Volume 4 No. 1 July 1977
NEXT ASSOCIATION MEETING
Date: August 4, 1977
Location: Golden Carriage
Time: 12:00 Noon
Topic: "What's It Like to Run a State Capital City"
Speaker: Tom Allen
Mayor of Olympia
DISTRIBUTED DATA PROCESSING COMMITTEE
Several months ago a committee composed primarily of WDPSC users was formed to determine the requirements for distributive data processing equipment in state agencies. Members of the committee are:
Twila Perry General Administration
Norm Gunther Department of Natural Resources
Mike Phillips Liquor Control Board
John Aikin The Evergreen State College
Cliff Cotey WDPSC
Bobbi Giovannini DPA
Bill Fischer Department of Revenue
Jim Andersen Department of Ecology
Don Smith Labor and Industries
Kepi Dahl WDPSC
Dick Nelson Utilities and Transpiration Commission
The committee soon discovered that distributive data processing was not defined well enough to allow the development of system or hardware requirements. Presented below is the committee's definition of distributive data processing and a proposed guideline for the application of distributive data processing by state agencies. Your comments related to this definition and proposed guideline are welcome.
Distributed Data Processing, a working definition for application by state agencies.
Definition of Distributed Data Processing:
Functional Definition
Distributed processing denotes the distribution of data and processing capability at the most appropriate geographic, organizational and functional locations within an integrated systems architecture to satisfy end user requirements.
Hardware Definition
Distributed system utilize the best attributes of centralized computer facilities and dispersed minicomputers in single systems tied together through communication networks.
Amplification:
Distributed data processing is:
Distributed data processing is not simply:
Guidelines:
Distributed data processing can be a viable, cost effective computing alternative. It should be considered when examining alternatives to satisfy new data processing requirements. It should be recognized that applications must be defined and specifically designed to make effective use of distributed processing. Distributed data processing tends to be best suited for applications in the following categories:
1. Control systems where the computer is used to control such things as; security, environmental conditions, machines, or traffic flow.
2. Simple transaction systems such as, point of sale, attendance reporting, reservations, circulation or testing.
3. Special purpose systems where a highly formatted set of data is used repetitively.
4. Systems in support of statistical analysis, technical computation, table lookup, directory assistance, text editing, and conversion calculations where the programs, tables and calculations are of relatively small size.
5. Systems that require local access and updating of local data bases.
In general, distributed data processing is not well suited to large production systems with large amounts of input/output and large data bases. Some of the key factors to be considered in evaluating applications for distributed processing systems are:
In most cases users will find that a combination of centralized host processing and data storage, and distributed data processing offers the widest range of capabilities and the most effective solution to their data processing needs.
Distributed data processing is not a total replacement for centralized data processing. The objective is to distribute the functions to achieve the most efficient and effective processing environment. This distribution must consider the effectiveness of the end user--machine interface, Offered below are some guidelines for functional distribution.
Centralize processing power to obtain economy of scale, ease of coordination, tightness of control and security.
Centralize processing power to provide a facility for development and operation of state interagency systems,
Centralize processing facilities to obtain a focal point for the interface of agency systems to state systems.
Centralize file organization for efficient use and control of magnetic storage.
Centralize the host computer for efficiency of the terminal network.
Decentralize input, placing the error detection and correction at the source.
Decentralize output to place information squarely in the hands of the user,
Decentralize intelligence into user locations for better local control, and economic preprocessing and post processing.
The requirements for a distributed data processing system should be developed in close coordination with the intended host computer center to insure that the selected distributed processor will be properly supported and integrated into the total system architecture.
Cliff Cotey,
WDPSC
TRAINING PAR EXCELLENCE
Do you know the difference between asynchronous and synchronous transmission? Or how to create a teleprocessing network that is both cost effective and efficient? Or that baud rate and bits per second are not the same?
On June 20 & 21 the Pacific Northwest Bell Company conducted a course on Teleprocessing and Network Planning and Concepts. All but two of the participants were Washington State Employees. The course was organized better than any other course I've been aware of, Its productiveness should be a big shot in the arm to state agencies toward the enhancement of practical teleprocessing concepts.
