Planning Considerations for Workstations

[Rivanciw@darcom-hq]

Let me first say that I have been pleased with the substance of
the messages on personal workstations. I am one of the program
managers for office automation at the US Army Development and
Readiness Command (DARCOM). I wanted to become a part of this
roundtable discussion in the hopes that I could keep abreast of
the technology of personal workstations. The messages generated
to date have proved very enriching.

Here are a couple of topics that I would like us to discuss:

COST/BENEFITS:

It seems that many of you are interested in this topic
based on the number of messages making mention of prices.
What most planners seem to forget is that there are too sides
to cost/benefits - the costs and the benefits. A lot of emphasis
over the past several years has been geared to an evaluation of
the benefits of office automation. Ther is another side - the
cost.

DARCOM is taking an aggressive role in personal workstations
from the cost/user aspect. I am confident that the benefits to
managers, professionals, and administrative employees will be evident.
What we are concerned with now is the cost of providing this service.

Our initial goal was to provide a workstation for Office
Automation users at a ONE TIME cost of $5K. Quite an aggressive
goal in 1976. We now have available micros that can provide
electronic mail, calendar systems, tickler functions, meeting
scheduler, word processing, electronic filing system, suspense
tools, and some other tools to a user for a ONE TIME cost of
$5800.00. Of course, system maintenance would be an annual cost,
but that will always be around with any system.

Now, when one can get all those services for under $6K
purchase price/user, a whole lot of the cost/benefit problems
are moot. On a pure lease deal, it is now possible for a
user in DARCOM to "try" office automation for a very low price.
If the user decides the service is worth $5K they buy it. If
they decide it is not, they quit leasing having made a very
small expenditure.

The demands for office automation, based on this low
cost methodology, have grown astronomically in DARCOM. We
can no way supply all the OA users with systems. Therefore,
in the rare instance when a user finds he/she does not want
to continue service, the equipment is immediately picked up
by another user.

PERSONAL WORKSTATION ARCHITECTURE

The key to DARCOM's OA success has been an aggressive,
carefully thoughtout architecture for the personal workstation.
Several messges mentioned that a large computer should be
used to process the big runs to keep the samll micro from
becoming overburdened. That's communications. Several messages
asked the question of whether the personal workstation should
be a small version of a big machine, just running slower.

DARCOM's architecture has addressed both of these issues.
Our personal workstations run the SAME software as our large
CPUs and Minicomputers. The SAME.

Likewise, our personal workstations communicate with
larger machines to process large runs, querry central
data bases, or send documents/mail. Right now I am composing
this messages many communications links away from the ARPANET.
My computer is NOT a host on the arpanet. However, when I give the
send command to my mail system, it gets my message to the ARPANET.
And if I want to send a message to someone on another computer that
is not on the arpanet it will get it ther also. From the user's
viewpoint the command is the same. We have a RELAY computer that
handles the multi-port communications requirements in the background.

At any rate, what I am trying to say is that the key to
a successful implementation of personal workstations in office
automation is not just the bells and whistles of one system vs
another. The key is an well planned, long range, flexible
architecture that allows for the extensive amount of communications
necessary to support today's office.

Randy