Hyper-G Web Enhancement Project

Since this article was written, Foresight has revised its Web Enhancement strategy to emphasize other approaches in addition to Hyper-G.

Critical Discussion Experiment to Begin on Web

by Chris Peterson

Foresight Institute has a special interest in systems to improve the evolution of knowledge and to enhance the quality of discussion and decisions on complex issues. Currently there is no good way to carry out such discussions: paper is too slow and inconvenient, while Internet discussions--whether they be in the form of newsgroups, static web pages, or chat sessions--are too unstructured.

Our Web Enhancement Project aims at adding features to the World Wide Web needed to better carry out critical discussion. These features have been described in the essay "Hypertext Publishing and the Evolution of Knowledge" available on the web (www.asiapac.com/Hypertext/HypertextPublishingKED.html) and on paper from the Foresight office.

Hyper-G has true hypertext publishing features

We had originally thought that this would require Foresight to produce the needed software, but fortunately this has become unnecessary by the introduction of Hyper-G by an international team (originating, as did the web itself, in Europe). Hyper-G, known as HyperWave in its commercial version, has almost all of the features on our wish list for hypertext publishing, because it was based on the early concepts of hypertext from Ted Nelson, and also explicitly designed for structured discussion:

Bidirectional links: links that work in both directions between linked documents. On today's web, links work in one direction only.
Extrinsic links: links that can be made visible without the document author's permission. Needed for critical discussion.
Link typing and sorting (filtering): the ability to label a link with a keyword describing its type (e.g., criticism or example) and sort for only those links to be displayed. Needed when a document has many extrinsic links. Annotations are an explicit link type already coded into Hyper-G.
Fine-grained links: as on the web, links can be made to only part of a document, even one word or one letter. In Hyper-G, up to six links can overlap.
Links in non-ASCII documents: links can be made and viewed in PostScript, gif, jpeg, tiff, video, VRML, and soon audio and PDF (Acrobat) documents. This is possible because unlike conventional HTML documents, which encode links within the document itself, Hyper-G supports the ability to maintain a list of links in a database, which can be merged into the document at display time. Links are treated as objects in their own right, with their own attributes and permissions ("rights").
Fine-grained access control: for each document, one can specify who can see the document, who can see each link, who can edit the document itself, and who can edit the links.
Advanced structuring facilities for documents, including graphical and even 3D visualization of the structure
Sophisticated search abilities, including the ability to search Postscript document contents (i.e., special Hyper-G tools can "read" PostScript document contents and make a full text index of them). A search engine is built into the server, which cooperates with a more lightweight search ability on the Hyper-G client.
Compatibility with the existing web: Hyper-G documents can be viewed using standard web browsers.

Hyper-G client software is available for UNIX, Windows NT, Windows 95, and is in preparation for the Macintosh. A line-oriented terminal version of the client is also available. However, the UNIX client is the most advanced, and can be run under the operating system Linux on Wintel machines, as Foresight plans to do.

Partly because they include commerically-useful features such as subscriptions and licensing, Hyper-G or Hyper-G spinoffs are already in use at publishing companies such as Springer, Academic Press, Wiley, and Oxford University Press. It is also used extensively by the European Space Agency.

Foresight's experiment

Foresight can experiment with Hyper-G without betting on its long-term success as a standard. The goal is to use the basic capabilities of second-generation hypertext publishing systems by building information structures with real content. This content could later be transferred to another system that provides the same basic capabilities. Foresight hopes to show the usefulness of the advanced hypertext publishing features listed above: we may be instrumental in spreading these back into the World Wide Web as a whole. Thus, our efforts don't depend on Hyper-G and HyperWave commercial success, but on how well we demonstrate the feature set.

