[Public WebGL] Requirements for bufferData don't make sense given the IDL

Tim Johansson [email protected]
Mon Apr 2 23:37:33 PDT 2012

On 2012-04-02 20:01, Gregg Tavares (勤) wrote:
> On Mon, Apr 2, 2012 at 7:22 AM, Boris Zbarsky <[email protected] 
> <mailto:[email protected]>> wrote:
>     In particular, it says:
>      If the passed data is null then an INVALID_VALUE error is generated.
>     But the IDL declarations for this method don't allow |data| to be
>     null; an attempt to pass null will throw during the WebIDL
>     overload resolution algorithm.  Was this requirement meant to be
>     on bufferSubData (which does allow null to be passed)?  Or is
>     bufferData supposed to have an overload that allows null?
> bufferData should (and does) take null
> bufferSubData does not take null
I agree that is how it should be, but last time I checked you cannot 
pass the conformance test unless bufferSubData accepts null. I just 
tried to update the spec to match what the conformance tests require, 
but this is something I think everyone agrees should change.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://khronos.org/pipermail/public_webgl_khronos.org/attachments/20120403/b405e247/attachment.html>

More information about the public_webgl mailing list