Agencies in need of media-neutral docs should try out SGML

Are you ready to shift your World Wide Web site into neutral?

The growing popularity of ''media-neutral'' publishing systems is beginning to have a
big impact on the way documents are created for storage on Web servers. The idea: Store
all your agency's documents in one format in a back-end document management system. Then
retrieve them over your intranet for easy updates, and convert them on the fly for Web or
CD-ROM publication.

If that sounds familiar, it's probably because you remember similar claims made several
years ago for the Standard Generalized Markup Language. SGML's Web subset, the Hypertext
Markup Language, took another path when browser builders kept adding their own extensions
for ever-fancier pages. Now the differences between SGML and HTML further complicate
document conversion.

Another thing that's slowed Web document management is that most products in this
category are proprietary and don't mesh well with the Internet's open systems orientation.
But as Web sites grow, system developers and resellers are working to adapt existing
products for Web use.

Most use either HTML or SGML as the primary format for storing documents. But if
formatting is kept simple, you can generate basic Web pages from SGML with few problems.
Some streamlined SGML documents can even be read directly by Web browsers with no

The real promise of an SGML-based system is that you can store a single copy of a
document and run it through conversion filters whenever requested. If you publish to
several media, just build a variety of filters. If you publish mostly to the Web and
spread the files across two or three servers with different addresses and pointers, even
that can be built into filters.

This on-the-fly publishing will protect agencies from having to hand-edit documents if
they later need to update a site.

The most Web-savvy document management systems store things in a Structured Query
Language (SQL) or object-oriented database. If you have a choice, go for object-oriented.
It's potentially better in the long run, because it's easier to integrate into an
all-encompassing system--an ''editor's workstation''with version tracking and integrated
tools to control document rights and publishing filters.

To get an idea of how all this works, look at the DynaText and DynaWeb products from
Electronic Book Technologies of Providence, R.I., recently acquired by Inso Corp.

DynaWeb has its own scripting language, customizable templates for translations and
configuration controls.

Visit to learn more.

Documentum Inc of Pleasanton, Calif., offers RightSite, a Web-oriented extension of its
Documentum Enterprise Document Management System. It delivers only the information
relevant to, and authorized for,

a particular user. It will filter and assemble pages based on the user's choice of
topic, format and document status. Details appear at

Other Web-savvy document management systems:

Most of these packages have the right management power, but they aren't
point-and-click. To get to that level, you'll need a real Web whiz with solid SGML

Someone who's now implementing one of these systems told me he plans to start slow with
a single directory holding about 100 documents. He'll roll the management system out
slowly and recruit end users for testing. That conservative approach should reveal the
system's quirks before he ramps up.

If you think Web-based document management is for you, first learn about document
management itself. That will help you understand how to adapt it to your Web site.

For a short definition, visit

html. For more detail, check out Interleaf Inc.'s magnum opus at

Shawn P. McCarthy is a computer journalist, webmaster and Internet programmer for GCN's
parent, Cahners Publishing Co. E-mail him at [email protected]

Stay Connected

Sign up for our newsletter.

I agree to this site's Privacy Policy.