Re: ISSUE-179 av_param: Chairs Solicit Proposals

As I pointed out in the change proposal rationale, x-* attributes are
unacceptable for use by other standards fora (and W3C WGs) for defining
specifications that involve additional A/V parameters. Such usage impedes
standardization.

Further, proposing specific parameters is not necessarily possible or
practical:

(1) the set of additional parameters is not fixed, and grows/changes over
time;
(2) even if specific parameters were identified, obtaining consensus on
their standardization as new attributes may not be feasible;

The bottom line is that object supports param, and that the intent of A/V
elements was to supplant (and extend) object semantics. To do this in a
practical, backward compatible manner requires param or an effective,
standards compliant equivalent.

It has been suggested that data-* or x-* may serve this purpose. I have
shown how data-* does not unless the current prohibitions in HTML5 are
removed. That leaves only x-*, which is anathema to any standards based
extensions.

Anne, regarding your comment "counter to the whole exercise", could you
elaborate what you mean by that? Also, you seem to use the term "loophole"
in a possible negative fashion. I view param as an extensibility mechanism.
If HTML5 were solely based on XHTML syntax, I would not be making this
proposal, since it would be easy to add attributes in another XML
namespace. However, the emphasis on HTML syntax by the HTML5 community has
reduced the number of extensibility options, making param a viable and
necessary feature for A/V elements as well as object.

Regards,
Glenn

On Sat, Oct 29, 2011 at 4:28 PM, Anne van Kesteren <annevk@opera.com> wrote:

> On Sat, 29 Oct 2011 09:27:13 +0200, Silvia Pfeiffer <
> silviapfeiffer1@gmail.com> wrote:
>
>> That's ok. If you need it for UA interpretation, use x-* attributes or
>> standardize them if all browsers should support them, as the counter
>> change proposal says.
>>
>
> Indeed, making arbitrary extensions to <video> and <audio> possible via a
> <param> loophole goes counter to the whole exercise.
>
>
> --
> Anne van Kesteren
> http://annevankesteren.nl/
>