This extension also defines how such an executable interacts with the
    fixed functionality vertex processing of OpenGL 1.4.

Issues

    1) What does need to be said about invariance (Appendix A of the OpenGL
       1.4 spec)?

    DISCUSSION: Numerous rules and proposals have been discussed. In the
    end, simplicity is an important goal. It is always possible to later add
    more invariance rules, if it turns out to be desirable.

    RESOLUTION: The same shader will produce the same result when run
    multiple times with the same input. The wording 'the same shader' means
    a program object that is populated with the same source strings, which
    are compiled and then linked, possibly multiple times. This program
    object is executed using the same GL state vector.

    Position invariance between a vertex shader and the fixed functionality
    pipeline is important for certain applications. Therefore there must be
    a way to guarantee that a vertex shader produces the exact same position
    for a vertex as the fixed function pipeline does. This is achieved by
    using the built-in function ftransform in a vertex shader, and assign it
    to the output variable gl_Position. The shader does this instead of
    computing the product of the ModelViewProjection matrix with the vertex
    position. See section 2.15.6 for more.

    2) Should we take the API commands to load generic vertex attributes
       from ARB_vertex_program?

    DISCUSSION: At first glance these commands, for example
    VertexAttrib3fvARB, look like exactly what this extension needs.
    However, if we re-use these commands here, how does the GL
    implementation decide where attributes go when using shaders written
    using ARB_vertex_program or this extension? What happens to the
    attribute value when switching between an ARB_vertex_program and an
    ARB_vertex_shader shader? How does this resolve with respect to
    undefined aliasing defined in ARB_vertex_program and the no aliasing
    defined in this extension?

    RESOLUTION: Resolved, we will re-use these entry points. It is very
    convenient for an ISV to be able to re-use the entry points to load
    data, and only have to switch shaders when it wants to go from an
    ARB_vertex_program shader to an ARB_vertex_shader shader. This extension
    allows for no aliasing among generic attributes and standard vertex
    attributes, whereas ARB_vertex_program has the concept of undefined
    aliasing. In order to resolve this the interactions section extends
    ARB_vertex_program to expose this no aliasing approach. Note that we do
    not lift the restriction in the grammar on invalid attribute binding
    pairs, see Table X.2.1 in the ARB_vertex_program specification.

    3) Should we take the API commands to load generic vertex arrays from
       ARB_vertex_program?

    DISCUSSION: See issue 2.

    RESOLUTION: Resolved, yes we will re-use these entry points. See also
    the previous issue resolution.

    4) Should we take the API commands to query the values of generic
       attributes from ARB_vertex_program?

    DISCUSSION: This goes hand in hand with issue 2.

    RESOLUTION: Yes we will.

    5) Do we need to specify how Evaluators and AUTO_NORMAL generation
       interacts?

    DISCUSSION: ARB_vertex_program specifies that for evaluated normals the
    normal is no longer automatically normalized when a vertex shader is
    active. This is discussed in detail in issue 92 in the
    ARB_vertex_program specification.

    RESOLUTION: We will follow ARB_vertex_program.

    6) What should the initial value of the generic vertex attributes be?

    DISCUSSION: ARB_vertex_program leaves their initial values undefined. It
    seems to make sense to do so, because of the undefined aliasing. In this
    extension there is no aliasing among conventional and generic
    attributes. Hence it could be desirable to define the initial values to
    be (0, 0, 0, 1).

    RESOLUTION: The initial values are (0, 0, 0, 1).

    7) Do generic attributes push and pop?

    DISCUSSION: There has been some concern that adding more state than
    strictly necessary is a performance issue for applications that push and
    pop a lot. Conventional attributes are pushed and popped, as well as
    generic attributes, by ARB_vertex_program.

    RESOLUTION: YES, they are part of GL state, and such should push and pop
    under the 'current' bit. Note that an OpenGL implementation could track
    if a generic attribute has been set, and as long as it has not, never
    push/pop that attribute.

    8) How does user clipping work?

    DISCUSSION: The OpenGL Shading Language provides a gl_ClipVertex
    built-in variable in the vertex shading language. This variable provides
    a place for vertex shaders to specify a coordinate to be used by the
    user clipping plane stage. The user must ensure that the clip vertex and
    user clipping planes are defined in the same coordinate space. Note that
    this is different than ARB_vertex_program, where user clipping is
    ignored unless the position invariant option is enabled (where all
    vertex transformation options are performed by the fixed functionality
    pipeline). Here are some tips on how to use user clipping in a vertex
    shader:

      1) When using a traditional transform in a vertex shader, compute the
         eye coordinates and store the result in gl_ClipVertex.

      2) If clip planes are enabled with a vertex shader, gl_ClipVertex must
         be written to, otherwise results will be undefined.

      3) When doing object-space clipping, keep in mind that the clip planes
         are automatically transformed to eye coordinates (see section 2.11
         of the GL 1.4 spec). Use an identity modelView matrix to avoid this
         transformation.

    RESOLUTION: Resolved. See also section 2.11.

    9) Do we need a vertex shader point size enable?

    DISCUSSION: ARB_vertex_program provides an enable,
    VERTEX_PROGRAM_POINT_SIZE_ARB. This extension uses the same enable. If
    VERTEX_PROGRAM_POINT_SIZE mode is enabled, and a vertex shader is
    active, then the point size is taken from the vertex shader's point size
    result, otherwise it is taken from the PointSize command. The OpenGL
    Shading Language provides a built-in gl_PointSize that a vertex shader
    can write to, to specify the point size in pixels. If gl_PointSize is
    not written while the point size mode is enabled then the point size
    used in the rasterization stage is undefined.

    This point size enable is convenient for the majority of applications
    who do not change the point size. Not setting the enable means the point
    size is picked up automatically, instead of having to write a shader
    which always outputs the same point size.

    RESOLUTION: Yes.

    10) How to handle the color sum stage?

    DISCUSSION: It is the vertex shader's responsibility to write a primary
    and secondary color into the appropriate built in varying variables, if
    subsequent stages in the pipeline use these. Otherwise results are
    undefined. Section 3.9 of the OpenGL 1.4 spec states that the color sum
    stage is always applied when lighting is enabled, ignoring the value of
    the COLOR_SUM enable bit. Should we take out this automatic enable when
    lighting is enabled?

    RESOLUTION: This automatic enable will only have effect when no vertex
    shader is active. This behavior is the same in ARB_vertex_program.

    11) Does generic attribute zero also provoke (signal the end of a
        vertex), or just the glVertex command?

    DISCUSSION: To be more compatible with ARB_vertex_program this would be
    desirable. It may be that no attribute directly expresses a vertex
    position. The shader may derive the position based on multiple
    attributes, where there is no clear attribute that can be marked as "the
    vertex". On the other hand, it is mandatory in a vertex shader to write
    to gl_Vertex. Thus it makes some sense that the glVertex API has to be
    used to signal the end of a vertex.

    RESOLUTION: Resolved. For compatibility reasons, attribute zero will
    also provoke.

    12) How to handle texture access from a vertex shader?

    DISCUSSION: The OpenGL Shading Language allows for texture access from a
    vertex shader. What is the OpenGL support necessary to make that
    actually happen? Two approaches seem feasible. 1) Virtualize the number
    of texture image units available in OpenGL. Deprecate the concept of
    binding a texture object to a texture image unit. Instead, bind a
    texture object directly to a sampler. 2) A not so drastic approach that
    makes a maximum number of texture image units available to a vertex
    shader as well as to the fragment processing part of the OpenGL
    pipeline. The sum of the available texture image units cannot exceed the
    constant MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB.

    RESOLUTION: Option 2) is what is used in this specification. Option 1)
    is a more invasive approach, and left for a future extension.

    This approach allows for two types of hardware. One where the hardware
    resources to perform the functions associated with a texture image unit
    are not shared between a vertex and fragment shader, and one type of
    hardware where they are shared. The maximum number of texture image
    units available to a vertex shader is
    MAX_VERTEX_TEXTURE_IMAGE_UNITS_ARB. The maximum number of texture image
    units available to the fragment stage of OpenGL is
    MAX_TEXTURE_IMAGE_UNITS_ARB. Now consider, for example, an
    implementation that supports up to 16 texture stages and shares its
    texture mapping resources between the vertex shader and fragment shader.
    For this implementation MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB,
    MAX_VERTEX_TEXTURE_IMAGE_UNITS_ARB and MAX_TEXTURE_IMAGE_UNITS_ARB would
    all be set to 16. A hardware implementation which does not share its
    texture mapping resources, and who can support 2 texture stages for a
    vertex shader and 16 texture stages for the fragment processing, would
    set MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB to 18,
    MAX_VERTEX_TEXTURE_IMAGE_UNITS_ARB to 2 and MAX_TEXTURE_IMAGE_UNITS_ARB
    to 16. A vertex shader cannot use more than
    MAX_VERTEX_TEXTURE_IMAGE_UNITS_ARB texture image units. The fragment
    processing part of OpenGL cannot use more than
    MAX_TEXTURE_IMAGE_UNITS_ARB texture image units.

    Both the vertex shader and fragment processing combined cannot use more
    than MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB texture image units. If a
    vertex shader and fragment processing part of OpenGL both use the same
    texture image unit, then that counts as separate units against the above
    mentioned limits.

    Automatic LOD calculation does not happen during vertex processing in a
    vertex shader, hence there is no automatic selection of an image array
    level. Minification or magnification of a texture map is controlled by a
    LOD value passed as an argument in some of the built-in texture lookup
    functions. These built-in texture lookup functions, defined in the
    OpenGL Shading Language, come in two flavors. One set of lookup
    functions that take a LOD value as an input, and one set that does not
    take a LOD value.

    If the former set of functions is used, then the LOD value passed in is
    used directly as the LOD value lambda' in section 3.8.8 equation 3.16 on
    page 140 of the OpenGL 1.4 spec. This in turn defines lambda, which then
    is used as usual to determine magnification or minification and, when
    mipmapping, which mipmap array(s) to select.

    If the latter set of functions is used, then lambda, the result of
    equation 3.16, is set to zero. This effectively means a texture object
    is considered magnified, and only paragraph 3.8.9 applies.

    13) Do we split a texture unit into a texture image unit and a texture
        coordinate set?

    DISCUSSION: This is what ARB_fragment_program does, and
    ARB_fragment_shader as well. The OpenGL Shading Language allows for
    texture lookups in a vertex shader. This means that a vertex shader
    should have access to a texture image unit (texture maps and associated
    parameters). A vertex shader also has access to state stored with a
    texture coordinate set, which encompasses vertex texture coordinate
    attributes, as well as the texture matrix stack and texture generation
    state. The number of supported texture image units and texture
    coordinate sets is not necessarily the same on a particular
    implementation.

    RESOLUTION: Yes

    14) Do we need to specify different texture coordinate sets for a vertex
        shader and fragment processing?

    DISCUSSION: So far there are MAX_TEXTURE_COORDS_ARB texture coordinate
    sets. This constant is defined in ARB_fragment_program, and re-used in
    ARB_fragment_shader as well as the OpenGL Shading Language (see built-in
    constants, section 7.4). Only a vertex shader has access to vertex
    texture coordinate attributes. But both an ARB vertex and fragment
    shader has access to the other state associated with a texture
    coordinate set. That state is texture generation and the texture matrix
    stack.

    RESOLUTION: NO.

    15) What is the minimum value for MAX_VERTEX_TEXTURE_IMAGE_UNITS_ARB?

    DISCUSSION: Should this be zero, or more?

    RESOLUTION: Resolved, zero.

    16) Do we also need a built in function to guarantee position invariance
        with respect to user clip planes?

    DISCUSSION: Issue 1 discusses invariance with respect to the vertex
    position. A vertex shader however can also output coordinates used for
    subsequent user clipping, by writing to gl_ClipVertex. In other words,
    do we need something like gl_ClipVertex = ftransformClip()?

    RESOLUTION: NO, position invariance with respect to user clipping seems
    like a rarely needed feature. The OpenGL Shading Language does not have
    a ftransformClip() built-in function. If this functionality is ever
    needed, it is easy to add later.

    17) What happens to the point size when a vertex shader is active?

    DISCUSSION: The point size is taken from the built-in gl_PointSize when
    VERTEX_PROGRAM_POINT_SIZE_ARB is enabled, but what to do with distance
    attenuation and point fade (see Section 3.3)?

    RESOLUTION: Resolved. The vertex shader should compute distance
    attenuation. Fixed function computes it as a function of, among others,
    eyeZ. It is not required for a vertex shader to compute eyeZ, and the
    fixed function eyeZ might have nothing to do with the vertex position
    computed in the vertex shader. Point fade, however, should still be
    fixed function. The vertex shader could compute a point fade, but it'll
    produce incorrect results if the application issues the following
    commands:

      glPolygonMode(GL_FRONT, GL_FILL);
      glPolygonMode(GL_BACK, GL_POINT);

    In this case you would want to fade if a vertex belonged to a
    back-facing triangle, but not if it could belong to a front-facing one.

    18) What is the deal with range and precision of colors once output from
        the vertex shader?

    DISCUSSION: Section 2.13.6 of the GL 1.4 spec specifies that all
    components of both the primary and secondary colors be clamped to the
    range [0,1]. Section 2.13.9 furthermore specifies that each color
    component be converted to a fixed-point value with enough bits to
    satisfy the frame buffer depth. Section 2.1.1 says that the maximum
    representable magnitude of a floating-point number for colors must be at
    least 2^10.

