Information Processing Managers Association Newsletter banner

December 1987

Monthly Association Meeting

January 7, 1988
Kings Table
1818 E. 4th

12:00 Noon

Speaker: Bob Marlett Executive Consulting Group
Topic: "Strategies for Systems Development"


We want your input!

The IPMA Newsletter staff is always looking for articles of specific interest to our State information processing community.

Send your drafts to:
Kathy Marston, Editor
DIS - 1310 Jefferson
Mail Stop: PE- 11


Table of Contents

December 3 Meeting Min.
Alphabet Soup
Forum Statistics
"Superconductivity - Revolution or Just Research"
November 24 Board Min.
December 16 Board Min.
Identifying Data Entities


AGENDA January 7, 1988

  1. Introduction of Guests
  2. Guest Speaker
  3. Bob Marlett
    Executive Consulting Group
    "Strategies for Systems Development"
  4. Approval of Minutes
  5. IPMA Board Report
  6. ISB Announcements
  7. Old Business
  8. New Business
  9. Correspondence
  10. Other Comments/Announcements
  11. Adjournment

December 3 Meeting Minutes

Darrel Riffe, Chairman, opened the meeting at 12:00.

Luncheon Speaker

Program Chairman, Jeff Boyce introduced Darlene Hildebrandt, Administrator of the University of Washington Computing Information Services at the Academic Computing Center, who spoke about computing resources. Ms. Hildebrandt indicated that they have resources that are very broad based and directed toward all aspects of the computing arena. Resources are available in a variety of formats including books, publications, audio and video. The fee structure is designed to complement the level of the research requirements. For more information on these services call Ms. Hildebrandt at (206) 543-8519, University of Washington.

O1d Business

IPMA is still working on a revised DP management matrix. Leonard Roberts is the official IPMA representative.

Chairman, Darrel Riffe announced that summary statistics about this year's Fall Forum will appear in the next newsletter.

New Business

Name tag price issues are still being researched.


Alphabet Soup

We had several responses to last month's Alphabet Soup game. Cliff Gott pointed out that the acronym PSF stands for Print Service Facility. Employment Security staff submitted these creative responses to the bogus acronyms; PSF = Pronoun Spellchecking Facility, APY = Apiary, a natural language for insect lovers. One of our other readers submitted a new list of acronyms to challenge your skills. There are no fakes in this list. Let's see how you do.

SQL   OSI
VGA   SMB
RISC   SDLC
TFT   APA
ESD   INIU

Answers are at the end of the newsletter.


Forum Attendance Statistics

Following are highlights of the statistical summary that Ruben Marti compiled from the registrations and the 75 FORUM questionnaires that were returned.

There is considerably more information and statistics available than is represented here. If you are interested in more detail, please contact Kathy Marston (586-2832) or Ruben Marti {586-1348) for a copy.


"SUPERCONDUCTIVITY - REVOLUTION OR JUST RESEARCH?"

-- Submitted by: Jim Michal, CEP Partitipant

Scientists have known since 1911 that some materials lose all their electrical resistance when chilled to absolute zero - which is minus 459 degrees Fahrenheit. Even in a good conductor like copper wire, the speeding electrons that make up electrical current usually waste energy ricocheting around. In a 'superconductor,' the electrons skip among the frozen atoms without colliding.

If superconductivity could be achieved in a practical manner, it could revolutionize the way we live. It could make possible shoe box-size supercomputers, inexpensive electric cars, 300 mph magnetic trains that float on a cushion of air 5 inches over the tracks and incredibly cheap electric power.

In late 1986, two IBM scientists (Professor K. Alex Mueller and J. Georg Bednorz) in Zurich discovered a new solution to one of the toughest problems in physics; finding a cheap, simple substance to conduct electricity without resistance. The substance they discovered is a new class of ceramic composed of copper oxide, laced with barium and lanthanum.

