Re: HTML.next and Rechartering

On Thu, Aug 11, 2011 at 7:20 PM, Silvia Pfeiffer
<silviapfeiffer1@gmail.com> wrote:
> This is where it becomes problematic: if you call the next version of
> HTML "HTML5", then you can't release it as a complete spec unless you
> have many of the still immature features ready. For example, I would
> say that the multitrack specification part of HTML5 video is still
> very immature and should not move to "CR". But we cannot take it out
> of HTML5 because it satisfies some core requirements for HTML5 video
> accessibility.

The usual way this is handled would be to say something like "HTML5
does not define any mechanism for authors to include accessibility
information for media elements, but it is expected that HTML6 will do
so."  This lets the spec get to REC, but also points readers to the
information they need (we can have a link to HTML6 there as an
informative reference).  There were a bunch of things in CSS2.1 that
were changed to be undefined so that they could get it to REC.  I
think this is all silly, but it does work after a fashion.

The bigger problem is when you have the exact same feature in HTML5
and HTML6, but they have different conformance requirements for
whatever reason.  Then people are going to get confused.  Hopefully
this will not be a huge issue, because anything that winds up with a
full test suite and two interoperable implementations should be very
stable anyway.  This will probably wind up being only a small fraction
of what we now call HTML5 -- CSS2.1 is vastly smaller than HTML and
took ten years to get to that point.

I would hope that whatever falls out, we would have prominent and
permanent notices in HTML5 advising people to check HTML6 for more
up-to-date but less stable requirements, similar to the notices we
have now in WDs.

Received on Friday, 12 August 2011 16:10:17 UTC