timm@runxtsa.runx.oz.au (Tim Menzies) (07/16/90)
this is my FOURTH attempt to post this to the net. forgive me if there is
a strong element of deja vu in this article
---------------------------------------------------------
Dear Net,
I am looking for tools to support a large-ish hypertext application. What I'd
like to find are:
o the fax numbers and email address (if any) of the vendors of the
hypertext tools;
o anyone with experience with these tools (and hypertext in general);
o anyone who can tell me if the application description I give below can
be done using commercially available hypertext tools.
If there is enough interest, I will post a summary.
Thanx in advance,
Tim Menzies
HYPERTEXT TOOL & VENDORS
Here's a list of products that I;ve got from the literature. Any additions
anyone?
HyperBase Cogent Software Ltd. 21 William J. Heights, Framingham, Mass 01701
(508) 875-6553
Hyperties 55 Princeton-Hightstown Road, Princeton Junction, N.J. 08550
(609) 799-5005
IBM Linkway IBM Educational Systems, P.O. Box 2150, Atlanta, Ga, 30055
(404) 238-3245
KnowledgePro Knowledge Garden Inc. 473A Malden Bridge Road, RD 2,
Nassau, N.Y. 12123
(518) 766-3000
MaxThink MaxThink Inc. 44 Rincon Road, Kensington, Calif, 94707
(415) 428-0104
Guide Review Owl International Inc. 14218 N.E. 21st St. Bellevue, Wash.
98007
(206) 747-3203
Window Book Review Window Book Inc. 61 Howard St, Cambridge Mass, 02139
(617) 661-9515
APPLICATION DESCRIPTION:
We are looking into the construction of a hypertext application with the
following parameters:
o The system must run on a standard business machine; i.e. a DOS box with
640K or RAM **only**.
o Our market research suggests that the whole package will have to sell
for less than $1500 Australian. Any tools we use will have to support
a licensing arrangement that allows us to meet this sales price.
o Due to the sales cost, solutions that require CD or optical readers are
not acceptable. Our system will be hard-disc-based.
o The system must support four hyperdocuments:
o a 10,000 document that is updated quarterly. 8,000 pages of this
document are replaced annually.
o a 1500 page document that is updated quarterly.
o a 12 page document issued weekly
o each organisation will be able to define their own "local
document" that describes connections between parts of the above
three "basic documents".
o Each of these four documents may have an arbitrary set of cross-links
to each other as well as a simple word index and a table of contents.
We estimate that we are looking at 100,000 hypertext nodes, each with
10 references into it. There is no guarantee that these cross-references
will form a hierarchy. Our cross-references might be quite arbitrary.
o The documents will be initialised from their existing hard-copy
versions. We plan to import them from their typesetting files. Our
hypertext system must support a batch-import utility.
o The documents are updated in three ways:
o Users will be able to define their own links. Hence, the system
must support a user-friendly way of establishing connections
between parts of the hyperdocument(s).
o Organisations want to collect together the documents of individual
users. The system will have to support collection of links from
numerous sources and the collection of same into one
organisational hyper-document (e.g. read/ write permissions for
sections of the documents, automatic cross-filing of several
documents).
o Regular updates are issued by the authors of the basic documents.
These updates will be automatically filed into the existing
hypertext document.
o Our first goal is a multi-windowed environment with different parts of
the hyperdocuments being displayed in different windows. Words/ phrases
that have cross-indexes are highlighted and the user can click on this
to open a new window that shows more information about that word/
phrase.
o Our next goal is to write a set of "monitors" that observe how people
move around this initial system. These observations will be used to
design subsequent versions. We plan to build this iteratively. A tool
that supports rapid-prototyping (e.g. fast compile/link/run/debug
cycles; access to interface tools; interpretative development
environment) would be ideal.
o The system aims to help busy consultants with searching the literature.
Hence, we want to write "smarts" into our system so that the system can
automatically suggest interesting cross-references and filter out un-
important cross-references. Path histories and a "play-back" mechanism
would be useful as well. Also, we are an expert systems group. We can
see a great utility in adding a rule-based expert systems to all the
above.
o Parts of the hyper-documents will be "active models" (formulas and flow
charts). We want to be able to specify and execute these active models
so that the users can feed in the particulars of their own work, and see
some answers.