Semantic Tagging in Web Objects (was: Comments on HTML+ Request For Comments)

George Forman - GHF (forman@cs.washington.edu)
Tue, 1 Feb 1994 10:29:38 -0800


> I think it would be cleaner and more flexible, and would preserve the
> focus of HTML+, to do semantic tagging with SGML tags from outside the
> HTML+ tag set. These semantic tags would be invisible to an HTML+
> browser, but would be known to a set of specialized indexing engines...

Steve,
You've got a good idea there-- having the set of semantic tags
separate from HTML so that they're effectively invisible. Then HTML browsers
don't have to be concerned with implementing these tags. Dave commented that
getting into all the semantic tags becomes a bottomless pit.

But on the other hand, I think it'd be really useful if the HTML
browser knew at least a few semantic tags. HTML already has the TITLE tag;
you wouldn't think of taking that out. I propose adding AUTHOR, ABSTRACT,
DATE, and VERSION. Why?

1) The browser uses this info to decide how to display things, just
as it does with <TITLE>.

2) Users may have personal preferences on how AUTHORS/ABSTRACTS are
displayed.

3) A better history list can be constructed if the HTML browser knows
these things about the document. I'd want a list of
documents I've read listing: title, author, abstract, and source.

4) Suppose you want to glance at a document without pulling all of it
across the Internet-- HTTP could send just these key elements.
This lets you look at the abstract before you decide to fetch the
whole thing.

So, I believe that HTML needs to specify at least a few tags so that
browsers can use the semantic information. Semantics aren't just for
automatic indexers.

We agree that orthogonality is important-- for this reason the ROLE
attribute shouldn't be coupled with <EM>phasis. I believe orthogonality can
be achieved even with certain semantic tags added to HTML+.

-GHF