The course covered 2 full days. There were 3 instructors from PNB. The course was organized to be educational, practical, entertaining, and enlightening. Not only were current technological concepts discussed but also what lies ahead.
There were demonstrations covering the use of telephones as terminals, PNB's present line of terminals, and lesser beam technology. Animated cartoons announced the beginning of coffee breaks, stretch breaks, and lunch, in a manner making the 14 hours of course material as painless as possible; even enjoyable.
The state does an excellent job in its training program. Even so, I'm certain that state training personnel could pick up a few pointers from PNB's approach.
Dick Nelson
WUTC
IBM JCL - XREF
Have you ever wanted to know what the system-wide usage is of a given data set or program? If you occasionally have a problem coming up with the answer, you might be interested to know that this information can be placed at your fingertips.
A small utility procedure has been developed, mainly through the efforts of like Phillips, of the Liquor Control Board, and myself, that creates a variety of useful JCL cross reference and other listings. The utility has been used, as of this writing, by six (6) different agencies with evident satisfaction.
The procedure has the ability to concatenate card input JCL, Panvalet JCL and your procedures in a procedure library. It then uses the above system-wide data to create the following reports:
1) A listing of your non-catalogued jobs.
2) A current listing identifying your members in a PROCLIB,
3) A current listing of all the above members.
4) An "abbreviated" JCL list emphasizing the chronological usage of programs and data sets within each procedure or job.
5) A system-wide program usage listing with program as major and procedure as minor.
6) A system-wide data set usage listing with data set name as major, procedure or job as intermediate and program as minor. This procedure internally resolves symbolic parameters and back references for those that take advantage of this flexibility.
Information as to how to obtain this procedure without costs can be had by calling Mike Phillips at 753-6232 or myself at 7531472.
Howard Glastetter
Dept. of Natural Resources
COMPUTERS AND HUMAN REASON
by Joseph Weizenbaum: A REVIEW
Dr. Joseph Weizenbaum is a computer scientist at MIT whose work includes the by now famous program ELIZA, which "simulated" a psychiatric interview using a computer as "therapist." He has just published a book which is of major importance to the practitioners of data processing. The book is titled Computers and Human Reason--From Judgement to Calculation and is published by W. H. Freeman.
Dr. Weizenbaum's book is controversial: in part because he argues that there are limits to what computers can do in the way of simulating human intelligence, but much more deeply so in that he argues that there are limits to what computers ought to do. Many of those who have been saying that in the "visible future" (to use Dr. Herbert Simon's phrase) computers will be able to mimic every intellectual accomplishment of man will be outraged. Weizenbaum offers evidence that machine "intelligence" (however great it may become) differs qualitatively from human "intelligence" and that thus computers can never simulate the "whole man."
To use a simple analogy, birds and jet planes both "fly," but to say that the jet plane is therefore a faithful representation of the bird is foolish. Something of the bird has been lost in the jet plane (and, of course, something of the jet ply-ne is lost in the birds). The kind of "thinking" of which computers are or could be capable is just not the same as that produced by the human brain in the human body. Why? Well, for one thing the human is socialized by humans in a human culture. The growing child learns to deal with human problems which have human solutions that are not just computations on strings of numbers. An intelligent computer, if one were possible, would not share that socialization and therefore no matter how "smart" could not see the world as humans do. Likewise, computers do not have human bodies; even robots would not have human bodies. A lover's touch or an angry blow could not possibly be the same for such a machine.
In short, one of Weizenbaum's central tenets (and one must read the book to grasp all his arguments) is that an intelligent computer (if one is ever created) would have an alien intelligence. It would be as different from human intellect as a dog's intellect is different from a man's. Intelligence is not just a question of quantity, but also of quality.
Weizenbaum neatly sidesteps the question of how intelligent a computer can possibly become by saying that nobody knows. That is a question which will be answered (if it can be answered at all) only through research; through trying to build faster, bigger, and (maybe) more "intelligent" computers. The heart of this argument is that it doesn't matter: no matter what kind of computer we can build, we ought to think more carefully than we have about what to do with it.
