Re: HTML 5 (sic) and A11y

On Mon, 24 Jan 2011 12:36:34 +0100, Eihab Ibrahim <eihabibrahim@gmail.com>  
wrote:

> On 1/24/2011 1:28 AM, Joshue O Connor wrote:
>> Hi y'all,
>>
>> I just saw Ian Hicksons post on the WHATWG Blog where the HTML 5
>> "technology is not versioned and instead we just have a living document
>> that defines the technology as it evolves".
> Hi Josh,
>
> That's a WHATWG decision and does not really affect the W3C. As far as I  
> know, we still call the next iteration "HTML 5" here.

Indeed. And publish snapshots...

>> What this move effectively means is that HTML (5) will be implemented in
>> a piecemeal manner, with vendors (browser manufacturers/AT makers etc)
>> cherry picking the parts that they want. It could be argued that this is
>> the way it _already_ is however as a specification isn't a movable feast
>> there is more chance for consistency and stability. This current move by
>> the WHATWG, will mean that discussions that have been going on about how
>> best to implement accessibility features in HTML 5 could well become
>> redundant, or unfinished or maybe never even implemented at all.
> Then follow the W3C's recommendation when it's finished and call your  
> documents or engine W3C HTML5 compliant.

Right. IF you want to follow a standard with a consensus process, patent  
policy that gives some legal teeth to Open Web rhetoric, and lets you  
state clearly waht you are comlying to you have W3C. If you want to follow  
a bunch of ideas that just keep moving, WHAT-WG is one of the places you  
can get that in the context of HTML. They do fundamentally different  
things, and both are important.

>> What is implemented will be dependent on where the "living
>> specification" is at any point in time. If I am flying in a plane, I
>> don't want to know the engineers were still not in agreement about how
>> to design the engines after the plane has taken off.

True. On the other hand, if I have Parkinson's disease I would rather try  
a slightly experimental treatment that is considered likely to have a  
greater chance of more success in managing it.

>> I think this will mean piecemeal implementation by vendors, with the
>> caveat that "the spec is in flux so we can only implement the most
>> stable parts of it" which is a perfect get out clause for a poor user
>> experience.
> The plane engine is not a good analogy in this situation. I have yet to  
> author a non-conforming document that seriously injured somebody.

Ah. I guess you never wrote documentation for machinery then.

> In all seriousness, as you have stated before that already is the case  
> even with a numbered version of the spec. Browsers implement the  
> features they deem appropriate and document authors write to the most  
> widely deployed parts of the standard.

Hmmm. At the bleeding edge, people make up ideas, test them out and refine  
them. That includes browsers adding shiny new features, authors doing  
things with them, people figuring out what is wrong with that and how to  
improve it. You *want* that stuff to be unstable - baking things into the  
Web too early has got us into as many problems as it has solved, or  
perhaps more.

At the trailing edge, pretty much anyone (except a seriously abusive  
monopolist or a totally wacko control freak) wants the <p> element to work  
the same way, so we work to make sure we understand even the wierd edge  
cases. Everyone wants that stuff to be stable, so we standardise it.

> The way I see it, the WHATWG is (or will become) a forum for the latest  
> developments in HTML, and the W3C will be the standards body you can  
> refer to when you need a specific snapshot in time.
>
>> This is a disappointing move, and a retrograde step that could well make
>> the fine tuning of important accessibility aspects of HTML (5) even  
>> harder.
> To the contrary, I think this is a great move. It will allow the WHATWG  
> to polish and iterate over features in a fast paced and somewhat formal  
> environment, while giving the W3C the opportunity to cherry pick widely  
> deployed features that have been tested for a while.
>
> I know that the WHATWG is still very much involved in the W3C's HTML 5  
> effort, and I doubt that this collaboration will cease to exist in  
> future revisions of the standard.

There is always a tension - different people have different priorities,  
and giving one group control over a W3C standard is generally a bad idea.  
On the other hand, making sure that it gets implemented - something that  
didn't happen to HTML4 but has been part of W3C process since shortly  
afterwards - is an important reality check for a standard.

Frankly, I am underwhelmed by the overall level of accessibility knowledge  
and practical suggestion I see coming from WHAT-WG. WHich is fine - they  
are not the only people involved in developing stuff for HTML - and some  
of the specific things they have done in accessibiltiy are IMHO extremely  
good (albeit others are not). If there were nobody but that one group  
making HTML5 I would indeed have the same concerns as you. But I don't  
think that is the case, so I am not really worried.

cheers

Chaals

-- 
Charles McCathieNevile  Opera Software, Standards Group
     je parle français -- hablo español -- jeg lærer norsk
http://my.opera.com/chaals       Try Opera: http://www.opera.com

Received on Monday, 24 January 2011 13:29:19 UTC