WORLD CONFERENCE ON RECORDS AND GENEALOGICAL SEMINAR COMPUTER INPUT AND OUTPUT TOOLS AND TECHNIQUES

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP73-00402R000100140010-8
• Release Decision: RIFPUB
• Original Classification: K
• Document Page Count: 6
• Document Creation Date: December 19, 2016
• Document Release Date: July 19, 2006
• Sequence Number: 10
• Publication Date: August 5, 1969
• Content Type: REPORT

Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
WORLD CONFERENCE
ON RECORDS
AND GENEALOGICAL SEMINAR
Salt Lake City, Utah, U.S.A.
5-8 August 1969
COMPUTER INPUT AND OUTPUT TOOLS AND TECHNIQUES
Part I
The Input and Output Syndrome
By
Barbara G. Fisher
"Re (',ri/ J'r',lr (ln,u In
(1/i 1 'lit e ll uil I I "alii?
{{11< 14 /j -6 & 7 a
Approved For Release 2006/09/25 : CIA-RDP73-00402R000100'140010-8
 Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
COMPUTER INPUT AND OUTPUT TOOLS AND TECHNIQUES
Part I
The Input and Output Syndrome
By
Barbara G. Fisher
A syndrome is a given set of symptoms characteristic of a particular disease. While the
input-output complex, which I describe as the I/O syndrome, is not indicative of disease, input
and output content, methods, and machines sometimes lead to a sense of frustrating malaise.
Choice of the right or the best machine can be an irritating factor in the development of
automated systems because input and output content and form are compounded of many cells
of logic, each of which must be diagnosed and shaped to machine capability. First of all,
however, we must be physicians to our own intelligence, and ask ourselves what we want to
automate and why.
The prescription for any automated or mechanized system calls for a combining formula
of readability, compatability, and retrievability. In short, an integrated processing system must
respect data requirements, machine components, their capabilities, and potential user needs.
The effectiveness of any machine system, however, first requires that exercise of judgment
known in computer jargon as feasibility. A feasible project is simply one that is practical
in terms of data preparation, use of available hardware, and over-all cost.
Feasibility studies of both manual and automatic information systems are basic diagnostic
steps that we must take before any information or management system can be committed or
converted to computer control. Computers can do no more than they are instructed to do. We
must prepare those instructions on a common sense basis of intelligent analysis of what we
want the machine to do and why.
Before we decide what data to make machine-readable, what programs will be clearly
compatible with the machines available, or how we would like to retrieve stored data, we must
grapple with stern realities-hardware limitations, operational costs, man-hours and machine
time, expected results, and basic purpose. In addition, distinctions must be made between
operations that can best be handled by unit record or data processing equipment, and those
operations that require total computer system design.
A computerized system is practical only after a hard look at data accumulation problems,
projected benefits, and at the underlying principles of computer technology. Poor or useless
data will not assume excellence or usefulness simply by putting it on magnetic tape or punched
Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
 Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
cards. Meaningless manipulations of data will not suddenly assume meaningful form simply
because the manipulations have taken place at great speed. If input on a sophisticated IBM
2741 terminal is made up of "garbage," this is exactly what the terminal will transmit and
exactly what the printer will print--"garbage." Only transmigrations, not transmutations, take.
place in the computer.
The computer configuration-the machine battery that includes memory units, data
processing machines, tape and disk drives, terminals--is increasingly becoming a
communications rather than a processing configuration. Engineering design is turning more and
more to computer storage and on-line access, and away from simple one-time or batch
processing. Archivists and genealogists are also increasingly concerned with accumulation and
storage of great masses of reference information, centralized in one file but accessible to single
inquiry as well as to retrieval for future selected assembly of data for publication.
Evolution of input equipment from the keypunch machine to direct access or on-line
terminals illustrates a continuing effort to narrow the gap between man-generated source data
and computer-stored information, and to reduce the time between conversion of information
to machine-readable form and to the ultimate production of output.