Computer security debate

Our first experimental debate will be in the field of computer security, specifically language and operating system security: how can we maximize cooperation without vulnerability? We will start by examining Java-style languages. This topic has several advantages for an initial debate:

It is important to the safe and widespread deployment of nanotechnology, i.e. it is critical to our shared future.
It is already of great current interest for commercial reasons.
It will be debated by those familiar with computer technology: early adopters who already use the web and may be willing to install the Hyper-G client software so they can participate actively in the debate.
It is complex enough to demonstrate the usefulness of our target feature set for debating complex issues.
It is relatively theorem-like: propositions can be clearly stated--"Given these assumptions, this security violation is impossible"--which can then be tested and possibly disproved either theoretically or experimentally. Thus, while the topic is complex, it is not as messy as human systems. Our first debate should be one in which actual progress is possible.

The funds for Foresight's Hyper-G server were raised at this year's Senior Associates Gathering. This machine has now arrived and is being configured by Russell Whitaker, technical leader of the project. We will be putting in a skeleton argumentation structure, and then inviting specific security experts to join the debate one by one. The reason for this controlled build-up of participants is that we expect to encounter glitches in the process which will have to be solved using social rules, rather than the procedures we can enforce using the software. We will also have to evolve filtering procedures.

Once it's clear that the debate software is working well, and we are being successful at adding needed social rules, we will open up access to the debate first to Senior Associates, later to Foresight members and some relevant professional groups, and eventually to the general public.

This computer security debate is only the first of many Foresight plans to conduct on advanced technologies of public policy importance. We hope that the debate procedures we evolve can be of use to those debating other topics as well-including "messy" human issues.

Hyper-G information sources

Those interested in assisting the project at this stage can start to familiarize themselves with the software by reading the book HyperWave: The Next-Generation Web Solution (by Hermann Maurer, Addison Wesley, 1996; available free online at http://www.hyperwave.de/hw_book) and by installing the client software available free online (ftp://ftp.ncsa.uiuc.edu/Hyper-G or ftp://ftp.utdallas.edu/pub/Hyper-G).

See the FAQ at http://www.hyperwave.com/faq. You may also join the discussion group comp.infosystem.hyperg, or join its mailing list mirror hyper-g@iicm.tu-graz.ac.at by sending email to listproc@iicm.tu-graz.ac.at with the message body: subscribe hyper-g <Your Name>

Next steps

In addition, funds are needed immediately to pay for (1) Foresight's client machine, about $3900; (2) people's time in organizing the project, tech support to targeted experts, and uploading reference documents; (3) $500 membership fee for Foresight to join the Hyper-G Consortium, and thereby become eligible for R&D grants from the Consortium; and eventually (4) our own connection to the Internet when we outgrow our initial shared T1 connection.

In the longer term, we invite all Foresight members--and eventually all web users-- to join us in debate online. We believe that full hypertext publishing capabilities are a breakthrough equal in importance to the invention of the library. No other tool is sufficient to deal with the complex problems to be solved in successfully implementing nanotechnology and the other advanced technologies now on the horizon.

For project updates, visit our web site www.foresight.org. Donations may be discussed with Chris Peterson at peterson@foresight.org or tel 415-917-1122, or mailed to Foresight Institute, PO Box 61058, Palo Alto, CA 94306 USA, and are tax-deductible in the U.S.

Special thanks to Russell Whitaker for technical leadership; those who donated funds for the server machine: Hughgie Barron, Ken Blakeslee, Steve Burgess, Warren Freeman, Dan Fylstra, Jim Lewis, David Lindbergh, Joy Martin, Chris Portman, Gary Pullar, Dick Smith, and J. Tory.

Since this article was written, Foresight has revised its Web Enhancement strategy to emphasize other approaches in addition to Hyper-G.

For ongoing nanotechnology information, Register for a
To help prepare for nanotechnology	Make a Donation to Support Foresight
To play a key role in Foresight's efforts	Become a Senior Associate

| Enhancing 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 18May97. The URL of this document is: http://www.foresight.org/WebEnhance/HyperG.html
Send requests for information about Foresight Institute activities and membership to inform@foresight.org.
Send comments and questions about material on this web site and reports of errors to webmaster@foresight.org.