Re: [WebIDL] Allowing specifications to opt in to throwing on sets of enumerated attributes to invalid values from Boris Zbarsky on 2012-11-15 (public-script-coord@w3.org from October to December 2012)

From: Boris Zbarsky <bzbarsky@MIT.EDU>
Date: Thu, 15 Nov 2012 08:26:20 -0800
To: Anne van Kesteren <annevk@annevk.nl>
CC: Marcos Caceres <w3c@marcosc.com>, "public-script-coord@w3.org" <public-script-coord@w3.org>
Message-ID: <50A517AC.5080102@mit.edu>

On 11/15/12 8:17 AM, Anne van Kesteren wrote:
> On Thu, Nov 15, 2012 at 8:03 AM, Marcos Caceres <w3c@marcosc.com> wrote:
>> If Version 1 switches to [strict], then you need the above check to catch Version 0, no?
>
> You wouldn't switch.
>
> It would be interesting to know though what the WebAudio scenario is.

For example, PannerNode currently has this:

     // Panning model
     const unsigned short EQUALPOWER = 0;
     const unsigned short HRTF = 1;
     const unsigned short SOUNDFIELD = 2;

     attribute unsigned short panningModel;

and the desire is that bogus assignments to panningModel throw instead 
of silently being ignored, as I understand.

There's similar stuff all over the spec ("type" on BiquadFilterNode and 
OscillatorNode are other examples, and there are lots more).

-Boris

Received on Thursday, 15 November 2012 16:26:57 UTC