Re: ACollab Work Groups 1.1 Content accessible?]

> > Creating valid HTML, in the sense that it passes the W3C Validator, is
> > a relatively mechanical process.  Creating semantically valid HTML
> > from a purely physical document is an AI problem.
> 
> I agree that creating valid HTML is a mechanical process...if you know 
> HTML. Most people don't; this why I believe commonly used tools must be 

I think the points I was trying to make here is that valid was the
wrong term to use, as it is normally understood to imply only syntactic
validity, and that non-human creation of truly accessible material from
a WYSIWYG marked up topic is still, to a large extent, a research topic.
(The best hope would be a reverse style sheet that told the program what
a particular physical markup really meant.)

You hint that the Adobe tagged PDF from Word tools are far from perfect,
but my guess, from what I've seen written, is that they use Word's
own object model and quite a lot of processing, to try their best.

As such, I don't think there is any real technological solution and 
one has to decide whether it is cheaper to educate the document
originators, or emply someone to, with the aid of tooling, clean up
the documents.

(The original aim of HTML was that it would be so simple to use that
anyone could use it with maybe half an hour of instruction, but there is
no money in such a tool, so commercialisation tried to convert a product
(Mosaic), developed with non-commercial funding, into a more visual art
related tool, rather than sell the original concept.)

Received on Saturday, 19 June 2004 05:14:11 UTC