Weizenbaum analyzes the growth and impact of technology generally and computers particularly on our society in the last thirty years and he is not impressed. Through a series of fascinating examples he shows how the tools available at any time have dictated the form of the solution and introduced their own new problems; how human solutions to human problems have been increasingly ignored in favor of technological solutions.
For example, it is often asserted that the banking system would have collapsed from the onslaught of paperwork in the late forties and early fifties if computers had not arrived "just in time." It is true that the volume of checks to be processed each night was growing and it was clear that even hordes of clerks could not keep up much longer. But consider: what would really have happened if the computer had not been invented? Would banking have collapsed? Of course not. Some other way would have been found to solve the problem. We don't know what kind of solution--it might have been better or it might have been worse, but it surely would have been different. And as Weizenbaum notes, such solutions were not even considered precisely because technology (the computer) offered a solution. The choice of that solution has brought with it a whole series of side effects and foreclosed other solutions that might have been better for society as a whole. We will never know.
Weizenbaum recoils at some of the "uses" which are now beginning to be proposed for computers: the computer as judge and jury; the computer as doctor; the computer as teacher; cyborgs (combinations of computers and animals) which Weizenbaum labels "monstrous." In the end, Weizenbaum is pleading for all of us to address that basic question: what kind of society do we want to live in? Shall we use technology as our tool, or shall we become technology's tool?
These are gut issues and those of us who are the high priests of this new technology--data processing managers, programmers, and analysts--cannot abdicate our responsibility to step back and say "Wait a minute. Just because the computer can do this, does that mean it should? What other choices do we have?"
I recommend this book to everyone who works with computers or who has a fascination with science. You will not agree with everything Weizenbaum says, but the book is more than worth reading. And the author's style, although well laced with big words, is engaging to say the least.
John Aikin
The Evergreen State College
PERSONNEL LIAISON COMMITTEE REPORT
Personnel Liaison Committee for fiscal year 1977-1978. Members of the committee are:
Clifford Cotey WDPSC Chairman 3-4360
Paul Newman DSHS 3-2208
Larry Seaberg SC #3 3-0987
The Department of Personnel has reassigned Fern Morrow. She will no longer be responsible for the data processing classes. The committee would like to express its sincere appreciation to Fern for the tremendous support she has given us over the last several years.
Effective July 1, 1977, Gary Hull is the Personnel Analyst responsible for the data processing classes. Gary's number is 3-2862. Welcome to the world of data processing Gary.
The committee will address the following areas during the year.
If you have any comments, or additional areas of concern please call any of the members of the committee.
Cliff Cotey
WDPSC
Thanks to all of you who supported me for Chairman of the Association for the upcoming year. With your continued support, I believe we can make this a very productive year.
Over the past couple of weeks, I have given much thought to what I would like to do as your new chairman. Some of the things, I would like to see accomplished during the upcoming year are:
JOB MATRIX:
Either drop it or complete implementation. This task has been assigned, with a highest priority, to the Personnel Liaison Committee. Cliff Cotey will be this year's Chairman.
D.P. SALARY BENCHMARKS:
A new salary survey will be conducted by D.O.P. this fall. The Personnel Liaison Committee will review existing benchmarks and will identify required changes.
ADPM NEWSLETTER:
This has been one of the more important communication devices for the Association. Patti Palmer will be this year's Editor. Patti has many good ideas for producing the NEWSLETTER.
IMPROVE ATTENDANCE:
I am not so concerned about the number of attendees, but rather, who attends. It's distressing that many key top data processing managers never attend our monthly meetings. I hope I can change this situation.
ASSOCIATION BYLAWS:
It bothers me that there are many highly qualified data processing managers who will probably never have the opportunity of holding office because of the Associations one voting member concept. A committee will be called to review our Bylaws and to make recommendations.
I am sure that there are many other tasks or goals that the Association should address during the upcoming year. Please feel free to call me at 3-2957 if you have any suggestions or thoughts.