What is remarkable about their discovery is that this substance is normally an insulator that blocks the flow of electricity because its electrons bind tightly, each to its own atom. The IBM scientists discovered that chilling this new substance altered the electrons causing the atomic forces to tug the electrons smoothly from atom to atom. Even more remarkable was that they achieved superconductivity at minus 396 degrees Fahrenheit.

But then in February, 1987 Professor Paul Ching-Wu of the University of Houston startled the scientific world by reporting that using a variation of the IBM scientists formula (by mixing the copper oxide with materials such as yttrium and barium carbonate) he achieved superconductivity at the amazingly high temperature of minus 283 degrees Fahrenheit. At this temperature, the ceramic materials can be cooled using liquid nitrogen rather than liquid helium, which is significantly more expensive. Overnight the impractical suddenly became practical. Using superconductive power lines cooled by liquid nitrogen, Alaskan hydroelectric stations could power Los Angeles, nuclear reactors in the middle of the desert could send power to both coasts and power could be stored in superconducting generators, with no energy loss, saving billions of dollars.

Since Professor Ching-Wu's discovery, a frenzy of research has shown evidence of the possibility of superconductivity at higher temperatures, even as high as room temperature - around plus 70 degrees Fahrenheit. However, the arduous task on transforming lab experiments to commercially available products is apparently proving difficult. The new ceramics are finicky.

Engineers wonder how to make the brittle material into wire that meet requirements for flexibility, s~e and strength. Electronic experts report that electricity flows adequately horizontally, but hardly at all vertically impeding efforts to make the material carry large amounts of electricity.

Despite the fabrication problems, most superconductor experts say that it will take five to l0 years to develop the new ceramics for large commercial applications, many of which probably haven't yet been imagined. That may seem like a long time, but it would be about half what was needed to move lasers and computers from the lab to major commercial use. Regardless, these recent breakthroughs promise to affect virtually every USA industry, firm and product line. It could be as significant as the transistor.

Note: Superconductivity is the basis for particle accelerators, called 'supercolliders.' Washington is among several dozen states making bids to capture a proposed new lab estimated as a $4 billion to $6 billion project. (At present, these supercolliders do not use the new ceramics mentioned in the article.)

BIBLIOGRAPHY

'Research Electrifies Industries', John Hillkirk, USA Today, April 22, 1987.

'Fabrication of Products From Superconductors Poses Major Challenge', David Stipp, The Wall Street Journal, July 9, 1987.

'Superconductivity Electrifies Japan Inc.', Stephen Krider Yoder, The Wall Street Journal, August 12, 1987.

'How 2 IBM Physicists Triggered the Frenzy Over Superconductors', Richard L. Hudson, The Wall Street Journal, August 19, 1987.

'Rush to Exploit New Superconductors Makes Japan Even More Patent-Crazy', Stephen Krider Yoder, The Wall Street Journal, August 27, 1987.

'Superconductor Research Is Heating Up', Jim Erickson, Washington State Energy Office - Dispatch', July/August 1987.

'Superconducting 'Thin Film' Tests IBM's Talent, Patience', Dennis Kneale, The Wall Street Journal, September 11, 1987.


November 24 Board Meeting Minutes

Darrel Riffe, Chairman, opened the meeting at 11:00 a.m. with a quorum present (N. A. Stussy, Gary Longmire, Jeff Boyce, Glenn Medeiros, Kathy Marston, Phil Coates and Leonard Roberts).

Darrel presented a proposal to clarify the service award procedures. His proposal was discussed and adopted as a board resolution.

Phil provided an information update on the FORUM. He reviewed the results of the two questionnaires, conference attendees and vendors. The board discussed the questionnaire findings and assigned a special committee the task of formulating recommendations regarding the 1988 FORUM.

Leonard gave an overview of the November 5th meeting he facilitated regarding the role of the IPMA standing committee of senior managers. Given the results of the meeting, he plans to develop a draft charter for the standing committee. He also encourages the IPMA to actively pursue development of a skills inventory. Gary will take the lead roll in developing a skills inventory in the near future.

