SGML: Report on the HyTime '96/SGML and The Web Conference
Subject: Report on the HyTime '96/SGML and The Web Conference
Date: Thu, 29 Aug 1996 16:48:48 -0500
From: "W. Eliot Kimber" <kimber@passage.com>
Newsgroup: comp.text.sgml
The GCA recently held two two-day conferences during "Information and
Technology Week", 18 Aug to 23 Aug. The first was two days of HyTime
madness followed by two days of SGML and the Web.
Below is my report on the conference, which includes a couple of
prognostications.
------------------------------------------------------------------
Executive Summary:
Intranets! Need SGML. Need HyTime. Need enabling architectures. Unisys
joins forces with TechnoTeacher; says standards for data representation
are key to its information management systems. In short: Corporate
America has realized that what the Web does is what they want but that
current Web approaches won't provide it.
New tools: Vicom Multimedia's multimedia development system, free IETM
brower, demonstration HyTime browser from TechnoTeacher. SoftQuad only
major SGML tool vendor represented; announces "Panorama-like" plug-in for
Internet Explorer.
No SGML database vendors in attendance. The general feeling expressed,
both by speakers and in the halls, was that this was very short sighted on
their parts.
There were about 80 people in attendance, about the same as last year.
------------------------------------------------------------------
The Details:
Most of the program concerned the HyTime Technical Corrigendum, the
alignment of HyTime with DSSSL, and the new concepts and implications that
come out of these efforts (groves and property sets).
The key things I came away with from the conference were:
1. Intranets will be very very big for SGML and HyTime. My prediction is
that within the next two years Corporate America will realize that only
SGML and HyTime enable the sort of long-term information structuring and
interoperation you need to realize the full potential benefits of
intranets.
The Web has shown people that internetworking can be easy and effective,
eliminating many of the scaling and maintenance problems that specialized
client-server systems presented. But they almost immediately realize the
limits of HTML and URL-based addressing. In short, it is patently obvious
to MIS types that HTML is insufficient to communicate the full richness of
business data to clients, both human and programatic. [One of the more
interesting talks was by Matt Fuchs on the use of SGML to communicate
information among agents in a network where the agents may be humans or
programs.]
Note that HyTime has to be part of the Intranet solution because intranets
involve the dynamic interlinking of heterogenous databases over networks
in a decentralized environment. The direct and file-based addressing
methods used with HTTP protocols (URLs, URNs, etc.) are simply not robust
enough to support this environment. In addition, the weak hyperlink
representation structures in HTML limit your ability to represent
databases of arbitrary relationships, a key part of decision support
systems, which should ideally be able to interlink any data anywhere in an
enterprise or "hyperprise" [I just coined this term: a set of
interconnected enterprises.]
Thus, I see demand for the following products and services increasing
sharply in the near term:
A. SGML- and HyTime-based end-user tools that are easier to use
and less expensive then today's crop. This includes both
client applications (browsers, editors, etc.) and servers,
such as databases, link managers, and so on.
B. Training on SGML and HyTime, both how to design for it and
how to implement support for it. In particular, MIS
professionals will need to understand the subtleties of
managing structured documents in a networked environment.
C. SGML- and HyTime-based integrated solutions for managing and
delivering information on intranets. Someone has to plug
these things together and make them work.
Note that no existing tool or suite built from existing tools can provide
the functions needed for intranets off the shelf. Big opportunity for the
company that can provide turn-key solutions and reliable custom
integrations.
2. DSSSL will be very important in all future SGML processing systems.
DSSSL provides a standard language for describing both style sheets and
transformations. It provides a processing abstraction layer that makes it
easy or possible to integrate different SGML and non-SGML applications
under a common configuration interface (where by "configuration" I mean
configuration of the presentation and processing applied to documents).
Customer demand for DSSSL support in browers, editors, composers, and
transformation systems will rise dramatically in the short term.
3. The use of HyTime hyperlinks to model abstract relationships has real
added value over current database and data modeling tools. This is
particularly important for decision support systems where you are
creating, managing, and analyzing relationships among data objects. Having
a robust standard for representing such relationships will be very
important, especially if I'm right about point 1.
