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

Re: [Public WebGL] hint(GENERATE_MIPMAP_HINT, *) never positively tested?



This was my thought. I was not trying to say that we should test if the hints had any effect, but rather that we should check that valid inputs do not generate errors.

-Jeff

----- Original Message -----
From: "Kenneth Russell" <kbr@google.com>
To: "Patrick Baggett" <baggett.patrick@gmail.com>
Cc: "Jeff Gilbert" <jgilbert@mozilla.com>, "public webgl" <public_webgl@khronos.org>
Sent: Tuesday, April 3, 2012 11:48:40 AM
Subject: Re: [Public WebGL] hint(GENERATE_MIPMAP_HINT, *) never positively tested?

On Tue, Apr 3, 2012 at 6:57 AM, Patrick Baggett
<baggett.patrick@gmail.com> wrote:
>
>
> On Mon, Apr 2, 2012 at 7:33 PM, Jeff Gilbert <jgilbert@mozilla.com> wrote:
>>
>>
>> I am just fixing a bug where Firefox always generates INVALID_ENUM for
>> perfectly valid calls to hint(GENERATE_MIPMAP_HINT, [FASTEST, NICEST,
>> DONT_CARE]). The tests appear to check that invalid calls to hint() generate
>> INVALID_ENUM, but they must not be checking if it ever works. Since this is
>> a hint, and can always be silently ignored, should this be allowed to fail?
>>
> I don't see how you can reliably test that a hint "worked" (as in, has an
> effect on rendering), but it could test for INVALID_ENUM when given a symbol
> that isn't in its accepted set. Beyond that, no, there really shouldn't be
> any testing of the results.

Agreed that it's difficult or impossible to test whether the hint
succeeded, but incoming invalid enums to "hint" should generate
INVALID_ENUM errors. Incoming valid enums should not generate
INVALID_ENUM errors.

-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
-----------------------------------------------------------