W3C home > Mailing lists > Public > public-web-perf@w3.org > June 2011

Re: ISSUE-7: FrameRequestCallback interface should be designated as Callback=FunctionOnly [Request Animation Frame]

From: Boris Zbarsky <bzbarsky@MIT.EDU>
Date: Wed, 01 Jun 2011 21:31:13 -0400
Message-ID: <4DE6E7E1.5010404@mit.edu>
To: public-web-perf@w3.org
On 6/1/11 6:04 PM, Cameron McCormack wrote:
> Web Performance Working Group Issue Tracker:
>> The change would look like so:
>>
>> [Callback=FunctionOnly, NoInterfaceObject]
>> interface FrameRequestCallback {
>>    void sample(in DOMTimeStamp time);
>> };
>
> I think that makes sense, and I’ve made the change to the spec.

I'd like us to think about that change for a minute.

Consider the case of a DOM with both ES and non-ES bindings, where he 
non-ES code wants to pass a FrameRequestCallback object to the ES code 
and have that ES code register the callback.  Or heck, ES code wants to 
pass a FrameRequestCallback to some other method implemented in a non-ES 
language which will then pass it back into ES which ends up registering it.

Is such a setup possible in the FunctionOnly case?  If not, do we really 
want to preclude it?  What are the benefits of using FunctionOnly?

-Boris
Received on Thursday, 2 June 2011 01:31:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 2 June 2011 01:31:44 GMT