home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Simtel MSDOS 1992 September
/
Simtel20_Sept92.cdr
/
msdos
/
hypertxt
/
hyper.arc
/
FILE20
< prev
next >
Wrap
Text File
|
1988-02-01
|
9KB
|
168 lines
How will CD-ROMs affect Hypertext?
==================================
Thinking about all the epic storage devices becoming available in CD-ROMS,
WORMS, and laser disks, might cause one to have dreams filled with hypertext.
But, there are also several reasons why hypertext is not suited for CD-ROMs.
I'll cover this in a moment. But first, let's consider the storage space:
Type of storage Approximate Capacity
=============== ====================
Laser disk (12) Stores 54,000 pictures (1,200 megabytes)
CD-ROM Compact-disk read-only memory (600 megabytes)
WORM Write-once read-many (400-600 megabytes)
Sounds great. However, is it relevant to hypertext <file65 KNOWLEDGE>?
I think the main purpose of hypertext is to allow users to browse through
information. This process has two implications concerning language and
structure.
LANGUAGE With respect to language, users will vary greatly in their
======== understanding of the jargon, vocabulary, or classifications
in a system of knowledge. For beginning users, an ideal
hypertext system should help them understand the
categorization methods used to describe the ideas within
the field of search. <file55 IMPORTANCE OF LANGUAGE>
STRUCTURE This is easily accomplished by organizing all information
========= into clear hierarchies <file50 DEFINITION>. As users
access a component of information organized within a
hierarchical system, they can't help but expand, modify, or
reaffirm their understanding of the structure of all
information within the system. That's language
acquisition!
At the other limits, with expert users who already
understand the language of the field, an ideal hypertext
system should index information not by language, but by
idea content.
The important observation is that both of these processes first require
the information to be intelligently organized. That's the value added by
builders of hypertext systems <file43 EXAMPLES> . . . and that's not the way
most large systems of information are currently organized.
The large data systems that might use the 500-1,200-megabyte systems are
already biased into free-file relational approaches -- using either key word
searches or set intersections methods. See example in <file23>.
That's great . . .,
- if you already know the language in the system of knowledge,
- if you understand set operations,
- if you want absolute pattern matches rather than "close fits,"
- if you know all the synonyms for what you're looking for, or
- if you are willing to search using all combinations of synonyms.
The truth is that with even the largest most sophisticated commercial
free-field relational systems in the legal field, even experienced users
typically find only 21 to 23 percent of the relevant cases. <file69 EXAMPLE>
Nevertheless, I imagine many firms put large amounts of information into
relational CD-ROM or WORM systems for two reasons: It is easy to copy paper
to disk (relational systems do work at a primitive level), and it avoids the
time and expense of intelligently reorganizing the information using hypertext
approaches.
So, you'll see encyclopedias, dictionaries, manuals of style, and other such
tools on CD-ROM disks -- with the computer adding only speed to the page turning.
But there is nothing new there to help users acquire the language or structure
of the information (i.e., knowledge) contained. <file75 KNOWLEDGE>
If the information format didn't transmit knowledge before, putting it on the
computer will gain nothing. Knowledge is not merely a composite of references,
but rather understanding the structure that has been embedded into the system.
Receiving information without the structure simply converts information into
data. <file62 VALUE OF INFORMATION>
It's easy to celebrate the Information Age <file51 DEF.> and all the tools
that produce information (word processors, databases, spreadsheets, radios,
magazines, etc.). But it is more difficult to acquire understanding as you
may be drowning in information . . . and parched dry in knowledge.
If there is only an ounce of knowledge in every ton of available
information, putting it on CD-ROMs doesn't help many people. Without
getting into the differences between data, facts, information, insight,
knowledge, wisdom, values, and behavior <file46 DEF.>, let me just say that:
For one client:
┌──────────────────────────────────────────────────────────────────┐
│ We hypertexted 3 feet of paper (2 megabytes of technical │
│ rulings) onto three floppy disks. │
│ │
│ We will update the system with the new rulings (one disk per │
│ month containing 500K of information). │
│ │
│ We think the CD-ROM cross-over point exceeds 30 megs (cheaper │
│ CD-ROM than hard disk). │
│ For details, see <file43> │
└──────────────────────────────────────────────────────────────────┘
The reason we created a floppy disk hypertext systems centered on:
- the size of the knowledge system (ultimately 8,000 files - 40K links)
- the number of users (1000-10,000)
- the computer technology on the desks of the users (all hard disk)
- the percentage of changes to the system with each update (5%-20%)
- the frequency of update (every 3-6 weeks)
From my understanding of how new markets open, most hypertext opportunities
<file80 OPPORTUNITIES> will first occur in information niches, where the
benefit gained is not the size of the system in megabytes, but in the value of
the information on the disk. See <file29 TYPES OF INFORMATION>.
Whether this value comes from improvement in accessibility, completeness,
currentness, skill expansion, or time savings, both the people adding this
value as well as the end users of such hypertext systems may have initial
biases against using CD-ROM technologies. <file44 HYPERTEXT AND SOCIETY>
Again, if you consider the formats and sources, most information suitable for
hypertext treatment (whether governmental or private) exists in text formats.
Whether specifications, rulings, laws, depositions, hearings, etc., the ratio
of pictorial to text information in the printed world is extremely low.
For these reasons, I think CD-ROMs are not now crucial to many forms of
hypertext. However, there's an exception.
The Chinese have said that one picture is worth a thousand words. But in
current computers, a picture might take from 3,000 to 5,000 words in storage
space while communicating perhaps 100 words of information. CD-ROM or WORM
disks <file22 STORAGE DISCUSSION> will be important only if you need a large
number of high-resolution pictures in your hypertext system.
For my conclusion, the current floppy/hard disk configurations can support
impressive textual knowledge systems. However, with hypertext systems
centering on graphics, pictures, audio, and video segments, CD-ROMs are
necessary.
1/10/88 Postscript -- I just finished reading about CD-ROMs in Microsoft's
================== SYSTEM JOURNAL magazine claiming:
CD-ROM mastering costs -- $3,000 to $10,000 per disk
Cost of each CD-ROM disk -- $3.00
Cost of CD-ROM drives -- $1,000 (using MS-DOS standard HSG format)
From my viewpoint, if the users of my 8 megabyte hypertext system all had
CD-ROM drives, I'll fill the remaining 492 megabytes of my CD-ROM with
supporting video. In effect, rather than add more hypertext links, I'd use
the storage space to enhance the communication of the contained knowledge.
See <file55 KNOWLEDGE DEF.>
However, this totally overlooks a completely different reason for shifts to
CD-ROM formats. For example, CD-ROMs are an attractive format for software
publishers who wish to reduce unauthorized copying (piracy) of their efforts.
Given the choice, every software publisher I know would gladly shift from
$1.00 a disk (floppy formats) to $3.00 a disk (CD-ROM formats).
Look at it this way. Users want storage space and developers want piracy
protection. The interests of both may bring CD-ROMs rapidly to the market.
Neil Larson 1/16/88 FILE20
44 Rincon Rd., Kensington, CA 94707
Copyright MaxThink 1988 -- Call 415-428-0104 for permission to reprint