4. Microsoft appears to be serious about supporting SGML in their
Web products (primarily Internet Explorer). I think Microsoft shares my
view of the future of Intranets. They have already demonstrated that they
understand structured information in a way that Netscape explicitly does
not. They have to see that Intranets are where the real money will be,
because corporations don't expect to get software for free, especially not
when it supports their business operations. They may not have the
Standards religion completely, but they understand the technical power at
least. If Microsoft pledged to support the standards completely and
without complaint and then produced the sort of software we know they're
capable of they would still own the market because they would get their
first with more. In other words, Microsoft can't lose by supporting the
standards.
As further evidence of this, SoftQuad announced that they are producing a
"Panorama-like" plug-in for Internet Explorer that will provide SGML
browsing functions.
5. The world must have a flavor of SGML that is as simple as possible to
parse. The W3C SGML ERB, of which I am a member, is working on this
specification now with the full knowledge of ISO and WG8 (I'm the informal
liason between the ERB and WG8). When this happens, SGML on the Web will
be even more compelling. Microsoft is involved in this project through
their membership in the ERB.
This leads me to the following conclusions:
1. The companies that can provide HyTime and DSSSL expertise to their
customers and build it into their products will have a *significant
competitive advantage* in the near term (next five years). In addition,
Microsoft is well positioned to capitalize on this if they support the
standard in their software.
2. Unless it drastically changes its technical attitude, Netscape and its
spinoffs will fail miserably or, at a minimum, be marginalized as players
in the Intranet world. In particular, their attempts to develop vertical
markets are almost certainly doomed to fail because they don't understand
the long-term information management issues and the role that data
representation standards must play in any solution.
3. In ten years, SGML will be an accepted and largely invisible aspect of
all corporate information management systems, just as relational databases
are today.
Tools:
I saw three exciting new tools at the conference:
1. Vicom Multimedia's TekDisc multimedia development, management, and
delivery system. This system provides a complete production management
system that combines content object management (like a document management
system), workflow and costing management, media production and capture
(scanning, etc.), and content authoring in a generalized system. It makes
it possible to create large-scale multimedia presentations that are highly
modular and easily configurable. The system is architected in such a way
that it is a natural to apply SGML and HyTime approaches to. For example,
it maintains a database of hyperlinks among the parts of a presentation.
This database could easily be expressed or exported as a HyTime
hyperdocument.
They are currently in the process of adding SGML and HyTime support to the
system. Everything they've said an done indicate's they're serious about
their commitment to standards *because it offers them a technical
competitive advantage*.
This is an industrial-strength tool, so it's pretty expensive (around 500K
$US for a development system) but it could be very important in the IETM
and educational markets, where the cost of producing and managing
multimedia information over time must be minimized. It should be
attractive to large educational publishers. I think we can expect the
price to drop as Vicom's develop costs are amortized and the potential
market expands.
This could be a very important tool, both for the functions it provides
and because it will demonstrate the power of the SGML and HyTime approach
to information management.
2. HyMID, a free "MID" browser. The "MID" is the metafile for interactive
documents, an SGML scripting language used to define the structure of IETM
presentations. The HyMID browser takes as input SGML documents conforming
to HyTime and the MID spec and produces dynamic interactive presentations.
It is free because it was developed by the U.S. Navy. It demonstrates how
SGML and HyTime can be used to enable easy-to-create configurable
interactive information. The MID is not necessarily the technical approach
I would have taken, but it does work.
3. HyBrowse. An SGML and HyTime browser built as a demonstration of
TechnoTeacher's HyMinder software. It is similar to Panorama in its
functions, but has specific functionality for applying style sheets to
HyTime links. In particular, it demonstrates the additional presentation
and behavior control you can have when style sheets can apply not only to
elements in context but to anchor roles in the context of particular link
types.
--
<Address HyTime=bibloc
homepage="http://www.squirrel.com/squirrel/drmacro">
W. Eliot Kimber, kimber@passage.com
Senior SGML Consultant and HyTime Specialist
Passage Systems, Inc., 10596 N. Tantau Ave., Cupertino, CA 95014-3535
(408) 366-0300 (Cupertino), (512) 339-1400 (Austin),
http://www.passage.com </Address>
"If I never had existed, would you still remember me?..."
--Austin Lounge Lizards, "1984 Blues"
(http://www.webcom.com/~yeolde/all/lllhome.html)