Darrel indicated that he planned to set up a meeting with Nancy Abraham, DIS Director, to introduce the IPMA Board members and discuss IPMA's role.

The meeting adjourned at 1:10 p.m.


December 16 Board Minutes

By: Tara Wolff, IPMA Secretary

Darrel Riffe, Chairman, opened the meeting at 11:00 a.m. with a quorum present (Gary Longmire, Leonard Roberts, Phil Coates, Kathy Marston, Jim Andersen and Tara Wolf0.

The meeting started with a discussion of the December 9th meeting with Nancy Abraham, DIS Director. The board highlighted a number of topics raised at the meeting. These included:

Phil Coates presented FORUM recommendations for 1988. The recommendations were approved by the board. The board will nominate a member to the 1988 FORUM committee at the January board meeting. In addition, the board asked Phil to continue to serve as the IPMA representative for six more months to assure continuity.

Gary Longmire presented a proposed outline for the IPMA skills inventory for board review and comment. Members will forward their comments to Gary prior to the January board meeting.

Leonard Roberts presented a draft of proposed DP classification changes. The board has nominated Leonard to be the official IPMA representative to the Department of Personnel on the proposed DP manager matrix changes and the establishment of a Computer Information Assistant classification.

Chairman Darrel Riffe adjourned the meeting at 1:10 p.m.


A special interest group on Artificial Intelligence is meeting monthly. The group has met three times with presentations by David Temple from Arthur Anderson & Co. Anyone interested in attending future meetings may contact Dr. Ruben Marti at 5861348 for information.


Identifying Data Entities

-- By: Michael Brackett, WSU

INTRODUCTION

The database design direction for the 80's and 90's appears to be the relational model and its associated techniques such as entity-relationship diagrams, data normalization and optimization, data naming and definition, integrity rule definition, etc. There is nothing foreseen that will completely replace the relational model, although enhancements wfil surely be made. Therefore, it would be relatively safe to plan on using the relational model for the design of corporate databases.

The relational model consists of three major features: structural, integrity, and manipulative. Database design addresses the first two features by defining the structure of the database and the integrity rules to control that structure and its contents. Database management systems (DBMS's) address the last two features by endorsing the integrity rules and providing the manipulative features to navigate through the physical database and store or retrieve data.

The major database design effort to date has been oriented toward defining the structure of the database, i.e. the data entities, the relations between those entities, and the data attributes that characterize each entity. Little or no attention has been paid to defining the integrity rules for controlling that structure and its contents. The major database management system effort to date has been determining whether or not a particular database management system meets the criteria of a true relational system, particularly the integrity rules and manipulative features.

All three features of the relational model are needed to have an effective, efficient corporate database, but there is a definite sequence in which that database must be defined and developed. The structure of the database must be defined first, followed by definition of the integrity rules to control the structure and its contents, followed by implementation and use of a database management system to implement the rules and provide the manipulative feature. However, before the structure can be completely defined there are several issues that must be addressed.

STRUCTURAL ISSUES

The definition of the corporate database structure has been largely oriented toward the mathematical aspects of the relational model and normalization theory. Although this technical orientation is certainly important, it is not the only consideration in developing a corporate database. The social issues of how an enterprise views the business world and how it operates in the business environment play a very important role in how the corporate database is defined and structured. Both the technical and the social aspects must be considered when designing the corporate database.

Since the technical aspects are known by most people the question becomes "what are the social aspects of database design?" To answer this question requires the definition of two terms: relation and subject. A relation is a table of data that has been defined by the mathematical techniques of the relational model and normalization theory, specifically normalization of data based on primary keys. A subject is a collection of data about a group of related occurrences in the business world as viewed by an enterprise. The combination of mathematical relations and social subjects become the data entities used in database design.

For instance, Enterprise A labels all of its equipment with a tag containing a unique equipment number that is permanently affixed to each piece of equipment. This unique equipment number is used as the primary key and normalization says that equipment is a relation. Enterprise B tracks vehicles by the license number, electronic equipment by the manufacturer serial number, and office furniture by a tag permanently affixed to each piece of furniture. Each of these numbers becomes the primary key for its respective group of data and normalization says there are three relations.