Jim Anderson
Ecology
The DPA is asking agencies to update their long-range ADP Plans. Submittal dates for annual reports of plans or plan updates are October 1, 1977, for agencies and December 1, 1977, for designated Service Centers. An added planning element is called for, identifying agency performance measures or cost benefits for major systems developments or acquisitions.
Bobbi Giovannini
DPA
The DPA has changed its regular meeting location to the Board Room, First Floor, Administrative Building of Olympia Technical Community College. The meeting dates are still the first Wednesday of each month at 2:00 P.M.
Bobbi Giovannini
DPA
JULY 7, 1977
The meeting was called to order at 12:30 by Chairman, Jim Anderson. There were 24 members and guests present.
The new officers for the coming year were introduced: Chairman, Jim Anderson; Secretary/Treasurer, Bobbi Giovannini; Program Chairman, Pat Mailey; Executive Committee Member, Don Brown; Past Chairman, Tom Jones.
The Treasurer s report was given by Bobbi Giovannini. The June balance was $99.28. During the month, $50.15 was paid to Tony' s Jewelers for plaques for the outgoing officers and $3.14 was spent for July's guest speaker's lunch. This leaves a new balance of $45.69.
Mr. Leroy Hittle was then introduced as our guest speaker. He has been a member of the Liquor Control Board since 1967. His presentation gave us his management perspective of data processing. Following are excerpts from his speech.
"I view data processing as an electronic file case and a sophisticated adding machine. Like a file case you can only take out what was put in. It is also important to take a periodic inventory of what you are maintaining in that file and what reports you are requesting. What is really necessary to keep? Any thing no longer necessary should be removed just like in a regular filing drawer.
Let me briefly describe the Liquor Control Board. There are three (3) members of the Board who are appointed by the Governor to serve 9 year terms. The Board has two major functions; regulatory and merchandising. Regulatory includes the licensing of the private manufacture and distribution of alcoholic beverages. It also includes enforcement of the state's liquor laws.
Merchandising includes the wholesale and retail sale of liquor. The Board operates 326 liquor stores throughout the state.
The Board provides an important source of revenue to the state. Revenue of $96 million is distributed to the state, cities and counties and to the two universities. In fact, liquor revenue ranks fourth as a source of funds to the state and third to the cities and counties.
The Board makes extensive use of data processing in its operations. Many of these operations would be almost impossible without data processing. In the regulatory area data processing maintains the license records. In merchandising, there is heavy usage of data processing in the operation of the warehouse, in purchasing stock with automatic reorder paints, etc. and in replenishing stock in the various stores and agencies, and in the loading and scheduling of the delivery trucks.
The Liquor Control Board is looking toward an exciting future with the development of point-of-sale applications. Last year a test of point-of-sale was made in an Olympia Liquor Store. The results were dramatic. Inventory was reduced from 113% of sales to 67% of sales while maintaining a service level in-stock of 99½%.
Before Point-of-Sale can be implemented three (3) major conditions must occur;
There has been good progress on the first two tasks and we are now drawing up specifications for the necessary equipment. We are excited about the future and this data processing application.
It is important for data processing people to remember that data processing shouldn't run the business. They have a tendency to talk in buzz words which sets them apart from the others in their organization. Data processing people need to make the managers aware of the capabilities of data processing--what can be done. Communications is essential. Buzz words or technical jargon make real communication difficult."
The chairman then called for old business. Bobby Giovannini gave the report on records management committee. The committee consisting of Rich Green, Sev Hurlen, Bill Lundberg, Roy Leicht, Phil Meek, Dick Nelson and Sherry Story drafted a records retention schedule. This was then circulated to the data processing managers for review and comment. The committee has incorporated the suggested changes and have now submitted the finalized draft to D. R. Ellenwood of the State Archives office. The task group has completed its job.
Don Brown gave a report on the Computer Report Audit Committee. Don has meet twice with Joe Coogan to brainstorm an approach to this project. Other committee members are Pat Mailey and Bill Lundberg. There will not be much activity in this committee during July since some members will be on vacation.
