Re: Why SGML?

[Charles Good <good -at- AUR -dot- ALCATEL -dot- COM>]

In a networked world full of different platforms, different operating
systems, and different documentation tool preferences, there is a need
for a document encoding methodology that is independent of platform, OS
and tool idiosyncracies. SGML can serve that purpose for companies that
want to embrace information product interchange. The U.S. government is
an example of an organization that receives documentation from vendors
around the world and everyone uses something different to develop and
publish the documentation. SGML is a way for the U.S. government to
eliminate storage and distribution problems associated with paper by
standardizing on an open systems form of electronic documentation.

Another consideration is the huge amounts of information most companies
and users are amassing which are not easily managed. SGML's Document
Type Definition (DTD) can specify uniform header information
requirements that can be easily searched by document management tools.
As a result, you can construct huge document libraries (databases) and
perform keyword searches for all documents matching the search phrase.

Since SGML is basically ASCII text, there is little overhead and file
sizes can be kept at a minimum. In addition, file compression routines
can work more efficiently. This results in a major savings in both
disk media and backup requirements (for disaster recovery preparedness).

Last but not least, SGML is wonderful in a multi output environment
where a single document may need to be presented in various forms
for the various customers. An SGML document might be formatted one
way for paper output, another way for CD-ROM ready output, another for
HTML WWW output, and another for online help. Large companies with
a diverse audience are addressing multiple output scenarios with SGML.

All of the above reasoning represents huge savings to companies if
they adopt SGML (and the related graphics standards). The only holdout
are the writers.  However, writers who grew up in UNIX cultures using
vi editors and paper formatting tagging schemes like nroff and troff
do not mind writing in SGML because it is a similar methodology to
what they are already using.

Usually, the writers who complain are those who like Word, Wordperfect,
Pagemaker, Framemaker, Interleaf and all the other GUI interface WYSIWYG
products on the publishing market. As a result, many of the larger tool
vendors developed SGML filters to import and export SGML from various tools.

In addition, some companies pioneered writer-assisted SGML authoring
tools (e.g.,  Datalogics, SoftQuad, Yard Systems, Software Exoterica).
These were usually pull-down menu approaches where the writer did not
need to remember the tags; simply pick from the list and the proper
tags (opening and closing) were inserted, then just type your text.

"Why SGML?"... Because the Information Systems and Production specialists
can do some amazing things with documentation maintained in this form.

"Why Write In SGML?"... I don't know that you need to provided your translation
filters keep pace with the information and publishing needs of your business.

"Why Learn SGML?"... It could provide future opportunities for you if you're
a contractor or freelance writer. Of course, there are more opportunities
right now if you know how to write in HTML ;-)