A pure technical database designer might say that this is not possible. One of those enterprises must be in error and should change their primary key so that both enterprises treat equipment the same way. Non-technical people from each enterprise, such as a manager or supervisor, would claim that the structure defined for their enterprise is the way they manage the equipment and that it works quite nicely. This non-technical approach is an example of the social aspects of defining the structure of the database.

Which definition is correct, the technical or the social? Is the equipment in the example above really one relation or is it three relations? The answer is that both structures are correct because the first enterprise views their equipment as one subject since they manage all equipment together, and the second enterprise views their equipment as three subjects since they manage the equipment as three subjects. In addition to these business views, the primary keys mathematically support both the one relation and the three relation structures.

Therefore, subjects are the way an enterprise views the business world and relations are the mathematical substantiation of those subjects. The two must go together and the result is a set of data entities that form the corporate data architecture.

DESIGN PROCEDURE

Design of a corporate database follows a definite pattern from strategic modeling, through retrofit modeling and project modeling, to physical implementation. Strategic modeling sets the basic framework of the corporate database, its data architecture. Retrofit modeling adds detail to this basic framework based on a review and evaluation of existing data files. Project modeling adds all the final details to the framework that are necessary to construct, implement, and use the database.

From this description it should be obvious that the basic data architecture is set during strategic modeling. This is where the mainline data entities and the data relations between those data entities are defined. The strategic model is the place where the basic decisions are made about the subjects of data that an enterprise manages and the primary keys to support those subjects. This is where the technical and social issues come together to define the backbone data entities.

The first step in strategic data modeling is to define the subjects of data that the enterprise manages, such as employees, customers, equipment, building, vehicles, etc. This is best done with key personnel from the enterprise that have a detailed knowledge about the business itself. A good set of business models will also help to define the subjects of data by identifying the real world occurrences the enterprise encounters. In other words, vehicles are occurrences, customers are occurrences, buildings are occurrences, etc., that the enterprise encounters and manages, and they become the subjects for which data is collected and stored.

Once these subjects are defined from the business side, a unique identifier is defined for each subject and technology is used to assure that they are mathematically unique subjects. Obviously, there is not always a perfect match the first time and some adjustment must be made. But, this is the nature of modeling, and the subjects and unique identifiers are repeatedly adjusted until there is harmony between the way the enterprise views the business world and the mathematical substantiation of that view.

Once this basic architecture is set, retrofit and project modeling follow the same general procedure to add detail to the strategic model. Existing data are reviewed and grouped by subject with each subject being verified mathematically. If existing data belongs to subjects already defined on the strategic model, there is no problem. However, ff the existing data belongs to another subject not defined in the strategic model, then the same balance between subject and relation described for the strategic model must be achieved before a data entity can be defined. The same procedure holds true for the project model.

CONCLUSION

This approach is the only way that a successful corporate database can be designed and developed. Defining subjects without mathematical substantiation leads to groups of data oriented toward business events or applications and results in the same problems that exist with the current application files. Defining relations without the business perspective leads to groups of data oriented solely toward the mathematical aspects of defining primary keys. Either approach will result in a database that is less than effective and efficient.

A good database design must include both the social and the technical aspects of design. The social aspect defines subjects of data based on the way an enterprise views the business world. The technical aspect verifies those subjects based on unique identifiers for each occurrence in the subject. The result is a structure o? data entities lhat, when implemented, will provide a very effective and efficient corporate database.


Answers to Alphabet Soup

SQL Structured Query Language
VGA Video Graphics Array
RISC Reduced Instruction Set Computer
TFT Thin Film Transistor
ESDI Enhanced Small Device Interface
OSI Open System Interconnect
SMB Server Message Block
SDLC Synchronous Data Link Control
APA All Points Addressable
NIU Network Interface Unit