SGML for Web Pages

The Big Glee Bopper thom at indiana.edu
Mon Dec 18 19:58:00 EST 1995


On Mon, 18 Dec 1995, Keith Engwall wrote:

> Well, actually, format is the least complicated aspect of electronic
> storage.  It's relatively trivial to create an interpreter that will
> translate from one format to whatever's current.  A much larger issue is
> how do we store this information?  Ink on paper has a certain amount of
> permanence, depending on the type of ink and paper.  As long as it will
> last (ie. acid-free paper, non-water-soluble ink, etc.), we can read it!

There are a couple of issues. One is how do _we_ store information and 
the other is how do _they out there_ want to store information. SGML 
sounds great but I doubt if Melville would have been all that interested 
in writing in either sgml or html. There is a site which checks for html 
valid documents. The highest number of valid documents out there is 3%. 
The problem is that the folks writing html documents want to use the 
language not as a document structuring language but as a display 
language. We can piss and moan about information retrieval issues but 
this is a natural trend which will not change by pushing SGML on folks. I 
think large organizations such as Boeing have an interest in all their 
documentation SGML-d but that is controlled info. Might be that documents 
will be assembled in a variety of files with header info: sgml, html, 
futureDTD, freeForm, etc. I wouldn't bet on files of homogenized info 
either. Document level maybe.

I think the physical medium is not important as long as there is serious 
information redundancy. Don't put all your eggs in one basket, comes to 
mind 8-)

--Thom


More information about the Web4lib mailing list