src/third_party/webgl/src/extensions/EXT_color_buffer_half_float/extension.xml

   1 <?xml version="1.0" encoding="UTF-8"?>
   2 <draft href="EXT_color_buffer_half_float/">
   3   <name>EXT_color_buffer_half_float</name>
   4
   5   <contact> <a href="https://www.khronos.org/webgl/public-mailing-list">WebGL
   6   working group</a> (public_webgl 'at' khronos.org) </contact>
   7
   8   <contributors>
   9     <contributor>Mark Callow, HI Corporation</contributor>
  10
  11     <contributor>Members of the WebGL working group</contributor>
  12   </contributors>
  13
  14   <number>14</number>
  15
  16   <depends>
  17     <api version="1.0"/>
  18
  19     <ext name="OES_texture_half_float" require="true"/>
  20   </depends>
  21
  22   <overview>
  23     <mirrors href="http://www.khronos.org/registry/gles/extensions/EXT/EXT_color_buffer_half_float.txt"
  24              name="EXT_color_buffer_half_float">
  25       <addendum>
  26         <p>All references to <code>R16F</code> and <code>RG16F</code> types
  27         are ignored.</p>
  28       </addendum>
  29     </mirrors>
  30
  31     <features>
  32       <feature>
  33         <p>The 16-bit floating-point types <code>RGB16F</code> and
  34         <code>RGBA16F</code> become available as color-renderable formats.
  35         Renderbuffers can be created in these formats. These and textures
  36         created with <code>type = HALF_FLOAT_OES</code>, which will have one
  37         of these internal formats, can be attached to framebuffer object color
  38         attachments for rendering.</p>
  39       </feature>
  40
  41       <feature>
  42         <p><span style="color: red">NOTE:</span> fragment shaders outputs
  43         gl_FragColor and gl_FragData[0] will only be clamped and converted
  44         when the color buffer is fixed-point and <code>BlendColor()</code> and
  45         <code>ClearColor()</code> will no longer clamp their parameter values
  46         on input. Clamping will be applied as necessary at draw time according
  47         to the type of color buffer in use.</p>
  48       </feature>
  49
  50       <feature>
  51         <p>The format and type combination <code>RGBA</code> and
  52         <code>FLOAT</code> becomes valid for reading from a floating-point
  53         rendering buffer. Note: <code>RGBA</code> and
  54         <code>UNSIGNED_BYTE</code> cannot be used for reading from a
  55         floating-point rendering buffer.</p>
  56       </feature>
  57
  58       <feature>
  59         <p>The component types of framebuffer object attachments can be
  60         queried.</p>
  61       </feature>
  62     </features>
  63   </overview>
  64
  65   <idl xml:space="preserve">module webgl {
  66   interface EXT_color_buffer_half_float {
  67     const GLenum RGBA16F_EXT = 0x881A;
  68     const GLenum RGB16F_EXT = 0x881B;
  69     const GLenum FRAMEBUFFER_ATTACHMENT_COMPONENT_TYPE_EXT = 0x8211;
  70     const GLenum UNSIGNED_NORMALIZED_EXT = 0x8C17;
  71   }; // interface EXT_color_buffer_half_float
  72 }; // module webgl</idl>
  73
  74   <additions>
  75     <p>In section 5.13.12 <cite>Reading back pixels</cite>, change the allowed
  76     format and types table to:</p>
  77
  78     <dl class="methods">
  79       <dd>
  80         <table>
  81           <tbody>
  82             <tr>
  83               <th>frame buffer type</th>
  84               <th>
  85                 <code>format</code>
  86               </th>
  87               <th>
  88                 <code>type</code>
  89               </th>
  90             </tr>
  91
  92             <tr>
  93               <td>normalized fixed-point</td>
  94               <td>RGBA</td>
  95               <td>UNSIGNED_BYTE</td>
  96             </tr>
  97
  98             <tr>
  99               <td>floating-point</td>
 100               <td>RGBA</td>
 101               <td>FLOAT</td>
 102             </tr>
 103
 104           </tbody>
 105         </table>
 106       </dd>
 107     </dl>
 108
 109     <p>Change the paragraph beginning "If <code>pixels</code> is null ..."
 110     to</p><blockquote>If frame buffer type is not that indicated in the table for
 111     the <code>format</code> and <code>type</code> combination, an
 112     INVALID_OPERATON error is generated. If pixels is null
 113     ...</blockquote>
 114   </additions>
 115
 116   <history>
 117     <revision date="2012/11/08">
 118       <change>Initial revision.</change>
 119     </revision>
 120
 121     <revision date="2012/11/13">
 122       <change>"Add reading-pixels-as-FLOAT feature to the Overview and related
 123       changes to WebGL section 5.13.12.</change>
 124     </revision>
 125
 126     <revision date="2012/11/26">
 127       <change>Move to draft.</change>
 128     </revision>
 129   </history>
 130 </draft>