Re: using Mosaic/WWW and ftp to distribute docs

Subject: Re: using Mosaic/WWW and ftp to distribute docs
From: Chet Ensign <Chet_Ensign%LDS -at- NOTES -dot- WORLDCOM -dot- COM>
Date: Wed, 21 Dec 1994 10:40:09 EDT

Interesting discussion on Mosaic. I'd like to throw in another thought.

The WWW model of document distribution is really a novel method of document
distribution, one that I think will become more and more important in the years
ahead. It is a client/server model of publishing information.

The standard model for distributing information has been based on what the
publisher could do, not necessairily what the customer wanted. Paper, the first
medium, was (and is) distributed in bound chunks. Various ways to reduce that
"granularity" have been developed -- ring binders, etc. But the customer, the
consumer, still basically gets it all.

The advent of electronic publishing didn't change the model. A set of books on
a CD-ROM is less expensive to produce and store, easier to distribute, faster
to search, but it is still the whole published thing.

Which may be fine for the consumer, or may be overkill Nobody ever really
thought to ask before because that's the way it was.

Enter the Web. Now we have a client/server model capable of deliverying
information in at a much more discreet level of granularity, and only on
demand. It is the ultimate in "on-demand publishing." Also, because it is
client/server, it is *not* product-dependent. The host of existing products
limit you to: a) a particular platform or set of supported platforms and b) a
particular vendor's idea of what constitutes a good data format. If you want to
use MS Viewer (and I do like that tool) you sacrifice Mac and UNIX and IBM
mainframe readers.

The client/server model skips the problem completely. Because the data format
is known, anybody can program readers to process it. Just look how fast Web
browsers have come into existence on Mac, Windows, OS/2 and UNIX machines!

The comment that HTML is a lame format is accurate but besides the point. HTML
is lame because its creators didn't even begin to envision the success it would
have. They were just cobbling together a mechanism for high-energy physicists
to share information on the Web. Tim Berners-Lee probably wakes up every
morning wondering how he wound up in the pages of Newsweek.

But now that it is successful, it is being extended. An even better approach is
the one being taken by Electronic Book Technologies in Providence RI. They have
developed a product called DynaWeb that lets you map elements from any SGML
source file to elements supported by HTML. When a document is requested across
the net, DynaWeb converts the datastream into HTML on the fly, as it feeds it
back to the requester. You don't have to write, store or convert docs into HTML
in order to make them available, and as HTML itself is extended (and browsers
extended to support it), you can change the output mapping.

This is the future. Let's enjoy it.


Chet Ensign
Director, Electronic Publishing
Logical Design Solutions, Inc.

Phone: 908-771-9221
Email: chet -at- lds -dot- com
Email (home): censign -at- interserv -dot- com

Previous by Author: HTML, WWW at the SGML Forum of NY
Next by Author: Re: Information needed
Previous by Thread: using Mosaic/WWW and ftp to distribute docs
Next by Thread: Computer Documentation Course Announcement

What this post helpful? Share it with friends and colleagues:

Sponsored Ads