    It is desirable to bypass clamping of colors when an ARB_fragment_shader
    shader is also active. It has utility to be able to exceed the range
    [0,1] and deal with extended range colors in a fragment shader.

    Varying variables, including the built-in color variables, are declared
    as floating point vectors in the OpenGL Shading Language, and
    interpolated as floating-point values. It does not seem to make much
    sense to specify a conversion to fixed-point after the ARB vertex shader
    writes colors, then interpolate those fixed-point colors, then convert
    the interpolated values back to floating point for input in the ARB
    fragment shader.

    ARB_fragment_program leaves the primary and secondary colors as the only
    varying variables that have less range (2^10) than any other varying. It
    promoted texture coordinates to full citizens with a maximum
    representable magnitude of at least 2^32.

    There are three cases to consider:

      1) An ARB_vertex_shader shader writing colors that are consumed by an
         ARB_fragment_shader shader.
      2) An ARB_vertex_shader shader writing colors that are consumed by an
         ARB_fragment_program shader.
      3) An ARB_vertex_shader shader writing colors that are consumed by
         fixed-function fragment processing.

    An argument can be made that a shader writer can bypass the clamping and
    conversion to fixed-point by defining its own varying variables, and use
    that to interpolate colors. It can then use the built-in gl_FrontFacing
    in a fragment shader for front or back face color selection. This is
    certainly true, but it bypasses the (fixed-function) front and back face
    color selection, which happens in a stage after the vertex shader. This
    front/back face selection stage can only be used when writing to the
    vertex shader built-in gl_FrontColor, gl_BackColor,
    gl_FrontSecondaryColor and gl_BackSecondaryColor.

    RESOLUTION: We anticipate future extensions will generalize the final
    color-processing clamp with an enable/disable clamp. Therefore, this
    specification will leave the final color clamp and conversion to
    fixed-point in place for gl_FrontColor, gl_FrontSecondaryColor,
    gl_BackColor, and gl_BackSecondaryColor.

    19) How do I get a list of all attributes and their types used by a
        shader and when do I bind generic vertex attributes?

    DISCUSSION: GetActiveAttribARB will return the name of an attribute and
    its type declared and actually used in the shader, as determined by the
    compiler and linker. Looping over all attributes and calling
    GetActiveAttribARB will generate the list of attribute names and types
    used in a shader. This list is called the active attributes list.
    GetActiveAttribARB can be called after a link attempt. It is not
    necessary for the link to have succeeded. After all, one of the reasons
    the link failed could be that there are too many active attributes. It
    is not required to call BindAttribLocationARB before issuing
    GetActiveAttribARB. BindAttributeLocationARB can be called any time. In
    order to have a BindAttribLocationARB command take effect, the program
    object will have to be (re)linked. Thus the order of commands to get the
    list of active attributes could be: Link the program object, call
    GetActiveAttribARB as many times as needed, call BindAttribLocationARB
    one or more times, then call link again.

    It is allowed to issue BindAttribLocationARB before any vertex shader
    objects are attached to a program object. Hence it is allowed to bind
    any name to a location, including a name that is never used as an
    attribute in any vertex shader object.

    RESOLUTION: Resolved, call GetActiveAttribARB.

    20) Who assigns vertex attribute variable names to indices used in the
        VertexAttrib*ARB commands?

    DISCUSSION: Both the GL and the application can do this. The linker will
    assign indices to vertex attributes declared in a vertex shader. These
    values can be queried by the application and used to load values for the
    generic vertex attributes. Alternatively the application can override
    this assignment, and bind attribute variable names to any generic vertex
    attribute index of their choice, within the range 0 to
    MAX_VERTEX_ATTRIBS_ARB. An application would have to do this when
    building a display list, for example. In this case it is possible to
    generate the display list before the program object has even been
    created.

    Only indices for generic attributes can be queried, or set. Conventional
    attributes are handled by the GL, and guaranteed to not alias with
    generic attributes.

    RESOLUTION: Both the GL and the application.

    21) How should this extension handle changes to material state within
        Begin/End?

    DISCUSSION: Supporting per-vertex material properties places additional
    pressure on the number of per-vertex bindings an OpenGL implementation
    can support, which already is a scarce resource.

    RESOLUTION: Materials are declared as uniforms in the OpenGL Shading
    Language. Any changes to the material properties inside a Begin/End
    through Material commands, or indirectly through ColorMaterial and Color
    commands, are not guaranteed to actually update material state until the
    next End command. At that point, all material property state is
    guaranteed to be updated.

    In practice, material properties are usually not changed within
    Begin/End. Applications needing to change material properties inside a
    Begin/End in a vertex shader can work around this limitation by storing
    the color in a conventional or generic vertex attribute and modifying
    the vertex shader accordingly.

    22) What space are gl_Position and gl_ClipVertex in?

    DISCUSSION: The GL treats the values in the built-in output variable
    gl_Position as being in clip space, for the purpose of clipping against
    the clip volume. See Section 2.11 of the OpenGL 1.4 specification. The
    GL treats the values in the (optional) built-in output variable
    gl_ClipVertex as being in eye-space for clipping against a
    client-defined clip plane.

    RESOLUTION: clip-space and eye-space respectively.

    23) What texture operations are not affected by a vertex shader
    performing a texture lookup?

    RESOLUTION: Whether or not a vertex shader is active, the following
    operations still behave as specified:

      * texture image specification (pp. 119-128)
      * alternate texture image specification (pp. 128-132)
      * compressed texture image specification (pp. 132-135)
      * texture parameters behave as specified even when a texture is
        accessed from within a vertex shader (pp. 135-147)
      * texture state and proxy state (pp. 148-149)
      * texture object specification (pp. 149-152)
      * texture comparison modes (p. 157)

New Procedures and Functions

    void VertexAttrib1fARB(uint index, float v0)
    void VertexAttrib1sARB(uint index, short v0)
    void VertexAttrib1dARB(uint index, double v0)
    void VertexAttrib2fARB(uint index, float v0, float v1)
    void VertexAttrib2sARB(uint index, short v0, short v1)
    void VertexAttrib2dARB(uint index, double v0, double v1)
    void VertexAttrib3fARB(uint index, float v0, float v1, float v2)
    void VertexAttrib3sARB(uint index, short v0, short v1, short v2)
    void VertexAttrib3dARB(uint index, double v0, double v1, double v2)
    void VertexAttrib4fARB(uint index, float v0, float v1, float v2, float v3)
    void VertexAttrib4sARB(uint index, short v0, short v1, short v2, short v3)
    void VertexAttrib4dARB(uint index, double v0, double v1, double v2, double v3)
    void VertexAttrib4NubARB(uint index, ubyte x, ubyte y, ubyte z, ubyte w)

    void VertexAttrib1fvARB(uint index, const float *v)
    void VertexAttrib1svARB(uint index, const short *v)
    void VertexAttrib1dvARB(uint index, const double *v)
    void VertexAttrib2fvARB(uint index, const float *v)
    void VertexAttrib2svARB(uint index, const short *v)
    void VertexAttrib2dvARB(uint index, const double *v)
    void VertexAttrib3fvARB(uint index, const float *v)
    void VertexAttrib3svARB(uint index, const short *v)
    void VertexAttrib3dvARB(uint index, const double *v)
    void VertexAttrib4fvARB(uint index, const float *v)
    void VertexAttrib4svARB(uint index, const short *v)
    void VertexAttrib4dvARB(uint index, const double *v)
    void VertexAttrib4ivARB(uint index, const int *v)
    void VertexAttrib4bvARB(uint index, const byte *v)

    void VertexAttrib4ubvARB(uint index, const ubyte *v)
    void VertexAttrib4usvARB(uint index, const ushort *v)
    void VertexAttrib4uivARB(uint index, const uint *v)

    void VertexAttrib4NbvARB(uint index, const byte *v)
    void VertexAttrib4NsvARB(uint index, const short *v)
    void VertexAttrib4NivARB(uint index, const int *v)
    void VertexAttrib4NubvARB(uint index, const ubyte *v)
    void VertexAttrib4NusvARB(uint index, const ushort *v)
    void VertexAttrib4NuivARB(uint index, const uint *v)

    void VertexAttribPointerARB(uint index, int size, enum type, boolean normalized,
                                sizei stride, const void *pointer)

    void EnableVertexAttribArrayARB(uint index)
    void DisableVertexAttribArrayARB(uint index)

    void BindAttribLocationARB(handleARB programObj, uint index, const charARB *name)

    void GetActiveAttribARB(handleARB programObj, uint index, sizei maxLength,
                             sizei *length, int *size, enum *type, charARB *name)

    GLint GetAttribLocationARB(handleARB programObj, const charARB *name)

    void GetVertexAttribdvARB(uint index, enum pname, double *params)
    void GetVertexAttribfvARB(uint index, enum pname, float *params)
    void GetVertexAttribivARB(uint index, enum pname, int *params)
    void GetVertexAttribPointervARB(uint index, enum pname, void **pointer)

New Tokens

    Accepted by the <shaderType> argument of CreateShaderObjectARB and
    returned by the <params> parameter of GetObjectParameter{if}vARB:

        VERTEX_SHADER_ARB                               0x8B31

    Accepted by the <pname> parameter of GetBooleanv, GetIntegerv,
    GetFloatv, and GetDoublev:

        MAX_VERTEX_UNIFORM_COMPONENTS_ARB               0x8B4A
        MAX_VARYING_FLOATS_ARB                          0x8B4B
        MAX_VERTEX_ATTRIBS_ARB                          0x8869
        MAX_TEXTURE_IMAGE_UNITS_ARB                     0x8872
        MAX_VERTEX_TEXTURE_IMAGE_UNITS_ARB              0x8B4C
        MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB            0x8B4D
        MAX_TEXTURE_COORDS_ARB                          0x8871

    Accepted by the <cap> parameter of Disable, Enable, and IsEnabled, and
    by the <pname> parameter of GetBooleanv, GetIntegerv, GetFloatv, and
    GetDoublev:

        VERTEX_PROGRAM_POINT_SIZE_ARB                   0x8642
        VERTEX_PROGRAM_TWO_SIDE_ARB                     0x8643

    Accepted by the <pname> parameter GetObjectParameter{if}vARB:

        OBJECT_ACTIVE_ATTRIBUTES_ARB                    0x8B89
        OBJECT_ACTIVE_ATTRIBUTE_MAX_LENGTH_ARB          0x8B8A

    Accepted by the <pname> parameter of GetVertexAttrib{dfi}vARB:

        VERTEX_ATTRIB_ARRAY_ENABLED_ARB                 0x8622
        VERTEX_ATTRIB_ARRAY_SIZE_ARB                    0x8623
        VERTEX_ATTRIB_ARRAY_STRIDE_ARB                  0x8624
        VERTEX_ATTRIB_ARRAY_TYPE_ARB                    0x8625
        VERTEX_ATTRIB_ARRAY_NORMALIZED_ARB              0x886A
        CURRENT_VERTEX_ATTRIB_ARB                       0x8626

    Accepted by the <pname> parameter of GetVertexAttribPointervARB:

        VERTEX_ATTRIB_ARRAY_POINTER_ARB                 0x8645

    Returned by the <type> parameter of GetActiveAttribARB:

        FLOAT                                           0x1406
        FLOAT_VEC2_ARB                                  0x8B50
        FLOAT_VEC3_ARB                                  0x8B51
        FLOAT_VEC4_ARB                                  0x8B52
        FLOAT_MAT2_ARB                                  0x8B5A
        FLOAT_MAT3_ARB                                  0x8B5B
        FLOAT_MAT4_ARB                                  0x8B5C

Additions to Chapter 2 of the OpenGL 1.4 Specification (OpenGL Operation)

    Modify Section 2.1.1, Floating-Point Computation (p. 6)

    (modify first paragraph, p. 6) ... are accurate to about 1 part in 10^5.
    The maximum representable magnitude of a floating-point number used to
    represent positional, normal or texture coordinates must be at least
    2^32. The maximum representable magnitude for colors must be at least
    2^10. The maximum representable magnitude for all other floating-point
    values must be at least 2^32.

    Modify Section 2.6, Begin/End Paradigm (p. 12)

    (modify last paragraph, p. 12) Each vertex is specified with two, three,
    or four coordinates. In addition, a current normal, multiple current
    texture coordinate sets, multiple current generic vertex attributes,
    current color ... Generic vertex attributes can be accessed from within
    vertex shaders (section 2.15) and used to compute values for consumption
    by later processing stages.

    Modify Section 2.6.3, GL Commands within Begin/End (p. 19)

    (modify first paragraph of section, p. 19) The only GL commands that are
    allowed within any Begin/End pairs are the commands for specifying
    vertex coordinates, vertex colors, normal coordinates, texture
    coordinates, generic vertex attributes and fog coordinates (Vertex,
    Color, SecondaryColor, Index, Normal, TexCoord and MultiTexCoord,
    VertexAttrib*ARB, FogCoord), ...

