home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Fritz: All Fritz
/
All Fritz.zip
/
All Fritz
/
FILES
/
WORDMISC
/
HYPER.LZH
/
FILE66
< prev
next >
Wrap
Text File
|
1988-02-01
|
3KB
|
71 lines
"Hypertext substitution for Neil Larson"
========================================
I spend most of my day on the phone, often providing the same information
to many different people. That sounds like a great opportunity for
hypertext. <FILE53 TALENTS>
I could collect all my knowledge, advice, and solutions into a hypertext
system, then let others give it out. But I won't do it for three reasons
-- change, measurement, and enjoyment. For a moment, let's consider each.
As for changes, I want to make whatever improvements I can to my software.
While some people believe that I single-handledly created MaxThink, HOUDINI,
and PC-Hypertext, most of the ideas contained in these products came from
comments or suggestions from users.
For that reason, while I may give out the same information to different
people, I receive invaluable hints on future commands and uses for my
software.
As for measurement, in one sense I improvise software -- meaning that I'll
put anything interesting in my software just to see if it may benefit
users. I'll often ask the people that call me if they use some of these
features or if a proposed enhancement would be important to them.
In both cases, I'm trying to measure the potential capabilities of my software.
That's how I do my market research . . . by talking to users.
Finally, I enjoy talking to users who share my interests in thinking,
planning, and writing. While my own academic training is in math and
marketing, thoughtful users have introduced me to new ideas in medicine,
philosophy, education, engineering, etc. That's nice -- as I believe I can
always learn something from others.
Now the twist!
If I did not want to measure or change my world, then putting my knowledge
into a hypertext decision-tree or into an expert system makes sense. But,
that's why most expert systems are a disaster.
If you build systems that provide advice, how will you know when the advice
needs changing? It is only through the monitoring of the questions and the
advice given that you discover whether your system is relevant.
Simply stated, it is one thing to build an expert system (seldom done), and
quite another thing to modify and revalidate the system (never done).
What has this to do with hypertext?
In an expert system, the stepping through of rules and linkages are
automatic (and hidden) from the user. <FILE21 EXPERT> In a hypertext
system, the stepping through rules and linkages is manual and always
visible to and easily changed by the user.
For these reasons, if you wish to codify a body of knowledge (i.e.,
establish access patterns to particular segments), hypertext may be the
better approach in terms of speed, visibility, and viability. However, at
the core, using a hypertext system is:
┌────────────────────────────────────────────────────────────┐
│ accepting someone else's thinking to ease your own. │
└────────────────────────────────────────────────────────────┘
While useful in most cases, hypertext is really only a representation of
reality, and not reality itself. If reality is changing, your hypertext
system needs constant attention to keep it closely matched to reality.
Neil Larson 1/14/88 FILE55
44 Rincon Rd., Kensington, CA 94707
Copyright MaxThink 1988 -- Call 415-428-0104 for permission to reprint