Re: Ideas for CSS 3.0?

> The analogy is false because those structures are not presented in multiple
> media.  They are in the physical world in one instantiation period.

But the machine readable document is a model of that physical world (as
are most documents), and the information they contain is not just the
the list of couuntry names, but the (approximate) location of their 
borders.  There is a continuum here - maybe communicating the borders is
not particularly important, but an image map used to name the parts in 
an assembly may well contain a lot of information.  Even with countries,
they are often used to define sales territories, which may well not 
be strongly tied to movements in political boundaries, especially at
a regional level within a country.

> But most people at least care about two media, screen and print, and
> increasingly numbers of people are caring about handhelds, phones and other

The information is there, in client side image maps; it may be there in
rather low level geometrical form (but that is an objection to SVG and
Flash as well - and I'm afraid that if IE ever implements SVG natively,
it will be the end of any semblance of structural markup on most web
sites, to a much greater extent than is threatened by image maps).  If
image map information is not available in printed media, that is, to
a large extent the responsibility of the user agent.  It can even be
presented in aural media, although to be realistically usable, one needs
to serialise the description at a rather higher level (long descriptions
of the image, etc.).

I think, though, that my original comment was based on the proposal that
the image itself should be introduced as a background image in the 
styling - background images are for cosmetic effects, and, for example,
don't have long descriptions, or alternative text.

Received on Thursday, 5 June 2003 02:23:40 UTC