How Can we Leverage Corporate Influence to Catalyze the Development
and Adoption of an Improved Hypertext Standard?
(to preserve the integrity of key corporate documents, making
lawyers generally happy about an otherwise troublesome situation)
Many companies are now using the tools developed for the World
Wide Web to create Intranets to share many types of information among
employees. There are important reasons for the popularity of Intranets:
unparalleled ease of use, low cost (since most companies already have an
installed base of computers), platform independence, and the availability
of off-the-shelf systems that are easy to set up and maintain.
But there is also an important unsolved problem (and it is not necessarily
obvious that this is a problem) that relates to current
World Wide Web hypertext standard:
In order to add a link, a document must be changed.
For dynamic documents like employee benefits and quality assurance procedures
this is generally fine, but in research environments that generate technical
reports this can lead to new problems that must somehow be addressed.
Problem 1: The Permanence of Documents
Traditionally we want technical memos, reports, correspondence, and so forth
to be permanent archives, frozen in time. The tradition stems from various
With the advent of hypertext publishing, the rules seem to have changed.
In order to take advantage of the power of hypertext publishing and enrich
the information content of a document by relating it to other documents,
authors need to add links at some later date. And since (with the current
hypertext standard) links are embedded in the document, the original must
be open to alteration. This opens the door for various other alterations:
fixing a typo here, a grammar correction there, rephrasing a sentence, .
. . and poof(!) - you no longer have the original.
- It's really a pain to recall hardcopy after it has been distributed,
make a change, and redistribute it.
- No one wants to read the same report over and over with only minor
changes each time. The accepted procedure is to publish a new report (when
there is enough new information worth publishing) that details the new information,
additional progress, and any corrections to the previous report.
- Unchanging hardcopy is an integral part of the paper trail leading
to patent disclosures and filings.
Problem 2: Multiple Copies
If a document is published electronically on a company's Intranet, it can
(and probably will) be downloaded to some client's computer. If the author
later changes the original and posts the change, the changed copy can be
downloaded. With multiple changes and multiple downloads, and the propensity
of certain individuals to print these copies, it is not hard to envision
a proliferation of many versions of documents existing within a company
in both hardcopy and electronic copy form. This can create a fair amount
of confusion for a development team and a nightmare for the patent attorneys.
Problem 3: Who Makes the Changes?
In the situation described so far, we've assumed that the author is the
person making the change. Eventually, authors leave the company and others
may take over their projects or at least refer to their work. It may often
be desirable to not only point to the original author's work, but to point
from that work to other documents. With the original author gone, who has
authorization to make this sort of change? How will the change be documented?
How will previous versions be maintained? How can we insure that only links
are added and the document remains otherwise unaltered?
In another scenario, the author is still with the company but - for reasons
of pride, jealousy, or obstinance - refuses to add links from his often-cited
documents to those of a worthy colleague. This may result in the alternative
information not reaching those who could benefit from seeing it.
One solution to Problem 1 is to implement a quality system to prohibit authors
from making non-link changes. It is important to realize that it is not
necessary to develop a technical solution to absolutely prevent people from
making non-link changes. Lawyers, Quality Assurance departments, and ISO
9000 auditors are satisfied as long as there is a system in place and
people are trained to comply with the system. The same would apply to Problem
2: develop a quality system prohibiting employees from downloading or printing
certain documents. And in Problem 3, a quality system would specify who
could add links to documents and what the procedure would be for doing this.
This solution is time consuming to implement and difficult to maintain.
Records should be kept regarding what links are added, by whom, and when.
It also does not completely solve the patent attorney problem that the original
document is no longer virgin, even though the browser-visible text is the
same except for some underlining and highlighting. And it does not guarantee
that people won't make other changes.
Retain a static archive of the original and maintain a modifiable working
copy on the Intranet. This doesn't work, because a series of modifications
to the dynamic copy may be important steps in leading to a patent. If not
saved, your "paper trail" is broken in the patent process. And
you are back to the colleague-confusion problem where Jane has a different
version than Hank's printout (and the author maybe didn't assign version
numbers at the top of each copy for each change).
This is a brute force approach: keep copies of every single revision but
only link to the most recent one. The lawyers will love you for this one.
Get a Better Hypertext Publishing System
This, of course, is what this proposal has been leading towards. If the
problem is that the links are intrinsic, make them extrinsic. This concept
was articulated by Eric Drexler in his
paper Hypertext Publishing and the Evolution of Knowledge.
If the links are extrinsic, then the document can remain in its original
form with corrections, comments, refutations, links to new developments,
and so forth added by anyone. Extrinsic links could be treated as distinct
objects with their own authors. They could be two-way, fine grained (attached
to small chunks of link-author-defined text), and subject to filtering.
This type of system would nicely solve two problems for us: (1) access restrictions
(who can and can't make links from documents) could be greatly relaxed,
and (2) documents could remain in their original form, cleaning up the paper
trail for patent development and corporate archiving in general. There are
other advantages, but for the purpose of this proposal it is sufficient
to restrict our focus to this narrow range of advantages for commercial
We believe that many other companies besides ours that have built Intranets
face similar issues. If so, there is an incentive (like a real live
commercial need) for these companies to use extrinsic links on their Intranets.
We want to hear your thoughts on how this might be accomplished. If there's
a commercial need, perhaps a software developer could be enticed by the
potential for economic gain. If this is simply a standards issue, maybe
the World Wide Web Consortium (W3C)
should fund development. In either case, we need to figure out how to encourage
them to do that - leveraging as much corporate support as we can muster.
This doesn't necessarily mean large cash donations - if enough corporate
voices sound an alarm, W3C will probably react. The Foresight
Institute has shown a willingness to pursue this
sort of project, given sufficient funds. A consortium of interested
parties could fund them or someone else until the development is mature,
then turn it over to W3C. Other ideas? Any thoughts about competing standards
like Xanadu, Hacklinks,
Published and maintained by Russell
Last updated: 23 September 1996
Moved from www.asiapac.com/ to www.foresight.org/ on March 28, 1997
the WWW | About the Foresight Institute
| Foresight Institute Home Page |
Foresight materials on the Web are ©1986-1997 Foresight Institute.
All rights reserved.
Last updated 28March97. The URL of this document is: http://www.foresight.org/WebEnhance/ImpHTForrest.html
Send requests for information about Foresight Institute activities and membership
Send comments and questions about material on this web site and reports
of errors to firstname.lastname@example.org.