Choice of input equipment depends primarily on the need for rapid source data
conversion, massive accumulation of data, and on the continuous necessity to update stored
information. The period of wide choice in input machine selection may be drawing to a close,
however, because more and more input devices are falling victim to the peculiarly American
idea of obsolescence. The characteristic American thought that old models should be replaced
or traded in when new models appear is nowhere more vividly demonstrated than in the rapid
evolution of IBM computers-- from the 1400 series to the 360 series. Indeed, the computer
industry has been so preoccupied with technical advance that it has seldom been conscious of
the human need to progress by gradual stages. Increasing competition in computer
manufacturing is producing a great proliferation of equipment and will undoubtedly lead to
substantial cost reduction, but it is also adding to the momentum of obsolescence. If the rate
of hardware competition continues to accelerate, paradoxically enough there will soon be less
and less choice of input equipment, as manufacturers tend to conform to best-selling models.
A brief history of input equipment demonstrates that computer design has been directed
to narrowing the gap between raw source data and computer-stored information. Initial
electric or electronic systems, still less than 50 years old, required:
1. Manual transcription of data from source documents to data sheets.
2. Keypunching of data from the data sheets.
3. Verification and sorting.
4. Printout on a tabulator-printer (a line printer or listing machine).
Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
 Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
The second machine effort, designed to reduce the processing time gap, was the
continuous roll paper tape punch. The Friden Flexowriter is representative of this kind of
input device. This system requires:
1. Manual transcription of data on data sheets.
2. Simultaneous punching onto paper tape.
3. Conversion of paper tape to magnetic tape (this generally requires special converter
hardware).
4. Conversion of paper tape to cards to magnetic tape (a required system when no
converter hardware is available).
An input device which permits input data to be transcribed directly onto mini-reels of
magnetic tape compatible with the bit, byte and track components of computer tape drives is
exemplified by Mohawk 1101 magnetic data tape recorder.
The optical scanner, developed about the same time as the paper tape input machine, was
designed to transfer data directly from source documents to computer magnetic tape. The
process greatly reduces the time required for data conversion although formatting of input
data is of necessity strictly standardized. A new direction in input devices, still in
developmental stages, is the laser beam sensor.
The most rapid input and access device currently in use is the terminal or on-line
keyboard-an electronic keyboard linked to the master computer by cable. The terminal can
carry messages and data directly into the computer, can interrogate stored data, and on
instruction can feed back rapid response to inquiry. The terminal input device provides almost
instantaneous communication between the computer input, data storage, and output
sequences.
All five input systems are currently in use, sometimes in the same computer center. Each
device has distinctive job advantages. The card keypunch system is an excellent teaching tool
for newcomers to computer processing. It is a visual aid to understanding the logic of input
design and the logic of data definition. It also teaches principles of automatic control of data
input through program drum control cards, and the use of ,automatic feed, duplication, and
skip controls. The card punch system permits total visual and intellectual operator control of
input data. Associated unit record equipment--verifiers, sorters, collators and interpreters--also
provide insight into error checking procedures, coding, and data manipulation. The card
keypunch machine is the most widespread of all input machines because rental is reasonably
moderate, and anyone who can type can easily learn to keypunch.
A by-product of card punching systems is the accumulation of a working card file during
the preliminary stages of testing or experimental programs. One of the advantages of the card
Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
 Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
system is the multiple usefulness of cards to many users. In addition, card files may be
processed in either the computer or unit record machines. Most other input devices are
restricted to machine specialists or operators, are generally housed in a closed shop, and input
data can be interpreted only by the computer. The disadvantages of punched cards are the
necessity for special storage and filing of card decks, the possibility of card damage by tearing
or warping, and the confining concept of only 80 columns for the transcription of data.
