[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Public WebGL] Why does the set of pname arguments for which the behavior of getProgramParameter is defined not match GL ES?



On Wed, Apr 18, 2012 at 1:42 AM, Cedric Vivier <cvivier@mozilla.com> wrote:
>
> Good catch.
>
> On Wed, Apr 18, 2012 at 04:43, Boris Zbarsky <bzbarsky@mit.edu> wrote:
>> At first glance, Gecko allows passing all 9 of the pname values defined in
>> GL ES (plus some CURRENT_PROGRAM value that's not in ES at all as far as I
>> can tell and should clearly go away).  Is the WebGL spec intending to
>> disallow passing INFO_LOG_LENGTH, ACTIVE_ATTRIBUTE_MAX_LENGTH, and
>> ACTIVE_UNIFORM_MAX_LENGTH?
>
> The removal of those enums in WebGL is specified in section 6.21.
>
>> If so, it might need to explicitly say so
>> somewhere....
>
> Yes, we should probably add something like "An INVALID_ENUM error is
> generated when the passed pname is not present in the table." to all
> pname-taking entry points in general (+ related conformance tests).

I don't think it's necessary to explicitly state that an INVALID_ENUM
error is generated. The WebGL spec delegates to the OpenGL ES 2.0 spec
for the majority of its behavior, and INVALID_ENUM is generated by
many of the APIs without explicit mention in the WebGL spec.

However, the conformance tests definitely should verify that the enums
that have been removed raise INVALID_ENUM. Cedric, would you be
willing to enhance the conformance tests in this area?
sdk/tests/conformance/programs/program-test.html looks like the best
place to put additional tests for getProgramParameter, and
sdk/tests/resources/desktop-gl-constants.js contains the enums that
have been removed.

-Ken

-----------------------------------------------------------
You are currently subscribed to public_webgl@khronos.org.
To unsubscribe, send an email to majordomo@khronos.org with
the following command in the body of your email:
unsubscribe public_webgl
-----------------------------------------------------------