Re: XML and SGML

[Rod Davison <rdavison@sprint.ca>]

I am dismayed that I have apparently been so unclear to some.  Since this 
question does come up in my classes, I suspect that there are probably a 
number of newcomers to this list, and mark-up in general,  who really could 
benefit from an authoritative reply.

I originally started to respond to Mr. Maden's points one by one, but I feel 
that rather than waste bandwidth with my comments... I'll let Charles 
Goldfarb summarize the situation.  The quote below is from the second edition 
of the XML Handbook, page xlv in the preface.  I hope it is enlightening. 

For those of you who do not know, Charles Goldfarb is the "father" of SGML 
and has been a key player in the mark-up world.

-------------------------------------------------------------

"Some claim that XML will replace SGML because there will be so much free and 
low-cost software.  Others assert that XML users, like HTML users before 
them, will discover that they need more of SGML and will eventually migrate 
to the full standard.

Both assertions are nonsense.. . XML and SGML don't even compete.

XML is a simplified subset of SGML.  The subsetting was optimized for the Web 
environment, which implies data-processing-oriented (rather than 
publishing-oriented), short life span (in fact, usually 
dynamically-generated) information.  The vast majority of XML documents will 
be created by computer programs and processed by other programs, then 
destroyed.  No human will ever see them.

Eliot Kimber, who was a member of both the XML and SGML standards committees, 
says:

There are certain use domains for which XML is simply not sufficient and 
where you need the additional features of SGML.  These applications tend to 
be very large scale and of long term; e.g., aircraft maintenance information, 
government regulations, power plant information, etc. Any one of them a 
larger volume of information than the entire use of XML on the web."

-------------------------------------------------------------------------------------------------------------------------

I hope this is useful to those of you who have been wondering.


Rod
-- 
====================================================
Rod Davison @ Critical Knowledge Systems Inc
rdavison@sprint.ca
====================================================

First things first -- but not necessarily in that order
====================================================