    Modify Section 2.7, Vertex Specification (p. 19)

    (modify second paragraph, p. 20) Implementations support more than one
    set of texture coordinates. The commands

        void MultiTexCoord{1234}{sifd}(enum texture, T coords)
        void MultiTexCoord{1234}{sifd}v(enum texture, T coords)

    take the coordinate set to be modified as the <texture> parameter.
    <texture> is a symbolic constant of the form TEXTUREi, indicating that
    texture coordinate set i is to be modified. The constants obey TEXTUREi
    = TEXTURE0 + i (i is in the range 0 to k-1, where k is the
    implementation-dependent number of texture coordinate sets defined by
    MAX_TEXTURE_COORDS_ARB).

    (remove the "Finally" from the second paragraph on p. 21)

    (add the following prior to the last paragraph in the section, p. 21)

    Vertex shaders (see Section 2.15) can be written to access an array of
    4-component generic vertex attributes in addition to the conventional
    attributes specified previously. The first slot of this array is
    numbered 0, and the length of the array is specified by the
    implementation-dependent constant MAX_VERTEX_ATTRIBS_ARB.

    The commands

        void VertexAttrib{1234}{sfd}ARB(uint index, T values)
        void VertexAttrib{123}{sfd}vARB(uint index, T values)
        void VertexAttrib4{bsifd ubusui}vARB(uint index, T values)

    can be used to load the given value(s) into the generic attribute at
    slot <index>, whose components are named <x>, <y>, <z>, and <w>. The
    VertexAttrib1*ARB family of commands sets the <x> coordinate to the
    provided single argument while setting <y> and <z> to 0 and <w> to 1.
    Similarly, VertexAttrib2*ARB commands set <x> and <y> to the specified
    values, <z> to 0 and <w> to 1; VertexAttrib3*ARB commands set <x>, <y>,
    and <z>, with <w> set to 1, and VertexAttrib4*ARB commands set all four
    coordinates. The error INVALID_VALUE is generated if <index> is greater
    than or equal to MAX_VERTEX_ATTRIBS_ARB.

    The commands

        void VertexAttrib4NubARB(uint index, T values)
        void VertexAttrib4N{bsi ubusui}vARB(uint index, T values)

    also specify vertex attributes with fixed-point coordinates that are
    scaled to a normalized range, according to Table 2.6.

    The VertexAttrib*ARB entry points defined earlier can also be used to
    load attributes declared as a 2x2, 3x3 or 4x4 matrix in a vertex shader.
    Each column of a matrix takes up one generic 4-component attribute slot
    out of the MAX_VERTEX_ATTRIBS_ARB available slots. Matrices are loaded
    into these slots in column major order. Matrix columns need to be loaded
    in increasing slot numbers.

    Setting generic vertex attribute zero specifies a vertex; the four
    vertex coordinates are taken from the values of attribute zero. A
    Vertex2, Vertex3, or Vertex4 command is completely equivalent to the
    corresponding VertexAttrib*ARB command with an index of zero. Setting
    any other generic vertex attribute updates the current values of the
    attribute. There are no current values for vertex attribute zero.

    There is no aliasing among generic attributes and conventional
    attributes. In other words, an application can set all
    MAX_VERTEX_ATTRIBS_ARB generic attributes and all conventional
    attributes without fear of one particular attribute overwriting the
    value of another attribute.

    (Change the first half of the last paragraph in the section, p. 21, to)

    The state required to support vertex specification consists of four
    floating-point numbers per texture coordinate set to store the current
    texture coordinates s, t, r, and q, three floating-point numbers to
    store the three coordinates of the current normal, one floating-point
    number to store the current fog coordinate, four floating- point values
    to store the current RGBA color, four floating-point values to store the
    current RGBA secondary color, one floating-point value to store the
    current color index, and MAX_VERTEX_ATTRIBS_ARB-1 four-component
    floating-point vectors for generic vertex attributes. There is no notion
    of a current vertex, so no state is devoted to vertex coordinates or
    generic vertex attribute zero. The initial texture coordinates are
    (S,T,R,Q) = (0,0,0,1) for each texture coordinate set. The initial
    current normal has coordinates (0,0,1). The initial fog coordinate is
    zero. The initial RGBA color is (R,G,B,A) = (1,1,1,1) and the initial
    RGBA secondary color is (0, 0, 0, 1). The initial color index is 1. The
    initial values for all generic vertex attributes are (0, 0, 0, 1).

    Modify Section 2.8, Vertex Arrays (p. 23)

    (modify first paragraph of section, p.23) The vertex specification
    commands described in section 2.7 accept data in almost any format, but
    their use requires many command executions to specify even simple
    geometry. Vertex data may also be placed into arrays that are stored in
    the client's address space. Blocks of data in these arrays may then be
    used to specify multiple geometric primitives through the execution of a
    single GL command. The client may specify up to 7 plus the values of
    MAX_TEXTURE_COORDS_ARB and MAX_VERTEX_ATTRIBS_ARB arrays: One each to
    store vertex coordinates, edge flags, colors, secondary colors, color
    indices, normals, fog coordinates, one or more texture coordinate sets,
    and one or more generic vertex attributes.

    The commands

        ...

        void VertexAttribPointerARB(uint index, int size, enum type,
                                    boolean normalized, sizei stride,
                                    const void *pointer)

    describe the locations and organizations...

    (add after the first paragraph, p.24) The <index> parameter in the
    VertexAttribPointerARB command identifies the generic vertex attribute
    array being described. The error INVALID_VALUE is generated if <index>
    is greater than or equal to MAX_VERTEX_ATTRIBS_ARB. The <normalized>
    parameter in the VertexAttribPointerARB command identifies whether
    fixed-point types should be normalized when converted to floating-point.
    If <normalized> is TRUE, fixed-point data are converted as specified in
    Table 2.6; otherwise, the fixed-point values are converted directly.

    (add to the beginning of the page, p. 25) An individual generic vertex
    attribute array is enabled or disabled by calling one of

        void EnableVertexAttribArrayARB(uint index)
        void DisableVertexAttribArrayARB(uint index)

    where <index> identifies the generic vertex attribute array to enable or
    disable. The error INVALID_VALUE is generated if <index> is greater than
    or equal to MAX_VERTEX_ATTRIBS_ARB.

    (modify Table 2.4, p.24)

                                          Normal
         Command                 Sizes    ized?   Types
         ----------------------  -------  ------  ---------------------------
         VertexPointer           2,3,4     no     short, int, float, double
         NormalPointer           3         yes    byte, short, int, float,
                                                  double
         ColorPointer            3,4       yes    byte, ubyte, short, ushort,
                                                  int, uint, float, double
         SecondaryColorPointer   3         yes    byte, ubyte, short, ushort,
                                                  int, uint, float, double
         IndexPointer            1         no     ubyte, short, int, float,
                                                  double
         FogCoordPointer         1         n/a    float, double
         TexCoordPointer         1,2,3,4   no     short, int, float, double
         EdgeFlagPointer         1         no     boolean
         VertexAttribPointerARB  1,2,3,4   flag   byte, ubyte, short, ushort,
                                                  int, uint, float, double

    Table 2.4: Vertex array sizes (values per vertex) and data types. The
    "normalized" column indicates whether fixed-point types are accepted
    directly or normalized to [0,1] (for unsigned types) or [-1,1] (for
    signed types). For generic vertex attributes, fixed-point data are
    normalized if and only if the <normalized> flag is set.

    (modify first paragraph, p. 25) The command

        void ClientActiveTexture(enum texture)

    is used to select the vertex array client state parameters to be
    modified by the TexCoordPointer command and the array affected by
    EnableClientState and DisableClientState with parameter
    TEXTURE_COORD_ARRAY. This command sets the client state variable
    CLIENT_ACTIVE_TEXTURE. Each texture coordinate set has a client state
    vector which is selected when this command is invoked. This state vector
    includes the vertex array state. This call also selects the texture
    coordinate set state used for queries of client state.

    (modify third paragraph, p. 25) The command

        void ArrayElement(int i);

    transfers the ith element of every enabled array to the GL. The effect
    of ArrayElement(i) is the same as the effect of the command sequence

        if (normal array enabled) {
          Normal3[type]v(normal array element i);
        }
        if (color array enabled) {
          Color[size][type]v(color array element i);
        }
        if (secondary color array enabled) {
          SecondaryColor3[type]v(secondary color array element i);
        }
        if (fog coordinate array enabled) {
          FogCoord[type]v(fog coordinate array element i);
        }
        for (j = 0; j < textureUnits; j++) {
          if (texture coordinate set j array enabled) {
            MultiTexCoord[size][type]v(TEXTURE0 + j,
                                       texture coordinate set j
                                       array element i);
        }
        if (color index array enabled) {
          Index[type]v(color index array element i);
        }
        if (edge flag array enabled) {
          EdgeFlagv(edge flag array element i);
        }
        for (j = 1; j < genericAttributes; j++) {
          if (generic vertex attribute j array enabled) {
            if (generic vertex attribute j array normalization flag
                is set, and type is not FLOAT or DOUBLE) {
              VertexAttrib[size]N[type]vARB(j, generic vertex attribute j
                                               array element i);
            } else {
                VertexAttrib[size][type]vARB(j, generic vertex attribute j
                                              array element i);
            }
          }
        }
        if (generic attribute array 0 enabled) {
          if (generic vertex attribute 0 array normalization flag
              is set, and type is not FLOAT or DOUBLE) {
            VertexAttrib[size]N[type]vARB(0, generic vertex attribute 0
                                             array element i);
          } else {
            VertexAttrib[size][type]vARB(0, generic vertex attribute 0
                                            array element i);
          }
        } else if (vertex array enabled) {
            Vertex[size][type]vARB(vertex array element i);
        }

    where <textureUnits> and <genericAttributes> give the number of texture
    coordinate sets and generic vertex attributes supported by the
    implementation, respectively. "[size]" and "[type]" correspond to the
    size and type of the corresponding array. For generic vertex attributes,
    it is assumed that a complete set of vertex attribute commands exists,
    even though not all such functions are provided by the GL.

    (modify the first paragraph on p. 26) ...with one exception: The current
    edge flag, texture coordinates, color, secondary color, color index, fog
    coordinate, normal coordinates and generic attributes are each
    indeterminate after execution of DrawArrays, if the corresponding array
    is enabled.

    (modify the first paragraph on p. 27) ...with one exception: The current
    edge flag, texture coordinates, color, secondary color, color index, fog
    coordinate, normal coordinates and generic attributes are each
    indeterminate after execution of DrawElements, if the corresponding
    array is enabled.

    (modify last paragraph of section, p. 30) If the number of supported
    texture coordinate sets (the value of MAX_TEXTURE_COORDS_ARB) is m and
    the number of supported generic vertex attributes (the value of
    MAX_VERTEX_ATTRIBS_ARB) is n, then the client state required to
    implement vertex arrays consists of 7+m+n Boolean values, 7+m+n memory
    pointers, 7+m+n integer stride values, 7+m+n symbolic constants
    representing array types, 3+m+n integers representing values per element
    and n boolean values indicating normalization. In the initial state, the
    boolean values are each disabled, the memory pointers are each null, the
    strides are each zero, the array types are each FLOAT, and the integers
    representing values per element are each four.

    Modify Section 2.10, Coordinate Transformations (p. 31)

    (add at the beginning of the section) This section and the discussion
    through section 2.13 describe the state values and operations necessary
    for transforming vertex attributes according to a fixed-functionality
    method. An alternate method for transforming vertex attributes is
    described in section 2.15.

    Modify Section 2.10.2, Matrices (p. 33)

    (modify first paragraph, p. 37) For each texture coordinate set, a 4x4
    matrix is applied to the corresponding texture coordinates. This matrix
    is applied as...

    (modify third and fourth paragraphs, p. 37) The command

        void ActiveTexture(enum texture)

    specifies the active texture unit selector, ACTIVE_TEXTURE. Each texture
    unit contains up to two distinct sub-units: a texture coordinate
    processing unit (consisting of a texture matrix stack and texture
    coordinate generation state) and a texture image unit (consisting of all
    the texture state defined in Section 3.8). In implementations with a
    different number of supported texture coordinate sets and texture image
    units, some texture units may consist of only one of the two sub-units.

    The active texture unit selector specifies the texture coordinate set
    accessed by commands involving texture coordinate processing. Such
    commands include those accessing the current matrix stack (if
    MATRIX_MODE is TEXTURE), TexGen (section 2.10.4), Enable/Disable (if any
    texture coordinate generation enum is selected), as well as queries of
    the current texture coordinates and current raster texture coordinates.
    If the texture coordinate set number corresponding to the current value
    of ACTIVE_TEXTURE is greater than or equal to the
    implementation-dependent constant MAX_TEXTURE_COORDS_ARB, the error
    INVALID_OPERATION is generated by any such command. The active texture
    unit selector also selects the texture image unit accessed by commands
    involving texture image processing (section 3.8). Such commands include
    all variants of TexEnv, TexParameter, and TexImage commands,
    BindTexture, Enable/Disable for any texture target (e.g., TEXTURE_2D),
    and queries of all such state. If the texture image unit number
    corresponding to the current value of ACTIVE_TEXTURE is greater than or
    equal to the implementation-dependent constant
    MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB, the error INVALID_OPERATION is
    generated by any such command.

