Re: structural markup

William F. Hammond (hammond@csc.albany.edu)
Wed, 24 Jul 1996 12:18:52 -0400 (EDT)


From: "William F. Hammond" <hammond@csc.albany.edu>
Date: Wed, 24 Jul 1996 12:18:52 -0400 (EDT)
Message-Id: <199607241618.MAA07223@phoebe.math.albany.edu>
To: tmb@best.com
Subject: Re:  structural markup
Cc: www-html@w3.org

Thomas Breuel <tmb@best.com> writes:

:> William F. Hammond writes in <199607192237.SAA25716@phoebe.math.albany.edu>:
:>  5.  HTML is not really very "structural".  For example, the SENTENCE
:>      is not something that is recognized in HTML.  Does it make sense
:>      to make HTML-Math more structural than HTML overall?
:     
: Yes, it does.  
: 
: Math proposals need to be evaluated on their own merit, not relative
: to whether they are more or less structural than the rest of HTML.
: The purpose of structural information in HTML/SGML is to allow better
: indexing/searching, and to allow formatting across a wide range
: of devices.  On the other hand, requiring extra structural information
: makes it harder and much costlier to input and to convert
: existing documents.  You have to strike a balance; more structural
: markup isn't intrinsically better.

Yes, presentation targets must include audio and robot.

I wonder if we are ready to deal with searching.  Do we know how
search-strings for math will be constructed by users from the
different presentation formats?  Will GUI-marking be one means of user
construction?  Do we give the renderer the job of re-packaging math
search-strings in a universal format?

I take the phrase "HTML-Math" to refer to the math markup subset of
the March '95 W3C draft (expired) definition of HTML-3 and the ultimate
evolutes of that subset.  Without explicit indication to the contrary
I assume that anything proposed as "HTML-Math" is under consideration
for inclusion in the content-type "text/html".

It may be a mistake to propose extensions of "text/html" that do not
get picked up by the mass market renderers.  (I say "may" rather than
"will" because of the possibilities that might be opened with the
eventual evolution of "web browsers" into "web browsing component
collections" [cf., web-related IPC].)

There is lots of room for creating undergarments using other
content-types to support needs beyond presentation.

The crucial question here is what the widest audience is going to see.

                                   -- Bill


P.S.  Now what was this below?

:     Eventually, HTML will give way to full SGML, and use of vi(1) and
:     other non-GUI editors (for non-casual editing) will give way to
:     GUI, HTML/SGML-aware editors that will enable easier use of
:     structural markup.  

Someone else was being quoted here.  I agree completely with Thomas
on this.

: The fact remains that requiring more structural markup
: results in higher costs for input and conversion, relative to
: less structural markup.  Tools only affect the baseline.
: 
: There is no reason to believe that GUI tools will make
: structural input for math so easy that the cost becomes
: negligible; in fact, based on my experience with GUI and
: non-GUI tools for math input, if anything, the opposite
: is true.