[Public WebGL] WebGL2 and no mapBuffer/mapBufferRange

Florian Bösch [email protected]
Sun Mar 15 05:11:47 PDT 2015


Having discussed this feature now thoroughly I will propose a resolution:

   1. mabBuffer[Range] and all its flags can be safely exposed
   2. It should be done to keep ES3 <-> WebGL 2 friction as low as possible
   and to avoid standards bifurcation
   3. It will require a bit more engineering (copying threads, read/write
   only array buffers) than is customary for GL features
   4. The performance might be better than buffer[Sub]Data, but it'll not
   be worse.
   5. All corner cases where the specification leaves the behavior
   undefined can be upgraded to GL errors

Can everybody involved agree to this resolution and can we hence put that
feature back into the WebGL 2 specification where it belongs if you want to
achieve ES 3 implementation certification?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://khronos.org/pipermail/public_webgl_khronos.org/attachments/20150315/39ae6b9c/attachment.html>


More information about the public_webgl mailing list