    ActiveTexture generates the error INVALID_ENUM if an invalid <texture>
    is specified. <texture> is a symbolic constant of the form TEXTUREi,
    indicating that texture unit i is to be modified. The constants obey
    TEXTUREi = TEXTURE0 + i (i is in the range 0 to k-1, where k is the
    larger of the MAX_TEXTURE_COORDS_ARB and
    MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB). For compatibility with old OpenGL
    specifications, the implementation-dependent constant MAX_TEXTURE_UNITS
    specifies the number of conventional texture units supported by the
    implementation. Its value must be no larger than the minimum of
    MAX_TEXTURE_COORDS_ARB and MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB.

    (modify third paragraph, p. 38) The state required to implement
    transformations consists of a 4-valued integer indicating the current
    matrix mode, one stack of at least two 4x4 matrices for each of COLOR,
    PROJECTION, each texture coordinate set, and TEXTURE and a stack of at
    least 32 4x4 matrices for MODELVIEW. Each matrix stack has an associated
    stack pointer. Initially, there is only one matrix on each stack, and
    all matrices are set to the identity. The initial matrix mode is
    MODELVIEW. The initial value of ACTIVE_TEXTURE is TEXTURE0.

    Modify Section 2.11, Clipping (p. 42)

    (Add a new paragraph between the first and second paragraph on page 43)
    When a vertex shader is active, the vector (Xe Ye Ze We)T is no longer
    computed by the GL. Instead, the OpenGL Shading Language provides a
    gl_ClipVertex built-in variable, which is used in the dot product with
    the client-defined clip planes. The user must ensure that the clip
    vertex and client-defined clip planes are defined in the same coordinate
    space. Note that the value in gl_ClipVertex is undefined if it is not
    written by the vertex shader.

    Modify Section 2.12, Current Raster Position

    (modify fourth paragraph, p.44) The coordinates are treated as if they
    were specified in a Vertex command. If a vertex shader is active, this
    vertex shader is executed using the x, y, z, and w coordinates as the
    object coordinates of the vertex. Otherwise, the x, y, z, and w
    coordinates are transformed by the current model-view and projection
    matrices. These coordinates, along with current values, are used to
    generate primary and secondary colors and texture coordinates just as is
    done for a vertex. The colors and texture coordinates produced using
    either method replace the colors and texture coordinates stored in the
    current raster position's associated data. If a vertex shader is active
    then the current raster distance is set to the value of the shader built
    in varying gl_FogFragCoord. Otherwise, if the value of the fog source
    (see section 3.10) is FOG_COORDINATE_SOURCE, then the current raster
    distance is set to the value of the current fog coordinate. Otherwise,
    the current raster distance is set to the distance from the origin of
    the eye coordinate system to the vertex as transformed by only the
    current model-view matrix. This distance may be approximated as
    discussed in section 3.10.

    Since vertex shaders may be executed when the raster position is set,
    any attributes not written by the shader will result in undefined state
    in the current raster position. Vertex shaders should output all varying
    variables that would be used when rasterizing pixel primitives using the
    current raster position.

    Modify section 2.13.1, Lighting

    (add just before the first paragraph on page 53) Additionally, vertex
    shaders can operate in two-sided color mode. When a vertex shader is
    active, front and back colors can be computed by the vertex shader and
    written to the gl_FrontColor, gl_BackColor, gl_FrontSecondaryColor and
    gl_BackSecondaryColor outputs. If VERTEX_PROGRAM_TWO_SIDE_ARB is
    enabled, the GL chooses between front and back colors, as described
    later. Otherwise, the front color output is always selected. Two-sided
    color mode is enabled and disabled by calling Enable or Disable with the
    symbolic value VERTEX_PROGRAM_TWO_SIDE_ARB.

    The selection between back color and front color depends ...

    Modify section 2.13.2, Lighting Parameter Specification (p. 53)

    (add a new paragraph between paragraph 2 and 3 on page 54) Material
    properties can be changed inside a Begin/End pair by calling Material.
    However, when a vertex shader is active such property changes are not
    guaranteed to update material parameters, defined in Table 2.8, until
    the following End command.

    Modify section 2.13.3, ColorMaterial

    (add a paragraph after the last paragraph, page 57) Material properties
    can be changed inside a Begin/End pair indirectly by enabling
    ColorMaterial mode and making Color calls. However, when a vertex shader
    is active such property changes are not guaranteed to update material
    parameters, defined in Table 2.8, until the following End command.

    Rename and Modify Section 2.13.8, Color and Texture Coordinate Clipping
    (p.59)

    (modify second paragraph, p. 60) Texture coordinates, varying variables,
    fog coordinates and point sizes computed on a per vertex basis must also
    be clipped when a primitive is clipped. The method is exactly analogous
    to that used for color clipping.

    Add Section 2.15 Vertex Shaders

    The sequence of operations described in Sections 2.10 through 2.13 is a
    fixed- functionality method for processing vertex data. Applications can
    more generally describe the operations that occur on vertex values and
    their associated data by using a "vertex shader".

    A vertex shader is an array of strings containing source code for the
    operations that are meant to occur on each vertex that is processed. The
    language used for vertex shaders is described in the OpenGL Shading
    Language Specification.

    Section 2.15.1 Creating and Using a Vertex Shader

    In order to create a vertex shader, use CreateShaderObjectARB with a
    <shaderType> of VERTEX_SHADER_ARB. The source code for the vertex shader
    can be specified in one call using the ShaderSourceARB command. A vertex
    shader is compiled by issuing CompileShaderARB and attached to a program
    object by calling AttachObjectARB. When LinkProgramARB is issued, all of
    the attached shader objects of type VERTEX_SHADER_ARB are linked
    together to create an executable program that can be used to process
    vertex values and their associated data. This executable program will
    become part of current state and used to process subsequent vertex data
    when UseProgramObjectARB is called.

    CreateShaderObjectARB will set the object specific parameter
    OBJECT_SUBTYPE_ARB to VERTEX_SHADER_ARB.

    Section 2.15.2 Vertex Attributes

    Conventional vertex attributes are communicated to the vertex shader via
    the Vertex, Normal, Color, etc. API commands (or their vertex array
    equivalents). Generic attributes are communicated via the
    VertexAttrib*ARB and VertexAttribPointerARB commands. The binding of the
    name of an attribute variable to a generic attribute index is assigned
    by the GL and can be queried. This generic attribute index can be used
    in the VertexAttrib*ARB and VertexAttribPointerARB commands.
    Alternatively, this binding can be overridden and specified explicitly,
    if so desired. These mechanisms make it possible for a vertex shader to
    use descriptive names for attribute variables rather than having generic
    variables numbered from 0 to MAX_VERTEX_ATTRIBS_ARB - 1. For example, an
    application can specify that the attribute variable "tangentVec" in a
    vertex shader will be mapped to generic vertex attribute number 6.

    When an attribute variable declared as a float, vec2, vec3 or vec4 is
    bound to a generic attribute index <i>, its value(s) are stored in the
    <x>, <x,y>, <x,y,z> or <x,y,z,w> components of the generic attribute at
    slot <i>. Recall that the VertexAttrib*ARB commands will set unused
    components to default values (either 0 or 1, see section 2.7).

    An attribute (either conventional or generic) is considered "active" if
    it is determined by the compiler and linker that the attribute may be
    accessed when the executable code is executed. The program object will
    fail to link if the sum of the active generic and active conventional
    attributes exceeds MAX_VERTEX_ATTRIBS_ARB. Attribute variables that are
    declared in a vertex shader but never used will not count against the
    limit. In cases where the compiler and linker cannot make a conclusive
    determination, the attribute will be considered active.

    To determine which of the declared vertex attribute variables an
    executable program uses and to determine their types, use the command:

        void GetActiveAttribARB(handleARB programObj, uint index,
                                sizei maxLength, sizei *length, int *size,
                                enum *type, charARB *name)

    This command provides information about the attribute selected by
    <index>. The <index> of 0 selects the first active attribute, and
    <index> of OBJECT_ACTIVE_ATTRIBUTES_ARB-1 selects the last active
    attribute. The value of OBJECT_ACTIVE_ATTRIBUTES_ARB can be queried with
    GetObjectParameter{if}vARB (see Section 6.1.12). If <index> is greater
    than or equal to OBJECT_ACTIVE_ATTRIBUTES_ARB, the error INVALID_VALUE
    is generated.

    The parameter <programObj> is a handle to a program object for which the
    command LinkProgramARB has been issued in the past. It is not necessary
    for <programObj> to have been linked successfully. The link could have
    failed because the number of active attributes exceeded the limit. If
    <programObj> is not of type PROGRAM_OBJECT_ARB, the error
    INVALID_OPERATION is generated.

    For the selected attribute, the attribute name is returned into <name>.
    The string <name> will be null terminated. The actual number of
    characters written by the GL into <name> is returned in <length>. This
    count excludes the null termination. If <length> is NULL then the GL
    ignores this parameter. The maximum number of characters the GL is
    allowed to write into <name> is passed in by <maxLength>. The returned
    attribute name can be the name of a generic attribute or a conventional
    attribute (which begin with the prefix "gl_", see the OpenGL Shading
    Language specification for a complete list). The length of the longest
    attribute name in <programObj> is given by
    OBJECT_ACTIVE_ATTRIBUTE_MAX_LENGTH_ARB, which can be queried with
    GetObjectParameter{if}vARB (see Section 6.1.12).

    For the selected attribute, the type of the attribute is returned into
    <type>. The size of the attribute is returned into <size>. The value in
    <size> is in units of the type returned in <type>. The type returned can
    be any of FLOAT, FLOAT_VEC2_ARB, FLOAT_VEC3_ARB, FLOAT_VEC4_ARB,
    FLOAT_MAT2_ARB, FLOAT_MAT3_ARB, or FLOAT_MAT4_ARB.

    If an error occurred, the return parameters <length>, <size>, <type> and
    <name> will be unmodified.

    This command will return as much information about active attributes as
    possible. If no information is available, <length> will be set to zero
    and <name> will be an empty string. This situation could arise if
    GetActiveAttribARB is issued after a failed link.

    After a program object has been linked successfully, the bindings of
    attribute variable names to indices can be queried.

    The command

        GLint GetAttribLocationARB(handle programObj, const charARB *name)

    will return the generic attribute index that the attribute variable
    named <name> is bound to. <name> has to be a null terminated string.
    <programObj> is a handle to a program object that has been successfully
    linked. If <name> is not an active attribute, or if <name> is a
    conventional attribute, -1 will be returned. If <name> is an attribute
    matrix, and it is active, then GetAttribLocationARB returns the index of
    the first column of that matrix. If <programObj> has not been
    successfully linked, or if <programObj> is not of type
    PROGRAM_OBJECT_ARB, the error INVALID_OPERATION is generated.

    The binding of an attribute variable to a generic attribute index can
    also be specified explicitly.

    The command

        void BindAttribLocationARB(handleARB programObj, uint index,
                                   const charARB *name)

    binds the attribute variable named <name> to generic vertex attribute
    <index> for the program object <programObj>. If <name> was bound
    previously, that binding information is lost. <name> has to be a null
    terminated string. If <name> is declared as an attribute matrix, <index>
    indicates the first column of that matrix. The other matrix columns are
    then automatically bound to locations <index> + 1 for a mat2, <index> +
    1 and <index> + 2 for a mat3 and <index> + 1, <index> + 2 and <index> +
    3 for a mat4. The error INVALID_VALUE is generated if <index> is equal
    or greater than MAX_VERTEX_ATTRIBS_ARB. The error INVALID_OPERATION is
    generated if <name> starts with the reserved "gl_" prefix. If
    <programObj> is not of type PROGRAM_OBJECT_ARB, the error
    INVALID_OPERATION is generated.

    Applications cannot bind any conventional attributes (the attributes
    starting with "gl_" defined in the OpenGL Shading Language
    specification). Conventional attributes used in a vertex shader will be
    bound automatically.

    In order to have attribute bindings take effect the program object needs
    to be linked after issuing the BindAttribLocationARB command. The linker
    will bind any active attributes not already explicitly specified through
    BindAttribLocationARB. After a successful link the complete list of
    bindings can be queried using the command GetAttribLocationARB.

    It is generally not known to the GL until link time if <name> is
    declared as an attribute matrix. Therefore LinkProgramARB can fail
    because binding of a matrix to a generic attribute index caused one or
    more columns of the matrix to be bound to indices outside the
    MAX_VERTEX_ATTRIBS_ARB range. LinkProgramARB can also fail because
    BindAttribLocationsARB commands have been issued leaving not enough
    contiguous indices for a matrix to be fitted in by the linker.

    It is allowed to issue BindAttribLocationARB before any vertex shader
    objects are attached to a program object. Hence it is allowed to bind
    any name (except a name starting with "gl_") to an index, including a
    name that is never used as an attribute in any vertex shader object.

    The values of generic attributes sent to generic attribute index <i> are
    part of current GL state, just like the conventional attributes. If a
    new program object has been made active, then these values will be
    tracked by the GL in such a way that the same values will be observed by
    attributes in the new program object that are also bound to index <i>.

    It is a link error to have a vertex shader statically consume the
    conventional attribute gl_Vertex and have the same vertex shader
    statically consume a generic vertex attribute that is bound to index
    zero. This would result in aliasing between a generic vertex attribute
    and the conventional attribute gl_Vertex, which is disallowed (see the
    description of VertexAttrib*ARB in Section 2.7).

