Re: Legacy elements (was : Complex Table Examples)

Philip & Le Khanh wrote:
> David Woolley wrote:
>
>> samp and kbd cover cases that are present in practically every user 
>> guide for any technological product.
> 
> But that is a very restricted universe of discourse.  Surely

I have a 9 inch pile of such documents, and I'd suggest that they are 
one of the most common sorts of document that people encounter.  Of 
those that people keep, the main other categories would be financial and 
legal, also rather specialist documentation, and often not even designed 
to be understandable by the user - at least user instructions have to 
attempt to do this.

> the element set of HTML should focus on the needs of /all/
> documents and not include (in its core set) elements of
> relevance only to one very small domain.  These elements

I disagree that it is a small domain (a lot of my web browsing is users 
instructions, as that is about the only place where you can get real 
facts about products - they are generally in PDF, but that is just part 
of the anomaly that the poor relation parts of marketing get to use the 
tool intended for up front marketing, and the advertising copywriters 
get to use the tool designed for supporting documentation!).

In any case, if they are not in HTML, supersets of them are needed in 
HTML, and preferably not such generic supersets as span.  HTML is 
supposed to provide basic capabilities, but support all document types; 
it's not supposed to be limited to the sorts of document that the man in 
the street writes, although it may not  be optimum for specialist documents.

> (which have existed since HTML's inception) serve only
> to remind us of the niche-like nerdy place that the web
> once was : now that the web is used by everyone, and not
> just by nerds, the element set should reflect this and not
> carry with it the intellectual baggage of its technological
> origins.

Ah! You mean span/div/a/img/script/embed.  That's what the average 
author wants!   In rejecting these semantic markup elements, I think 
you are actually supporting the presentationalist case.

var is certainly valid in almost any academic field, not just man 
machine interfaces, and man machine interfaces are a rather significant 
part of modern culture.

The other specific reason for retaining human computer interface 
concepts is that they are needed to document HTML itself.  Without them 
you cannot properly write the HTML specification in HTML!

Received on Sunday, 13 May 2007 11:33:20 UTC