Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 130
Personal Computing, Not
Personal Computers
Norman M. Epstein*
At E.F. Hutton we do not believe in personal computers, but we
do believe in personal computing. This is more than a semantic
difference. It affects the technology applied, the philosophy we
adhere to, and the planning that went into our system.
Although we think our approach is a very obvious and logical
one, apparently others do not. It reminds me of the famous story
of Louis Pasteur and the maggots. In Pasteur's day people be-
lieved that maggots came from decaying meat because every
place there was decaying meat there were maggots. But Pasteur,
being a scientist, took decaying meat, put it in a large bell j ar, and
covered the bell j ar with gauze. He came back three days later and
found the maggots on top of the gauze, proving that maggots
came from gauze.
In short, the system has to suit you. And if you look around, no
one is wearing the same suit. Our organization would seem to be a
perfect cancliclate for personal computers. We have 400 branch
offices around the world and a campus in lower Manhattan with
6,000 people in 12 buildings. We have 9,000 people in the field,
6,000 account executives, and 3,000 people in support positions.
Our goal has been to automate the operation in the field. There
*Norman M. Epstein is executive vice-president and director of E.F. Hutton Group
and E.F. Hutton and Company, Inc., New York, New York.
130
OCR for page 131
COMPUTING, NOT COMPUTERS
131
fore, our approach might not work for a company with 14,000 peo-
ple in one builcling.
Currently, each of our 6,000 account executives has a single ter-
minal on a desk. These terminals are caller} branch information-
processing system terminals (BIPs). The terminal comes out of a
dual-port CPU (central processing unit), and (lata comes from two
different networks, Bunker Ramo ant! our own.
When our system is complete everyone in the firm except tele-
phone operators and porters will have a terminal. Everyone. The
multipurpose use of this terminal is the most important aspect of
our system. We have a limited amount of space on a desk, and we
are going to accommodate it by having one terminal. That one
terminal may be supplied by Data General, by IBM, by Wang, by
Bunker Ramo. But no matter who makes it, each terminal will
have access into all the databases. The secondary function of each
terminal determines the reason for selecting a particular product.
For example, the Bunker Ramo terminal is selected because it is a
market data system, the Wang terminal because it has certain
word processing capabilities, the Data General terminal because
it ties into our operations and communications.
How are our operations and communications tied together? We
start with head-end computers that create the databases and sup-
ply information at corporate headquarters. The next level down
consists of about 35 Data General Eclipses located at regional
offices around the worIcI. Why do we have a distributed network
around the world? Most people involved with personal computing
know that when you hit the button and have to wait any length of
time for a response, you're in trouble. The reason for the ctistrib-
uted network is very simple: hit the button, get a response. Obvi-
ously, when the computer is in the branch you can talk much
faster.
This part of the system provides only the data. There is nothing
fancy about the technology, no state-of-the-art breakthroughs.
We have simply recognized that if having data in a central Toca-
tion does not give adequate response time economically, we must
move the data out.
At the next level is the branch information processing system.
It is the integrated architecture of the network. It consists of 400
Data General MV-4000s, one in each branch. The MV-4000 is
larger than the computer that ran all of E.F. Hutton's communi-
cations a dozen years ago. It is basically a 2-megabyte machine, .6
OCR for page 132
132
MANAGING MICROCOMPUTERS
MTPS (mega-instructions per second), and 350 million characters
of storage. This is an awesome box, and it is sitting in a branch
office. The terminal hangs off this box, with high-speed, dual
mode printers and a local database for very rapid response time
and for personal computing.
At the bottom of the hierarchy, the terminal level, the branch
information processing system has the ability to grow whereas
the personal computer does not. And the system can grow in the
same box, for the terminals do not need to change. They are essen-
tially just a light bulb and a screen. The brains of that terminal
sits in the branch office. We estimate that one MV-4000 can ac-
commociate 30 or 35 terminals. And, in general, we have three
terminals to one letter quality printer.
Theselevels headquarters, region, branch, anciterminal- are
the fabric of our communications and data processing capability,
ant! they are all interconnected. The host complex controls the
distributed information system and the distributed information
system controls the branch processing. Users can reach around it
or through it to get to whatever level they want.
From my perspective, the frequently asked question of who
owns the data is the wrong question. We ought to ask who main-
tains the data. With our system, who "owns" it is irrelevant, since
anybody who wants it and is approved can get it. This system
makes data available to anybody anywhere in the world. Each
machine is individually adciressable and assignable and has its
own name, and each user has his or her own sign-on cocle. The end
result is that one computer communications resource is providing
information to approximately 10,000 terminals. Basically, the
hardware is secondary; primarily what we JO is provide a termi-
nal, and the person using that terminal has the smarts of the com-
puter behind it.
One of the problems we used to have involved gateways. In the
past there was only one way out of a branch office. That was via a
teletype system that was part of our network. Information was
passed from functional areas to communications areas and then
out. The new system solves that problem. Everyone who has a
terminal has a gateway out of the office.
One of the most important management issues is that beyond a
certain point you can't pay people more money to HO a better job.
And people who are capable of doing a better job won't necessar-
ily stay with a company just for more money. To solve the prob
OCR for page 133
Representative terms from entire chapter:
personal computer
COA
134
AIANAGING MICROCOMPUTERS
branch offices, but also to our customers. With a million cus-
tomers, we are looking for a reasonable response of 5,000 clients
by the end of 1984.
This system cost 40 million dollars. For us, the cost-benefits are
really very simple. E.F. Hutton is a firm on the leading edge. As
such, we cannot afford not to have such a system; we cannot af-
forc3 to be second. Our alternative is not to be in business. There-
fore, it is not a question of if, it is a question of who. We chose Data
General because its hardware is significantly more powerful than
the other computers we investigated. We received an operating
system with add-one and improvements written in to keep us
competitive.
We estimate that we can justify the cost of the system over a
period of three to five years, and we anticipate its life cycle will be
more than seven years. This is unusual because we generally blow
our computers before their leases expire. In fact, I have never
kept a computer until term. The company simply cannot afford to
keep computers that are no longer efficient and economical.
The cost, as far as I'm concerned, is dependent upon the mis-
sion. And realizing the mission of an organization raises the issue
of management control. Yet, strangely enough, in all the talk
about personal computers, we don't hear much about managers'
ability to control the work. To me, on-line personal computing
means giving away or abdicating a great amount of responsibil-
ity. And, as a businessman, ~ must wonder how I can control what
I have given away.
For this reason I view the personal computer as a dangerous
weapon and I treat it as such. I think the first and most important
question to ask when considering a personal computing system is,
can ~ control this? If the answer is yes, and I am satisfied with the
level of control I can exercise over the system, then I would
choose it. If not, I wouic3 look for something else. There is a vast
array of options from which to choose. My advice is to take the
best and leave the rest.