    It is possible for an application to bind more than one attribute name
    to the same location. This is referred to as "aliasing". This will only
    work if only one of the aliased attributes is active in the executable
    program, or if no path through the shader consumes more than one
    attribute of a set of attributes aliased to the same location. OpenGL
    implementations are not required to detect violations of this rule. The
    compiler and linker are allowed to assume that no aliasing is done, and
    are free to employ optimizations that work only in the absence of
    aliasing. It is not possible to alias generic attributes with
    conventional ones.

    When a vertex shader is active, more (or less) vertex attributes might
    be sent to OpenGL than the vertex shader uses. When fewer attributes
    than the current vertex shader uses are sent, the vertex shader uses the
    values in the current state for the attributes not provided. In the case
    of a generic vertex attribute that current value comes from the slot
    index number the generic attribute is bound to. When more attributes are
    provided than the vertex shader uses the vertex shader ignores the extra
    attributes but these extra attributes continue to modify the current
    vertex attribute state.

    Section 2.15.3 Uniform Variables

    A vertex shader may define one or more "uniform" variables. These values
    are to remain constant over a primitive or a sequence of primitives. The
    OpenGL Shading Language specification defines a set of built-in uniform
    variables for vertex shaders that correspond to the state that GL
    manages for the purpose of processing vertices. The amount of storage
    that is available for vertex shader uniform variables is specified by
    the implementation dependent constant MAX_VERTEX_UNIFORM_COMPONENTS_ARB.
    This value represents the number of individual floating point values, or
    individual integer values or individual Boolean values that can be held
    in uniform variable storage for a vertex shader. A link error will be
    generated if an attempt is made to utilize more than the space available
    for vertex shader uniform variables.

    Section 2.15.4 Varying Variables

    A vertex shader may define one or more "varying" variables (see the
    OpenGL Shading Language specification). These values are expected to be
    interpolated across the primitive being rendered. The OpenGL Shading
    Language specification defines a set of built-in varying variables for
    vertex shaders that correspond to the values required for the fixed
    functionality processing that occurs after vertex processing. The number
    of interpolators that is available for processing varying variables is
    specified by the implementation dependent constant
    MAX_VARYING_FLOATS_ARB. This value represents the number of individual
    floating point values that can be interpolated, therefore interpolating
    a vertex will consume 4 (x, y, z, w) and interpolating a color will also
    consume 4 (R, G, B, A). A link error will be generated if an attempt is
    made to utilize more than MAX_VARYING_FLOATS_ARB interpolators.

    Section 2.15.5 Texture access

    Vertex shaders have the ability to do a lookup into a texture map. Zero
    or more texture image units are available to a vertex shader to access
    textures. The maximum number of texture image units available to a
    vertex shader is MAX_VERTEX_TEXTURE_IMAGE_UNITS_ARB. The maximum number
    of texture image units available to the fragment stage of OpenGL is
    MAX_TEXTURE_IMAGE_UNITS_ARB. Both the vertex shader and fragment
    processing combined cannot use more than
    MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB texture image units. If both the
    vertex shader and the fragment processing stage access the same texture
    image unit, then that counts as using two texture image units against
    the MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB limit.

    When a texture lookup is performed in a vertex shader, the GL computes
    the filtered texture value tau in the manner described in sections 3.8.8
    and 3.8.9, and converts it to a texture source color C_s according to
    table 3.21 (section 3.8.13). The GL returns a four-component vector
    (R_s, G_s, B_s, A_s) to the vertex shader.

    In a vertex shader, it is not possible to perform automatic LOD
    calculations using partial derivatives of the texture coordinates with
    respect to window coordinates as described in section 3.8.8. Hence,
    there is no automatic selection of an image array level. Minification or
    magnification of a texture map is controlled by a LOD value optionally
    passed as an argument in the texture lookup functions. If the texture
    lookup function supplies an explicit LOD value <l>, then the pre-bias
    LOD value lambda_base(x,y) = l (equation 3.16a). If the texture lookup
    function does not supply an explicit LOD value, then lambda_base(x,y) =
    0. The scale factor rho(x,y) and its approximation function f(x,y),
    pages 141-142, are ignored.

    Texture comparison requires the vertex shader to use the shadow versions
    of the texture lookup functions. This in turn means that samplers of
    type sampler1DShadow or sampler2DShadow or sampler2DRectShadow need to
    be used to indicate the texture image unit that has a depth texture
    bound to it.

    The results of a texture lookup function are undefined if:

      * The sampler used in a texture lookup function is of type sampler1D
        or sampler2D or sampler2DRect, and the texture object's internal
        format is DEPTH_COMPONENT, and the TEXTURE_COMPARE_MODE is not NONE.
      * The sampler used in a texture lookup function is of type
        sampler1DShadow or sampler2DShadow or sampler2DRectShadow, and the
        texture object's internal format is DEPTH_COMPONENT, and the
        TEXTURE_COMPARE_MODE is NONE.
      * The sampler used in a texture lookup function is of type
        sampler1DShadow or sampler2DShadow or sampler2DRectShadow, and the
        texture object's internal format is not DEPTH_COMPONENT.

    If a vertex shader uses a sampler which associated texture object is not
    complete, as defined in section 3.8.10, the texture image unit will
    return (R, G, B, A) = (0, 0, 0, 1).

    Section 2.15.6 Position Invariance

    If a vertex shader uses the built-in function ftransform() to generate a
    vertex position, then this generally guarantees that the transformed
    position will be the same whether using this vertex shader or the fixed
    functionality OpenGL pipeline. This allows for correct mixed multi-pass
    rendering algorithms.

    Section 2.15.7 Vertex Shader Execution

    If a successfully linked program object that contains a vertex shader is
    made current by calling UseProgramObjectARB, the executable version of
    the vertex shader is used to process incoming vertex values rather than
    the fixed functionality vertex processing described in Sections
    2.10-2.13. In particular,

      * The model-view matrix is not applied to vertex coordinates (pp.
        31-32)
      * The projection matrix is not applied to vertex coordinates (pp.
        31-32)
      * The texture matrices are not applied to texture coordinates (p. 37)
      * Normals are not transformed to eye coordinates (pp. 38-39)
      * Normals are not rescaled or normalized (pp. 39-40)
      * Normalization of AUTO_NORMAL evaluated normals is not performed. (p.
        191)
      * Texture coordinates are not generated automatically (pp. 40-42)
      * Per vertex lighting is not performed (pp. 48-53)
      * Color material computations are not performed (pp. 54-57)
      * Color index lighting is not performed (pp. 57-58)
      * All of the above applies when setting the current raster position
        (pp. 44-45)

    The following operations are applied to vertex values that are the
    result of executing the vertex shader:

      * Color clamping or masking (p. 56)
      * Perspective division on clip coordinates (pp. 31-32)
      * Viewport mapping (pp. 32-33)
      * Depth range (p. 33)
      * Clipping, including user clipping (pp. 42-44)
      * Front face determination (p. 53)
      * Flat-shading (pp. 58-59)
      * Color, texture coordinate, fog, point-size and generic attribute
        clipping (pp. 59-60)
      * Final color processing (p. 60)

    Section 2.15.8 Required State

    Besides the required state outlined in the ARB_shader_objects extension
    document, the following state is also required:

    The state required to support vertex shaders consists of:

      * A bit indicating whether or not vertex program two-sided color mode
        is enabled, initially disabled.
      * A bit indicating whether or not vertex program point size mode is
        enabled, initially disabled.

    The state required per program object consists of:

      * An array of strings holding the names of active attributes.
      * An array of integers holding the sizes of active attributes.
      * An array of integers holding the types of active attributes.
      * An integer holding the number of active attributes
      * An integer holding the length of the longest attribute name.
      * An array of integers holding the locations of active generic
        attributes.
      * An array of MAX_VERTEX_UNIFORM_COMPONENTS_ARB words that holds
        uniform values.

Additions to Chapter 3 of the OpenGL 1.4 Specification (Rasterization)

    Modify Section 3.3, Points (p. 66)

    (replace the first and second paragraph) If a vertex shader is not
    active, then the rasterization of points is controlled with

        void PointSize(float size)

    <size> specifies the requested size of a point. The default value is
    1.0. A value less than or equal to zero results in the error
    INVALID_VALUE. The requested point size is multiplied with a distance
    attenuation factor, clamped to a specified point size range, and further
    clamped to the implementation-dependent point size range to produce the
    derived point size:

        derived_size = ...

    (add a paragraph between the third and fourth paragraph)

    If a vertex shader is active and vertex program point size mode is
    enabled, then the derived point size is taken from the, potentially
    clipped, shader built in gl_PointSize and clamped to the
    implementation-dependent point size range. If the value written to
    gl_PointSize is less than or equal to zero, results are undefined. If a
    vertex shader is active and vertex program point size mode is disabled,
    then the derived point size is taken from the point size state as
    specified by the PointSize command. In this case no distance attenuation
    is performed. Vertex program point size mode is enabled and disabled by
    calling Enable or Disable with the symbolic value
    VERTEX_PROGRAM_POINT_SIZE_ARB.

    If multisampling is enabled, an implementation ...

    Modify Section 3.8.8 Texture Minification (p. 140)

    (modify subsection Scale Factor and Level of Detail. Replace its first
    paragraph)

    The choice is governed by a scale factor rho(x,y) and the level of
    detail parameter lambda(x,y), defined as

    lambda_base(x,y) = log2[rho(x,y)]                                  3.16a

    lambda'(x,y) = lambda_base(x,y) +                                  3.16b
                   clamp(texobj_bias + texunit_bias)

                  |------- LODmax     lambda' > LODmax
    lambda(x,y) = |        lambda'    LODmin <= lambda' <= LODmax      3.16c
                  |        LODmin     lambda' < LODmin
                  |------- undefined  LODmin > LODmax

    texobj_bias is the value of TEXTURE_LOD_BIAS for the bound texture
    object (as described in section 3.8.4), and texunit_bias is the value of
    TEXTURE_LOD_BIAS for the current texture unit (as described in section
    3.8.13). The sum of these values is clamped to the range [-max_bias,
    max_bias] where max_bias is the value of the implementation defined
    constant MAX_TEXTURE_LOD_BIAS.

    Modify Section 3.9 Color Sum (p. 160)

    (modify third paragraph) Color sum is enabled or disabled using the
    generic Enable and Disable commands, respectively, with the symbolic
    constant COLOR_SUM. If lighting is enabled and if a vertex shader is not
    active, the color sum stage is always applied, ignoring the value of
    COLOR_SUM.

    Modify Section 3.10, Fog (p. 160)

    (modify second paragraph) If a vertex shader is active, or if the fog
    source, as defined below, is FOG_COORDINATE, then c is the interpolated
    value of the fog coordinate for this fragment. Otherwise, if the fog
    source is FRAGMENT_DEPTH, then c is the eye-coordinate distance from the
    eye, (0, 0, 0, 1) in eye coordinates, to the fragment center. The
    equation and the fog source, along with either d or e and s, is
    specified with

Additions to Chapter 4 of the OpenGL 1.4 Specification (Per-Fragment
Operations and the Frame Buffer)

    None

Additions to Chapter 5 of the OpenGL 1.4 Specification (Special Functions)

    Modify Section 5.1, Evaluators (p. 188)

    (modify next-to-last paragraph, p. 191) For MAP_VERTEX_3, let q = p. For
    MAP_VERTEX_4, let q=(x/w,y/w,z/w), where (x,y,z,w) = p. Then let

                dq   dq
            m = -- x --.
                du   dv

    Then the generated analytic normal, n, is given by n=m if a vertex
    shader is active, or else by n=m/|m|.

    Modify Section 5.4, Display Lists (p. 198)

    (change second paragraph, p. 202)

    Certain commands, when called when compiling a display list, are not
    compiled into the display list but are executed immediately. These are:
    IsList,...Flush, Finish, GetAttribLocationARB, GetActiveAttribARB,
    BindAttribLocationARB, VertexAttribPointerARB,
    EnableVertexAttribArrayARB, DisableVertexAttribArrayARB as well as
    IsEnabled and all of the Get commands defined in Section 6.1.12.

    (add just before the final paragraph of the section on p. 202)

    When a program object is in use, a display list may be executed whose
    vertex attribute calls do not match up exactly with what is expected by
    the vertex shader contained in that program object. How this mismatch is
    handled is described in section 2.15.2.

Additions to Chapter 6 of the OpenGL 1.4 Specification (State and State
Requests)

    Modify Section 6.1.2, Data Conversions (p. 205)

    (add to end of last paragraph, p. 206) Queries of texture state
    variables corresponding to texture coordinate processing unit (namely,
    TexGen state and enables, and matrices) will produce an
    INVALID_OPERATION error if the value of ACTIVE_TEXTURE is greater than
    or equal to MAX_TEXTURE_COORDS_ARB. All other texture state queries will
    result in an INVALID_OPERATION error if the value of ACTIVE_TEXTURE is
    greater than or equal to MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB.

    Modify section 6.1.12 Generic Object Queries.

    The commands

        void GetObjectParameterfvARB(handleARB obj, enum pname,
                                     float *params)
        void GetObjectParameterivARB(handleARB obj, enum pname, int *params)

    return object specific parameter values for object <obj> in <params>.
    The parameter value to return is specified by <pname>.

    If <pname> is OBJECT_SUBTYPE_ARB, GetObjectParameter{if}vARB returns
    VERTEX_SHADER_ARB if <obj> references a shader object. If <obj> is not
    of type SHADER_OBJECT_ARB, the error INVALID_OPERATION is generated.

