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

Re: [Public WebGL] Please make the ArrayBufferView constructor visible



Thanks for posting. I agree that the ArrayBufferView constructor
should be exposed. It is a longstanding design flaw that it was not.
The spec previously only documented implementations' behavior. Since
it seems there is widespread desire to make this change and no
negative feedback, it's been made. Apologies for how long it took to
take care of this.

The typed array editor's draft has been updated:

  http://www.khronos.org/registry/typedarray/specs/latest/
  svn diff -r20018:20019
https://cvs.khronos.org/svn/repos/registry/trunk/public/typedarray

The typed array conformance tests have been updated:

  https://github.com/KhronosGroup/WebGL/pull/120

Longstanding bugs against Chromium and Firefox have been reopened:

  https://code.google.com/p/chromium/issues/detail?id=60449
  https://bugzilla.mozilla.org/show_bug.cgi?id=796327

I've verified that in WebKit the changes to make this work properly
are trivial, and will file and fix a bug in that repository shortly.

Please review the changes above and send any comments to the list.

I'll also review the current ECMAScript 6 draft and file a bug there
if necessary to modify the spec to track this change. (ES6 subsumes
all of the current Typed Array types.)

-Ken



On Tue, Dec 11, 2012 at 9:09 AM, Boris Zbarsky <bzbarsky@mit.edu> wrote:
>
> On 12/11/12 11:58 AM, Victor Costan wrote:
>>
>> So it seems that the Firefox developers don't want to expose
>> ArrayBufferView?
>>
>> https://www.khronos.org/webgl/public-mailing-list/archives/1206/msg00254.html
>
>
> More precisely, one or two of the people working on the JS engine didn't.  I
> personally think we should, and we should further have some sane way of
> doing things like instanceof cross-global (though that's not really a typed
> array specific problem).
>
>
>> On the Firefox and Chrome bug trackers, developers cite the spec as
>> the main reason why they didn't expose ArrayBufferView.
>> https://code.google.com/p/chromium/issues/detail?id=60449
>> https://bugzilla.mozilla.org/show_bug.cgi?id=796327
>
>
> I think you read that wrong.  In the Gecko case it wasn't exposed for
> whatever reason, then the spec was changed to reflect that, then we're not
> going to randomly go violate the spec unless there's a prospect of it
> changing back.  I do think the spec is wrong here, fwiw, as is the
> SpiderMonkey typed array implementation...
>
>
> -Boris
>
> -----------------------------------------------------------
> 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
> -----------------------------------------------------------
>

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