While paper tape and magnetic tape input devices permit the mass production of
converted data with automatic control of format errors, paper tape processing by the
computer is relatively slow. The computer can interpret only 60 tape-recorded characters per
second whereas punched cards can be interpreted at the approximate rate of 1280 characters
per second. Many tape input devices have presented substantial machine compatability
problems and many computer centers cannot handle a diversity of input forms economically.
The computer and its components operate on integrated and exacting timing schedules.
Irregular speeds at any phase of the processing--program assembly, data interpretation, core
storage, tape-to-tape search and retrieval, and printing--can badly backlog, delay, or disrupt the
system. Paper tape machines are particularly useful, however, for duplicate copy making, for
form letters and for administrative reports. They have demonstrated their usefulness
consistently in the data processing activities of the American Genealogical Society.
But input is only the ingress phase of the computer processing sequence. One of the
major dilemmas of the I/O syndrome at the output phase is the unmanageable masses of paper
printout that can result from a total output program. Endless paper output is one aspect of the
information explosion--a computer-generated information explosion, about which little
comment has been made. A computer store of 5,000,000 characters, for example, could
produce about 100,000 lines of text that could be printed in about 1 1/2 hours. At that rate, a
printout of stored data might run to over 2,000 pages. In the foreseeable future low cost mass
storage will accommodate 600,000,000 characters. This could conceivably result in a printout
of 12,000,000 lines of text approximating some 200,000 pages of print. It is awesome to
realize that a text this large could be processed by the high speed IBM 1403 printer in about 3
hpurs.
The high speed printer is a necessity in a well-integrated computer layout where many
small as well as large projects are processed. Integrated also with a terminal access system, the
high speed printer can produce selective hard copy output at the same time the computer is
responding directly to inquiry. The printer is designed to handle many jobs rapidly as well as
to produce extensive printouts. .
Resolution of the phenomena of too much information requires the same logical tests of
practicality as those that are applied to the determination of input. In fact, output should be
determined even prior to final definition of input. The consequences of any projected system
are as important as the sequences. In most present computer installations and in current
software packages, output must of necessity be selective.
Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
 Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8
As librarians, bibliographers, archivists, genealogists, we can estimate reasonably well, on
the basis of established criteria and accumulated experience, what elements of data in a master
record are most often needed. Selective retrieval and print programs can be designed to
produce both economical and pertinent professional indexes, listings or guides.
Increasing momentum toward on-line computer services, and wide dispersal of terminal
keyboards, however, will soon shift the initiative for meeting real needs for stored information
from the librarian to the user. The computer processing sequence of the future will expand
from an input-storage-output cycle to an input-storage-use record and output or response
cycle. Just as oral magnetic tape now preserves random or informal responses and opinions,
computer recording media can preserve a record of the uses made of stored information.
By-products of this enlarged processing cycle could include computer assembled thesauri
of search of key terms, subject bibliographies, and even work-in-progress or user-trend reports.
The computerized data bank of the future will yield to the on-line inquirer not only the
data he needs and wants; it will also record his method of search, accumulate search patterns,
and will reduce demand for massive printouts. Output will more and more become
user-generated and optionalized, tailored to meet specific user need.
The emphasis in input machine design is on reducing the necessity for preselection and
pre-editing of information. The emphasis in output is on increasing the immediacy of the
dialogue between inquirer and computer-stored information. The trend in computer
technology is toward automata and self-generating systems. To prepare for this text realization
of computer potential the newcomer to automation should study the history of machine
developments from the abacus to the IBM 360, from the typewriter to the terminal keyboard.
While machine concepts of data recording and manipulation do not provide insight into the
meaning of information, they do provide clues to methods of human information gathering
and communication.
Input and output concepts, methods and machines need not be a syndrome marked by
irritation and malaise. Built on the sound diagnosis of human and machine intelligence
systems, the I/O complex should represent a synthesis of cause and effect, and a unity of plan
and product.
Approved For Release 2006/09/25 : CIA-RDP73-00402R000100140010-8