    If <pname> is OBJECT_ACTIVE_ATTRIBUTES_ARB, GetObjectParameter{if}vARB
    returns the number of active attributes. If no active attributes exist,
    0 or 0.0f is returned. If <obj> is not of type PROGRAM_OBJECT_ARB the
    error INVALID_OPERATION is generated.

    If <pname> is OBJECT_ACTIVE_ATTRIBUTE_MAX_LENGTH_ARB,
    GetObjectParameter{if}vARB returns the length of the longest active
    attribute name, including a null termination. If no active attributes
    exist, 0 or 0.0f is returned. If <obj> is not of type PROGRAM_OBJECT_ARB
    the error INVALID_OPERATION is generated.

    The commands

        void GetVertexAttribdvARB(uint index, enum pname, double *params)
        void GetVertexAttribfvARB(uint index, enum pname, float *params)
        void GetVertexAttribivARB(uint index, enum pname, int *params)

    obtain the vertex attribute state named by <pname> for the generic
    vertex attribute numbered <index> and places the information in the
    array <params>. <pname> must be one of VERTEX_ATTRIB_ARRAY_ENABLED_ARB,
    VERTEX_ATTRIB_ARRAY_SIZE_ARB, VERTEX_ATTRIB_ARRAY_STRIDE_ARB,
    VERTEX_ATTRIB_ARRAY_TYPE_ARB, VERTEX_ATTRIB_ARRAY_NORMALIZED_ARB, or
    CURRENT_VERTEX_ATTRIB_ARB. Note that all the queries except
    CURRENT_VERTEX_ATTRIB_ARB return client state. The error INVALID_VALUE
    is generated if <index> is greater than or equal to
    MAX_VERTEX_ATTRIBS_ARB.

    All but the last query return information about generic vertex attribute
    arrays. The enable state of a generic vertex attribute array is set by
    the command EnableVertexAttribArrayARB and cleared by
    DisableVertexAttribArrayARB. The size, stride, type and normalized flag
    are set by the command VertexAttribPointerARB. The query
    CURRENT_VERTEX_ATTRIB_ARB returns the current value for the generic
    attribute <index>. In this case the error INVALID_OPERATION is generated
    if <index> is zero, as there is no current value for generic attribute
    zero.

    Note that all the queries except CURRENT_VERTEX_ATTRIB_ARB return client
    state. The error INVALID_VALUE is generated if <index> is greater than
    or equal to MAX_VERTEX_ATTRIBS_ARB.

    The command

        void GetVertexAttribPointervARB(uint index, enum pname,
                                        void **pointer)

    obtains the pointer named <pname> for vertex attribute numbered <index>
    and places the information in the array <pointer>. <pname> must be
    VERTEX_ATTRIB_ARRAY_POINTER_ARB. The INVALID_VALUE error is generated if
    <index> is greater than or equal to MAX_VERTEX_ATTRIBS_ARB.

Additions to Appendix A of the OpenGL 1.4 Specification (Invariance)

    Add to end of Section A.3 (p. 250):

    The same vertex shader will produce the same result when run multiple
    times with the same input. The wording 'the same shader' means a program
    object that is populated with the same source strings, which are
    compiled and then linked, possibly multiple times. This program object
    is then executed using the same GL state vector.

Additions to the AGL/GLX/WGL Specifications

    None

Interactions with ARB_vertex_program and ARB_fragment_program

    Mixing a high level ARB_vertex_shader shader with a low level
    ARB_fragment_program shader is allowed. However, a high level
    ARB_vertex_shader shader and a low level ARB_vertex_program shader
    cannot be active at the same time.

Interactions with ARB_vertex_program

    Enabling an ARB_vertex_shader shader by issuing the command
    UseProgramObjectARB(), with a handle which is not zero, results in any
    low level vertex shader to be ignored and overrides the enable
    VERTEX_PROGRAM_ARB. Note that the value for VERTEX_PROGRAM_ARB does not
    change by installing an ARB_vertex_shader shader.

    The following enums and entry points are re-used from
    ARB_vertex_program:

        MAX_VERTEX_ATTRIBS_ARB                          0x8869
        VERTEX_ATTRIB_ARRAY_ENABLED_ARB                 0x8622
        VERTEX_ATTRIB_ARRAY_SIZE_ARB                    0x8623
        VERTEX_ATTRIB_ARRAY_STRIDE_ARB                  0x8624
        VERTEX_ATTRIB_ARRAY_TYPE_ARB                    0x8625
        VERTEX_ATTRIB_ARRAY_NORMALIZED_ARB              0x886A
        CURRENT_VERTEX_ATTRIB_ARB                       0x8626
        VERTEX_ATTRIB_ARRAY_POINTER_ARB                 0x8645
        VERTEX_PROGRAM_TWO_SIDE_ARB                     0x8643
        VERTEX_PROGRAM_POINT_SIZE_ARB                   0x8642

        VertexAttrib*ARB
        VertexAttribPointerARB
        EnableVertexAttribArrayARB
        DisableVertexAttribArrayARB
        GetVertexAttrib*ARB
        GetVertexAttribPointervARB

    The ARB_vertex_program extension is modified to turn undefined vertex
    attribute aliasing into non-aliasing. That means that generic vertex
    attributes and conventional attributes do not alias on top of each
    other.

    The following edits are against version 36 of the ARB_vertex_program
    specification.

      * Section 2.7. Delete the paragraph above Table X.1, Generic and
        Conventional Attribute Mappings. Add the following sentence in its
        place: "There is no aliasing among generic attributes and
        conventional attributes."
      * Section 2.7. Delete the paragraph below Table X.1
      * Section 2.7. Delete Table X.1.
      * Change the last sentence in Section 2.7 to: "The initial values for
        all generic vertex attributes are (0, 0, 0, 1)."
      * Section 2.8. Edit the paragraph below the pseudo code. Change "As
        described in section 2.7, setting a generic vertex attributes ..."
        into "As described in section 2.7, there is no aliasing among
        generic vertex attributes and conventional attributes".
      * Section 2.14.3.1. Delete the first part of the paragraph above Table
        X.2.1. The new paragraph becomes "A vertex program will fail to load
        if it binds both a conventional vertex attribute and a generic
        vertex attribute listed in the same row of Table X.2.1."

Interactions with ARB_fragment_program

    The following enums are re-used from ARB_fragment_program:

        MAX_TEXTURE_IMAGE_UNITS_ARB                     0x8872
        MAX_TEXTURE_COORDS_ARB                          0x8871

Interactions with ARB_vertex_blend

    If ARB_vertex_blend is supported, and if a vertex shader is active, the
    executable version of the vertex shader is used to process incoming
    vertex values rather than the fixed functionality vertex blend
    processing described in ARB_vertex_blend. In particular,

      * Vertex coordinates are not transformed to eye space as described by
        the equation in Section 2.10 provided by ARB_vertex_blend.
      * Normals are not transformed to eye space as described by the
        equations in Section 2.10.2 provided by ARB_vertex_blend.

Interactions with ARB_matrix_palette

    If ARB_matrix_palette is supported, and if a vertex shader is active,
    the executable version of the vertex shader is used to process incoming
    vertex values rather than the fixed functionality matrix processing
    described in ARB_matrix_palette. In particular,

      * Vertex coordinates are not transformed to eye space as described by
        the equation in Section 2.10 of the ARB_matrix_palette
        specification.
      * Normals are not transformed to eye space as described by the
        equations in Section 2.10.3 of the ARB_matrix_palette specification.

GLX Protocol

    The ARB_vertex_program extension defines GLX protocol for the following
    rendering commands.

        VertexAttrib1fvARB, VertexAttrib1svARB, VertexAttrib1dvARB, 
        VertexAttrib2fvARB, VertexAttrib2svARB, VertexAttrib2dvARB,
        VertexAttrib3fvARB, VertexAttrib3svARB, VertexAttrib3dvARB,
        VertexAttrib4fvARB, VertexAttrib4svARB, VertexAttrib4dvARB,
        VertexAttrib4ivARB, VertexAttrib4bvARB, VertexAttrib4ubvARB,
        VertexAttrib4usvARB, VertexAttrib4uivARB, VertexAttrib4NbvARB,
        VertexAttrib4NsvARB, VertexAttrib4NivARB, VertexAttrib4NubvARB,
        VertexAttrib4NusvARB, VertexAttrib4NuivARB

    The following rendering commands are sent to the server as part of a
    glXRender request:

    BindAttribLocationARB

        2           12+n+p           rendering command length
        2           265              rendering command opcode
        4           CARD32           programObj
        4           CARD32           index
        n           LISTofBYTE       name, n = strlen(name) + 1
        p                            unused, p = pad(n)

    The following non-rendering commands are added.

    GetActiveAttribARB

        1           CARD8            opcode (X assigned)
        1           179              GLX opcode
        2           5                request length
        4           GLX_CONTEXT_TAG  context tag
        4           CARD32           programObj
        4           CARD32           index
        4           CARD32           maxLength
      =>
        1           1                reply
        1                            unused
        2           CARD16           sequence number
        4           m                reply length, m = (n+p)/4
        4                            unused
        4           CARD32           n
        4           INT32            size
        4           ENUM             type
        8                            unused
        n           LISTofBYTE       name, n = length+1
        p                            unused , p = pad(n)

    Note that n may be zero, indicating that a GL error occured.

    GetAttribLocationARB

        1           CARD8            opcode (X assigned)
        1           180              GLX opcode
        2           (12+n+p)/4       request length, n = strlen(name) + 1
        4           GLX_CONTEXT_TAG  context tag
        4           CARD32           programObj
        n           LISTofBYTE       name
        p                            unused, p = pad(n)
      =>
        1           1                reply
        1                            unused
        2           CARD16           sequence number
        4           0                reply length
        4           INT32            return value
        20                           unused

    GetVertexAttribdvARB

        1           CARD32           opcode (X assigned)
        1           181              GLX opcode
        2           4                request length
        4           GLX_CONTEXT_TAG  context tag
        4           CARD32           index
        4           ENUM             pname
      =>
        1           1                reply
        1                            unused
        2           CARD16           sequence number
        4           m                reply length, m = ((n == 1) ? 0 : 2*n)
        4                            unused
        4           CARD32           n

        if (n = 1) this follows:

        8           FL0AT64          params
        8                            unused

        otherwise this follows:

        16                           unused
        8*n         LISTofFLOAT64    params

    The ARB_vertex_program extension defines GLX protocol for the following
    non-rendering commands.

        GetVertexAttribfvARB and GetVertexAttribivARB.

    The following are entirely client-side commands and do not
    require GLX protocol.

        GetVertexAttribPointervARB, VertexAttribPointerARB
        EnableVertexAttribArrayARB, DisableVertexAttribArrayARB

Errors

    The error INVALID_VALUE is generated by any command that takes one or
    more handles as input, and one or more handles are not an object handle
    generated by OpenGL.

    The error INVALID_VALUE is generated by any of the VertexAttrib*ARB
    commands if <index> is equal or greater than MAX_VERTEX_ATTRIBS_ARB.

    The error INVALID_VALUE is generated by VertexAttribPointerARB if
    <index> is equal or greater than MAX_VERTEX_ATTRIBS_ARB.

    The error INVALID_VALUE is generated by VertexAttribPointerARB if <size>
    is not one of 1, 2, 3, or 4.

    The error INVALID_VALUE is generated by VertexAttribPointerARB if
    <stride> is negative.

    The error INVALID_VALUE is generated by EnableVertexAttribArrayARB or
    DisableVertexAttribArrayARB if <index> is greater than or equal to
    MAX_VERTEX_ATTRIBS_ARB.

    The error INVALID_OPERATION is generated by any command accessing
    texture coordinate processing state if the texture unit number
    corresponding to the current value of ACTIVE_TEXTURE is greater than or
    equal to the implementation-dependent constant MAX_TEXTURE_COORDS_ARB.

    The error INVALID_OPERATION is generated by any command accessing
    texture image processing state if the texture unit number corresponding
    to the current value of ACTIVE_TEXTURE is greater than or equal to the
    implementation-dependent constant MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB.

    The error INVALID_OPERATION is generated by GetActiveAttribARB if
    <programObj> is not of type PROGRAM_OBJECT_ARB.

    The error INVALID_VALUE is generated by GetActiveAttribARB if <index> is
    greater than or equal to OBJECT_ACTIVE_ATTRIBUTES_ARB.

    The error INVALID_OPERATION is generated by GetAttribLocationARB if
    <programObj> is not of type PROGRAM_OBJECT_ARB or if <programObj> has
    not been linked successfully.

    The error INVALID_OPERATION is generated by BindAttribLocationARB if
    <name> starts with the reserved "gl_" prefix.

    The error INVALID_VALUE is generated by BindAttribLocationARB if <index>
    is equal or greater than MAX_VERTEX_ATTRIBS_ARB.

    The error INVALID_OPERATION is generated by BindAttribLocationARB if
    <programObj> is not of type PROGRAM_OBJECT_ARB.

    The error INVALID_OPERATION is generated by GetObjectParameter{if}vARB
    if <pname> is OBJECT_ACTIVE_ATTRIBUTES_ARB or
    OBJECT_ACTIVE_ATTRIBUTE_MAX_LENGTH_ARB and <obj> is not of type
    PROGRAM_OBJECT_ARB.