Jim Anderson then thanked Paul Newman for his efforts as Chairman of Personnel Liaison Committee. He then introduced Cliff Cotey as the new Chairman of the committee with Paul Newman and Larry Seaberg remaining as members. Cliff then outlined the planned activities for the coming year. (These are detailed elsewhere in this newsletter). Special emphasis will be given to the Job Matrix and to the new salary survey which will be conducted next Spring. Garry Hull was introduced as the new Personnel Analyst for data processing classes in the Department of Personnel. The committee will be working closely with him during the coming year. Cliff also announced a meeting of the Personnel Board on July 8, 1977, to discuss the implementation of the pay raise approved by the legislature.
Jim Anderson than introduced Patti Palmer as the Editor of the Newsletter. Patti stated that the newsletter will be published monthly with articles due by the 10th of the month so that it can be published by the 25th, Pat Mailey and John Aiken will alternate as feature writers. Patti requested articles from the membership. Jim Anderson also requested volunteers to serve as assistant editor.
Jim Anderson then moved on to new business. He announced that Fern Morrow was reassigned from the data processing classes. Jim read excerpts of a letter that the Association sent to Leonard Nord to express our appreciation for Fern's excellent job and our willingness to continue in a cooperative relationship with the Department of Personnel.
Jim Anderson then presented a plaque to Tom Jones for serving as Chairman of the Association during the last year. Tom Jones then presented plaques to the other outgoing officers; Jim Anderson, Secretary/Treasurer, Bobby Giovannini, Program Chairman, and Bill Lundberg, Executive Committee Member. He also thanked these officers, plus the past chairman, Bob Vaughn for their support.
The meeting was adjourned.
Bobby Giovannini
Secretary/Treasurer
1977 OFFICERS
Executive Committee
Jim Anderson, Chairman 3-2957
Bobby Giovannini, Secretary/Treasurer 3-5465
Pat Mailey, Program Chairman 3-2872
Don Brown, Executive Committee 3-2218
Tom Jones, Past Chairman 3-5161
Standing Committees
Patti Palmer, Newsletter Editor 3-2946
Twila Perry, Newsletter Assistant Editor 3-3799
Cliff Cotey, Personnel Liaison 3-4360
ACQUIRING DP EQUIPMENT THROUGH STATEWIDE CONTRACTS
Many users and data processing managers recently have asked what equipment and software is available through statewide Master ADP Contracts. The following ten master contracts are presently in force:
Joe Coogan
DPA
STATE OF WASHINGTON
Statewide Master ADP Contracts
As of July 1, 1977
PRODUCT/VENDOR EXP. VENDOR REPRESENTATIVE
MODEL CONTRACT # DATE NAME* TELEPHONE NO.
Terminals
1. CMC/Remcom 2780 & 4780 G73-819 11/27/77 Joe Marvin 206/324-2600
2. Hazeltine 000, 1200, 2000 G74-004A 9/23/77 Wendy Johnson 2061242-0505
3. Perkin-Elmer Fox & Owl G77-012 5/17178 Bob Hampton 408/249-5540
4. Sanders (Harris Communications) 8160 & 8170 G76-005 9/01/77 Jeff Goldberg 206/363-3930
5. Tektronix 4006, 10, 12, 13, 14, & 15 G76-025 11/24/77 Richard C. Hjelte 206/575-0180
Key To-Disk Systems
6. CMC 3&5 G74-012-2 2/20/78 Joe Marvin 206/324-2600
7. Four Phase Systems 7001 & 7002 G74-012-1 1/01/78 Don Meek 206/827-9751
Phil Mustain
Key Entry Systems
8. IBM 3741 & 3742 G75-001 6/10/78 Dick Thompson 206/357-7726
Larry Anthony
Mini-Computers
9. Prime Computer, Inc. 3032, 3064, 3128 G76-025 12/22/77 William Brennan 206/454-9332
Data Base Management
10. Software ag ADABAS & ADAMINT S75-003 5/01/85 Bob Preger 415/383-5658
*For specific information about pricing, features, etc.