Re: longdesc="" in HTML5

Hi Ian,

Ian Hickson wrote:
> I hate to throw fuel on the fire here, but as far as longdesc="" is 
> concerned, the reason it isn't in HTML5 is that there has never been any 
> feedback sent that described a problem for which longdesc="" was even 
> remotely considered as a solution.

Firstly, longdesc (in its inception and execution) is pretty much 
exclusively an element of the accessibility domain. Successfully 
designing mechanisms in this domain is a matter of combining engineering 
skills, usability analysis with real users and trying to successfully 
future proof the core nuts and bolts of any new technical specification.

Who in the HTML 5 WG has this delicate mix of skills? Very few if any. 
So we must rely on each other to share what we know and the skills we 
have in our own domains of expertise. This was also the case during the 
last iteration of the HTML specification. What we do have though are 
some first generation (sic) elements and attributes from the last 
specification that we can draw on. Some are flawed, poorly supported etc 
but they grew out of a genuine need that the specification authors tried 
to support. We can learn a lot from this - if we choose to.

Secondly, what about the need for a long descriptor? The authors of HTML 
4 saw a need for it. Has this need gone away?

User needs are not problems, they are specific requirements that need to 
be fulfilled. A long descriptor is a mechanism that satisfies the needs 
of certain users, in this case mostly blind and visually impaired users. 
The question is does longdesc suitably satisfy the users need for long 
descriptions and if not what will?

Be careful when crunching the numbers to research usage in the wild or 
relying on authors and developers to tell explicitly state what there 
problems or their user problems are. The HTML language is a part of a 
toolkit that the author will just use, if it has features that support 
accessibility fine, they may use them, if not they will just get on with 
it anyway doing what they can. We have the power now to ensure the 
toolkit is a quality one.  Poor element support coupled with authors 
ignorance of what to do with longdesc etc does not mean that there was 
no need for it to begin with. The question is will this genuine user 
need for a long descriptor be properly satisfied by longdesc, that is 
maybe better supported, or is there another mechanism that you think will?

Cheers

Josh

Received on Sunday, 7 September 2008 11:34:21 UTC