    The error INVALID_OPERATION is generated by GetObjectParameter{if}vARB
    if <pname> is OBJECT_SUBTYPE_ARB and <obj> is not of type
    SHADER_OBJECT_ARB.

    The error INVALID_OPERATION is generated by GetVertexAttrib*ARB if
    <index> is zero and <pname> is CURRENT_VERTEX_ATTRIB_ARB.

    The error INVALID_VALUE is generated by GetVertexAttrib*ARB if <index>
    is greater than or equal to MAX_VERTEX_ATTRIBS_ARB.

    The error INVALID_VALUE is generated by GetVertexAttribPointervARB if
    <index> is equal or greater than MAX_VERTEX_ATTRIBS_ARB.

New State
                                                                   Initial
    Get Value                     Type  Get Command                Value      Description              Sec.   Attribute
    ---------                     ----  -------------------------  -------    -----------              ----   ---------
    VERTEX_PROGRAM_TWO_SIDE_ARB     B    IsEnabled                  False     two-sided color mode     2.13.1    enable
    CURRENT_VERTEX_ATTRIB_ARB    16+xR4  GetVertexAttributeARB      (0,0,0,1) generic vertex attribute 2.7       current
    VERTEX_PROGRAM_POINT_SIZE_ARB   B    IsEnabled                  False     point size mode          3.3       enable

    Table X.  New Accessible State Introduced by ARB_vertex_shader


    Get Value                        Type    Get Command    Initial Value Description         Section  Attribute
    -------------------------------  ------  -------------  ------------- ------------------  -------  ------------
    VERTEX_ATTRIB_ARRAY_ENABLED_ARB  16+xB   GetVertex-     False         vertex attrib       2.8      vertex-array
                                             AttribARB                    array enable
    VERTEX_ATTRIB_ARRAY_SIZE_ARB     16+xZ   GetVertex-     4             vertex attrib       2.8      vertex-array
                                             AttribARB                    array size
    VERTEX_ATTRIB_ARRAY_STRIDE_ARB   16+xZ+  GetVertex-     0             vertex attrib       2.8      vertex-array
                                             AttributeARB                 array stride
    VERTEX_ATTRIB_ARRAY_TYPE_ARB     16+xZ4  GetVertex-     FLOAT         vertex attrib       2.8      vertex-array
                                             AttributeARB                 array type
    VERTEX_ATTRIB_ARRAY_             16+xB   GetVertex-     False         vertex attrib       2.8      vertex-array
        NORMALIZED_ARB                       AttributeARB                 array normalized
    VERTEX_ATTRIB_ARRAY_POINTER_ARB  16+xP   GetVertex-     NULL          vertex attrib       2.8      vertex-array
                                             AttribPointerARB             array pointer

    Table X  New Accessible Client State Introduced by ARB_vertex_shader

                                                             Initial
    Get Value              Type   Get Command                Value             Description          Sec.   Attribute
    ---------              -----  -----------------------   -------------      -----------          ----   ---------
    OBJECT_TYPE_ARB        Z2     GetObjectParameterivARB   SHADER_OBJECT_ARB  Type of object       2.15.1        -
    OBJECT_SUBTYPE_ARB     Z2     GetObjectParameterivARB   VERTEX_SHADER_ARB  Sub type of object   2.15.1        -

    Table X Shader object state.

                                                                      Initial
    Get Value                     Type       Get Command              Value     Description                  Sec.    Attribute
    ---------                     -----      -----------------------  --------  -----------                  ----    ---------
    -                             0+xZ+      GetActiveAttribARB         -       Size of active attribute     2.15.2      -
    -                             0+xZ+      GetActiveAttribARB         -       Type of active attribute     2.15.2      -
    -                             0+xcharARB GetActiveAttribARB        ""       Name of active attribute     2.15.2      -
    -                             0+xZ       GetAttribLocationARB       -       Location of active generic   2.15.2      -
                                                                                attribute
                                  512+xR     GetUniformARB              0       Uniform value                2.15.3      -
    OBJECT_ACTIVE_ATTRIBUTES_ARB  Z+         GetObjectParameterivARB    0       Number of active attributes  2.15.2      -
    OBJECT_ACTIVE_ATTRIBUTES_
                MAX_LENGTH_ARB    Z+         GetObjectParameterivARB    0       Max attribute name length    6.1.12      -

    Table X Program object state.


New Implementation Dependent State

                                                                Minimum
    Get Value                               Type  Get Command    Value   Description                    Sec.  Attr.
    ---------                               ----  -----------   -------  -----------                    ----  -----
    MAX_VERTEX_ATTRIBS_ARB                  Z+    GetIntegerv     16     Number of active               2.15.2   -
                                                                         vertex attributes
    MAX_VERTEX_UNIFORM_COMPONENTS_ARB       Z+    GetIntegerv    512     Number of words                2.15.3   -
                                                                         for vertex shader
                                                                         uniform variables
    MAX_VARYING_FLOATS_ARB                  Z+    GetIntegerv     32     Number of floats               2.15.4   -
                                                                         for interpolated
                                                                         (varying) variables
    MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB    Z+    GetIntegerv      2     Number of texture image        2.15.5   -
                                                                         units that can be accessed
                                                                         by a vertex shader and the
                                                                         fragment processing part of
                                                                         OpenGL combined.
    MAX_VERTEX_TEXTURE_IMAGE_UNITS_ARB      Z+    GetIntegerv      0     Number of separate             2.15.5   -
                                                                         texture units that can
                                                                         be accessed by a vertex
                                                                         shader.
    MAX_TEXTURE_IMAGE_UNITS_ARB             Z+    GetIntegerv      2     Number of separate             2.15.5   -
                                                                         texture units that can
                                                                         be accessed by the fragment
                                                                         processing part of OpenGL.
    MAX_TEXTURE_COORDS_ARB                  Z+    GetIntegerv      2     Number of texture              2.7      -
                                                                         coordinate sets

    Table X  New Implementation Dependent State Introduced by ARB_vertex_shader

Sample Usage

    GLboolean init(GLcharARB *vertexShader)
    {
        GLcharARB *pInfoLog;
        GLboolean compiled  = GL_FALSE;
        GLboolean linked    = GL_FALSE;
        GLint     length, maxLength;

        //
        // Create shader and program objects.
        //
        ProgramObject      = glCreateProgramObjectARB();
        VertexShaderObject = glCreateShaderObjectARB(GL_VERTEX_SHADER_ARB);

        length = strlen(vertexShader);
        glShaderSourceARB(VertexShaderObject, 1, &vertexShader, &length);

        //
        // OpenGL made a copy of the shader, we can free our copy
        //
        free(vertexShader);

        //
        // Compile the vertex shader, and print out the compiler log file.
        //
        glCompileShaderARB(VertexShaderObject);
        glGetObjectParameterivARB(VertexShaderObject,
                                  GL_OBJECT_COMPILE_STATUS_ARB, &compiled);

        glGetObjectParameterivARB(VertexShaderObject,
                                 GL_OBJECT_INFO_LOG_LENGTH_ARB, &maxLength);
        pInfoLog = (GLcharARB *) malloc(maxLength * sizeof(GLcharARB));
        glGetInfoLogARB(VertexShaderObject, maxLength, &length, pInfoLog);

        printf("%s\n", pInfoLog);
        free(pInfoLog);

        if (!compiled) {
            printf("Compile failed\n");
            return GL_FALSE;
        }

        //
        // Populate the program object with the compiled shader
        //
        glAttachObjectARB(ProgramObject, VertexShaderObject);

        //
        // We want the shader object to go away as soon as it is detached
        // from the program object it is attached to. We can call delete now
        // to achieve that. Note that calling delete on a program object
        // will result in all shaders attached to that program object to be
        // detached. If delete has been called for the shader object,
        // calling delete on the program object will result in the shader
        // object being deleted as well.
        //
        glDeleteObjectARB(VertexShaderObject);

        //
        // Bind user attributes before linking the program object.
        //
        glBindAttribLocationARB(ProgramObject, 1, "myFirstAttribute");
        glBindAttribLocationARB(ProgramObject, 2, "mySecondAttribute");

        //
        // Link the program and print out the linker log file
        //
        glLinkProgramARB(ProgramObject);
        glGetObjectParameterivARB(ProgramObject, GL_OBJECT_LINK_STATUS_ARB,
                                  &linked);

        glGetObjectParameterivARB(ProgramObject,
                                 GL_OBJECT_INFO_LOG_LENGTH_ARB, &maxLength);
        pInfoLog = (GLcharARB *) malloc(maxLength * sizeof(GLcharARB));
        glGetInfoLogARB(ProgramObject, maxLength, &length, pInfoLog);

        printf("%s\n", pInfoLog);
        free(pInfoLog);

        //
        // If all went well, make the ProgramObject part of the current state
        //
        if (linked) {
            glUseProgramObjectARB(ProgramObject);
            return GL_TRUE;
        } else {
            printf("Link failed\n");
            return GL_FALSE;
        }
    }

    GLhandleARB  ProgramObject;
    GLhandleARB  VertexShaderObject;

    void main()
    {
        GLcharARB *vertexShader; // Holds source for the shader
        int       location;      // Uniform location
        float     uniformData[3] = {0.0, 1.0. 2.0};
        float     vertexData1[3] = {0.0, 1.0. 2.0};
        float     vertexData2[3] = {0.0, 1.0. 2.0};

        //
        // Fill in the source in vertexShader
        //
        .
        .

        //
        // Create shader and program objects, setup generic attributes,
        // compile the shader object and link the program object and make it
        // part of the current state.
        //
        init(vertexShader);

        //
        // Get the location of someUniformName in the vertex shader, then
        // load its data.
        //
        location = glGetUniformLocationARB(ProgramObject, "someUniformName")
        glUniform3fvARB(location, 1, &uniformData[0]);

        //
        // Render
        //
        glBegin(...);
        .
        .
        glVertexAttrib3fvARB(1, vertexData1);
        glVertexAttrib3fvARB(2, vertexData2);
        .
        .
        glEnd();
    }

Revision History

    Revision: 0.5 6/5/2002
      - First draft for circulation
    Revision: 0.51
      - Updated API entry points in the New Procedures and Functions section
      - Named and explained all parameters to all API calls
      - BindAttributeLocationGL2 and GetActiveAttributesGL2 now have a
        length argument for their strings
      - Updated the section that describes state required to support vertex
        specification
      - Changed how standard attribute values are preserved across program
        changes (always preserved)
      - Changed language describing what happens when too few or too many
        attributes are provided
      - Updated the list of OpenGL stages that are and are not replaced by a
        vertex shader
      - Added suggestion for what new vertex array functionality to include
      - Added content to the errors section
      - Expanded the issues section up to and including number 4.
    Revision: 0.52
      - Generic vertex attributes, except for vertex arrays, are loaded
        using the exact same API commands as ARB_vertex_program uses. Added
        discussion of this to the issues section.
      - Changed the vertex array section to be TBD.
    Revision: 0.53
      - Now references V1.0 of the OpenGL Shading Language Specification
      - Fixed minor typos
      - Accepted by the GL2 working group
    Revision: 0.54 10/17/2002
      - Removed IsObjectGL2().
      - Expanded GetObjectParameter{if}GL2. Added OBJECT_SUBTYPE_GL2.
    Revision: 0.55 10/18/2002
      - The index parameter to BindAttributeLocation is now a uint instead
        of a ubyte.
      - Updated Errors section.
      - Added language disallowing mixing GL2 high level vertex shader with
        low level ARB vertex program or fragment program.
      - Documented interaction with ARB_vertex_blend
    Revision: 0.56 10/18/2002
      - Added vertex array interface, and added an issue discussing it.
      - Changed re-using ARB_vertex_program VertexAttrib*ARB commands issue.
    Revision: 0.57 10/21/2002
      - Changed VA interface to be exactly the same as ARB_vertex_program
      - Replaced MAX_ACTIVE_ATTRIBUTES_GL2 with MAX_VERTEX_ATTRIBS_ARB.
      - Went back to a different set of VertexAttribute*GL2 calls (instead
        of re-using the ARB_vertex_program ones).
      - Updated issues 2 and 3.
      - Added code to the sample usage section.
    Revision: 0.58 10/23/2002
      - Spelled out GetObjectParameter, instead of using 'T value' in its
        parameter list.
      - Fixed usage of const keyword in GetAttributeLocations.
      - Added GetVertexAttribute{ifd}vGL2
      - Added GetVertexAttribPointervARB
      - Updated interactions with ARB_vertex_program section
      - Updated errors section.
      - Moved shader resources section to GL2_shader_objects.
      - Updated chapter 5, Display Lists.
    Revision: 0.6 10/29/2002
      - Fixed typos.
      - Changed Issue three to unresolved.
      - Added issue number four.
      - Second distribution to the GL2 working group.
      - Now written against the OpenGL 1.4 specification.
    Revision: 0.61 11/26/2002
      - Minimum for MAX_VARYING_FLOATS reduced from 40 to 32.
      - Allowed mixing of a high level GL2 vertex shader with a low level
        ARB_fragment_program shader.
    Revision: 0.62 12/04/2002
      - Resolved Issue 5 the same way as ARB_vertex_program. As a result,
        added language to section 5.1.
    Revision: 0.63 1/29/2003
      - Marked issues 2 and 4 as unresolved.
      - Updated list of contributors.
      - Added issue 6, 7, 8, 9.
      - Updated the 'state required' section of section 2.7 and 2.8.
      - Renamed 'user-defined attribute' to 'generic attribute' to match
        ARB_vertex_program terminology.
      - Added language to section 2.13.2 and 2.13.3 about material and
        ColorMaterial handling when a GL2 shader is active.
      - Added wording to section 3.10 for PointSize and Fog.
      - Added a two sided lighting enable to section 2.13.1.
      - Added wording to explain how raster pos works, section 2.12
      - Added and resolved issue 10. Added wording to section 3.9.
      - Added issue 11.
      - Third working group distribution.
    Revision: 0.64 2/03/2003
      - Added VERTEX_PROGRAM_TWO_SIDE_ARB to the new tokens section.
      - Resolved issue 6, 7, 8, 10, 11 and updated the spec to reflect this.
      - Added wording to section 2.11, clipping
      - Updated the New State section
      - Updated the New Tokens section
      - Changed the VertexAttribute*GL2 entry points to VertexAttrib*ARB,
        which are the same entry points as defined in ARB_vertex_program.
      - Made generic attribute zero provoke a vertex
      - Resolved issue 2,3 and 4. Edited the interactions with
        ARB_vertex_program section to reflect this.
      - Consistently used 'conventional attributes' instead of 'standard
        attributes'.
      - Tried to clarify the wording in section 2.15.2 regarding attribute
        tracking.
      - Updated the Errors section
      - Added issue 12, texture access
      - Fourth working group distribution.
    Revision: 0.65 2/10/2003
      - Updated section 5.4, commands excluded from a display list.
      - Added section 2.15.6, position invariance. Renamed the old section
        2.15.6 to 2.15.7.
      - Added a discussion section to issue 1, invariance
      - Updated Appendix A, invariance
      - Added a proposed resolution to issue 12. Related to that are the new
        issues 13, 14 and 15.
      - Added language to sections 2.7, 2.8, 2.10.2 and 6.1.2 to support
        texture access.
      - Added a new section 2.15.5 to support texture access.
      - Added MAX_FRAGMENT_TEXTURE_UNITS_GL2, MAX_VERTEX_TEXTURE_UNITS_GL2
        and MAX_TEXTURE_IMAGE_UNITS_GL2.
      - Fifth working group distribution
    Revision: 0.66 2/11/2003
      - Added issue 16. Position invariance with respect to user clipping
        and closed it.
      - Closed issue 1.
      - Updated 'implementation dependent state' section.
      - Renamed MAX_VERTEX_UNIFORM_FLOATS_GL2 to
        MAX_VERTEX_UNIFORM_WORDS_GL2.
      - Consistently used 'texture coordinate set' instead of 'texture
        coordinate unit'.
      - Distributed to HP
    Revision: 0.67 2/13/2003
      - Added wording about interaction with ARB_matrix_palette. Same
        applies as the interaction with ARB_vertex_blend.
      - Sixth working group distribution
    Revision: 0.68 2/20/2003
      - Renamed MAX_FRAGMENT_TEXTURE_UNITS_GL2 to
        MAX_TEXTURE_IMAGE_UNITS_ARB
      - Renamed MAX_VERTEX_TEXTURE_UNITS_GL2 to
        MAX_VERTEX_TEXTURE_IMAGE_UNITS_GL2
      - Renamed MAX_TEXTURE_IMAGE_UNITS_GL2 to
        MAX_COMBINED_TEXTURE_IMAGE_UNITS_GL2
      - Added wording to issue 12) explaining how LOD calculation works.
      - Updated section 2.15.5 with respect to LOD calculations.
      - Updated section 2.15.5. Accessing the same texture image unit from a
        vertex shader and fragment processing counts as using two image
        units against the limit MAX_COMBINED_TEXTURE_IMAGE_UNITS_GL2.
      - Updated the interactions with ARB_vertex and fragment_program
        section.
      - Added the 'interactions with ARB_fragment_program' section.
    Revision: 0.69 2/27/2003
      - Updated the interactions with ARB_vertex and fragment_program
        section, again.
      - Updated the texture access section 2.15.5. It is more carefully
        worded and references the relevant texturing sections in chapter 3.
      - Closed issues 2, 3 and 4. Attribute handling.
      - Closed issue 9 and modified section 3.3 to reflect the resolution.
      - Updated texture coordinates to full citizens with respect to their
        dynamic range. Section 2.1.1. ARB_fragment_program already does
        this.
    Revision: 0.70 3/3/2003
      - Treatment of PointSize wasn't complete with respect to point
        attenuation and point fade. Section 3.3.
      - Added issue 17 explaining what happens to point distance attenuation
        and point fade.
      - Texture access now takes the texture object and texture unit bias
        into account as well. See sections 2.15.5 and Section 3.8.8.
      - Fixed numerous small typos and tried to clarify wording here and
        there.
      - It is no longer a requirement to load and/or compile a vertex shader
        before specifying its attribute bindings.
      - Added section 2.15.8 Required State.
      - Rewrote the overview section.
    Revision: 0.71 3/3/2003
      - Upgraded colors to be full floating point citizens, section 2.1.1.
      - Added language about undefined results if a vertex shader does not
        clamp when fixed-function fragment processing is used or an
        ARB_fragment_program target is active. Section 2.15.7.
      - Added issue number 18, explaining color clamping and conversion to
        fixed-point.
    Revision: 0.72 3/31/2003
      - Updated PointSize. Its value is undefined unless written by the
        shader.
      - Changed issue 9. We will have a point size enable after all.
      - Updated IP Status section to point to the contributors agreement.
      - Renamed BindAttributeLocation to BindAttribLocation to be in line
        with the rest of the vertex attribute naming (VertexAttrib etc).
      - Renamed GetActiveAttributes to GetActiveAttribs for the same reason.
    Revision: 0.73 4/9/2003
      - Added to the example usage section. Shows how to use
        GetActiveAttribs.
      - GetActiveAttribs no longer returns the length of the strings. They
        are always null terminated, a length parameter is no longer needed.
      - GetActiveAttribs no longer returns a Boolean. It was never set.
      - Added life time to the pointer returned by GetActiveAttribs.
      - GetActiveAttribs now also returns the type of each attribute.
      - It is no longer a link error if generic attributes are not bound yet
        using BindAttribute.
      - Renamed "handle" to "handleGL2".
      - Updated section 2.15.8
    Revision: 0.74 4/17/2003
      - Added issues 19, 20 and 21.
      - Re-ordered section 2.15.2. Vertex Attributes
      - The GL will assign an index for an attribute variable in a vertex
        shader. This index is queriable using the new GetAttribLocationGL2
        command.
      - Clarified BindAttribLocationGL2. It can override any binding
        assigned by the GL. It can be called at any time. Binding takes
        effect once a program object is (re)linked.
      - Added text that clarifies how to load and bind attributes declared
        as a matrix to section 2.7 and 2.15.2.
      - Clarified issues 1, 7, 8, 9, 11, 18 based on Pat's feedback. I
        didn't change their resolution, just tried to make them clearer.
      - Numerous other small clarifications based on Pat's feedback.
    Revision: 0.75 4/18/2003
      - Made BindAttribLocationGL2 display-listable.
      - Query of active attributes is still allowed after a failed link.
        After all the link could fail because there were too many active
        attributes.
      - Added a size parameter to GetActiveAttribsGL2.
      - Clarified when the linker can generate errors based on incorrect
        user bindings of attributes.
      - More clarification's based on Pat's feedback, most of them in
        section 2.15.2 Vertex Attributes.
      - Re-did section 2.13.1, two sided lighting with a vertex shader
        enabled.
      - Refined the section about aliasing attributes in section 2.15.2.
      - Changed MAX_VERTEX_UNIFORM_WORDS_GL2 to
        MAX_VERTEX_UNIFORM_COMPONENTS_GL2
    Revision: 0.76 4/21/2003
      - Changed a few occurances of "char" to charGL2.
      - GetAttribLocation and BindAttribLocation no longer take a length,
        just like GetUniformLocation.
      - Updated the example section to reflect the latest API changes.
      - Changed the location of the Contributors License to a publicly
        accessible place on www.3dlabs.com.
      - Version approved by the ARB-GL2 working group.
    Revision: 0.77 5/09/2003
      - Fixed issues 17 and 19 where some language was out of sync with the
        actual resolution in the spec.
      - Change 'zero' to NULL in the text describing GetActiveAttribsGL2.
      - Fixed typo in one of the comments in the example section and in the
        revision history.
      - Removed the VertexAttribMatrix*GL2 entry points from the 'new
        procedures and functions' section. These were never intended to be
        part of the spec.
      - Added Kent Lin to the list of contributors.
      - Added state tables. Not all state defined in the spec was present in
        the tables.
      - Section 2.15.8, required state for program objects was incorrect.
      - Clarified the lifetime of the pointers returned by
        GetActiveAttribsGL2.
      - Clarified that you need to pass in three NULL pointers to
        GetActiveAttribsGL2, to only get the count of active attributes.
      - Listed (already defined in the Errors section) error conditions to
        GetActiveUniformsGL2, GetAttribLocationGL2 and
        BindAttribLocationGL2.
      - Updated example section to reflect the latest API names.
      - The first error rule should have been INVALID_VALUE (consistent with
        GL2_shader_objects).
      - Added Jon Leech, Evan Hart, Benjamin Lipchak and Jeremy Sandmel to
        the list of contributors.
      - Removed the two rules from Appendix A (invariance) and updated Issue
        1 to reflect this. Added the same repeatability rule as the fragment
        shader.
      - Assigned enum values.
    Revision: 0.78 5/15/2003
      - Deleted reference to AppendShader (this call no longer exist).
      - Renamed all GL2 occurrences to ARB.
      - Added SHADER_OBJECT_ARB to the list of new tokens.
      - Fixed typos.
      - Added to the contributors list. If anyone is still left out, let
        Bill Licea Kane, Jon Leech or myself know.
    Revision: 0.79 6/24/2003
      - Changed MAX_VERTEX_TEXTURE_IMAGE_UNITS minimum value to 0 (from 1).
        Updated issue 15 to reflect this.
      - Updates issue 18 (color clamping). As a result, updated Section
        2.1.1 and 2.15.7.
      - Changed GetActiveAttrib to no longer return a pointer to GL memory.
        This call now returns data for one active attribute, instead of all
        active attributes. As a results, added error conditions to the
        Errors section.
      - BindAttribLocationARB is not display-listable.
      - Added OBJECT_ACTIVE_ATTRIBUTE_MAX_LENGTH_ARB and
        OBJECT_ACTIVE_ATTRIBUTES_ARB as a queriable parameter to
        GetObjectParameter{if}vARB. As a result, changed the New State
        section, and the Required State section.
      - Ordered the Errors section in the order of API commands described in
        the spec. Also split errors in one error per API command.
      - Updated the examples section.
      - General language cleanup.
      - ARB approved version.
    Revision: 0.80 3/16/2004
      - Section 2.15.5 changed that zero or more, and not one or more,
        texture image units are available to a vertex shader.
      - Section 2.15.5 Added language (copied from the fragment shader spec)
        about samplers and depth textures and undefined behavior. Also
        copied the language that specifies what happens when a texture
        object is incomplete. Also copied the language that specifies what
        stays the same with texture operations.
      - Added ARB extension number (31) and edited the status paragraph.
      - Removed SHADER_OBJECT_ARB, OBJECT_TYPE_ARB and OBJECT_SUBTYPE_ARB
        tokens from the New Tokens section. They are already defined in
        ARB_shader_objects. Also removed references to these in Chapter 6,
        GetObjectParameter{fi}v and the Errors section.
      - Changed MAX_COMBINED_TEXTURE_IMAGE_UNITS_ARB to 2 (from 3).
      - Changed a typo: OBJECT_ACTIVE_ATTRIBUTES_MAX_LENGTH_ARB should be
        OBJECT_ACTIVE_ATTRIBUTE_MAX_LENGTH_ARB (without the 'S')
      - Specified that the return parameters for GetACtiveAttribARB are
        unmodified when a GL error occurs.
      - Section 2.13.8. Varying variables, not generic vertex attributes, are
        clipped.
      - GetAttribLocationARB() in the 'New Procedures and Functions' section
        takes a handleARB not a handle.
      - Explained the meaning of the parameters to GetVertexAttrib*ARB.
      - Added that the current values of secondary color, fog coord and generic
        attributes are undefined after a call to DrawArrays or DrawElements, if
        the respective arrays are enabled.
      - Added resolution 22.
      - Added a paragraph to section 2.15.2 that makes it a link error to have
        both gl_Vertex an a generic attribute bound to index zero be used in a
        vertex shader.
    Revision: 0.81 3/30/2004
      - Removed ARB from DEPTH_COMPONENT_ARB
      - GetObjectParameter{if}vARB instead of GetObjectParameter{fi}vARB.
      - Updated the version of the Shading Language document referenced to
        100.53
      - Updated description of MAX_VERTEX_UNIFORM_COMPONENTS_ARB in the
        state tables.
      - Moved spec language related to texture operations not affected by a
        vertex shader to issue 23.
      - ARB-GL2 workgroup approved version.
    Revision 0.82, 12/12/2006
      - Fixed spelling of glGetObjectParameterivARB in sample code.
    Revision 0.83  12/14/2009
      - Added GLX protocol.