PCs: Preliminary Concerns

Following are excerpts from a Sept. 11, 1982, internal memo written by Harvey “Joe” Wertz of the Information Processing Division. The letter poses interesting questions on the advent of desktop computers and sheds light on some of the issues Aerospace was grappling with as personal computers were becoming commonplace.

 

Some Observations on Personal Computers at Aerospace

The lifecycle cost of the personal computers scheduled for acquisition by Aerospace is substantially higher than one might expect. Some of the components of this higher cost are additional hardware, additional software, systems support, repairs, and “lost” MTS time. This document should not be construed as an argument against the purchase of personal computers. On the contrary, I believe personal computers and some available “non-scientific” software can be extremely useful for Aerospace purposes.

I expect, almost immediately, a clamor for hard disks and/or more memory. Although IBM has not yet announced a hard disk for the PC, they are available from several other sources. A 10–20 MB hard disk can currently be purchased for approximately $3500.

The addition of hard disks will bring on the canonical problem of shared disk space. I hope we can finesse this problem by considering the hard disk as permanent system storage and purely temporary user storage. This assumes users will back up their work on their personal floppy disks. Two events are certain: someone will lose critical work because someone (quite possibly themselves) has erased their file, and someone will erase some or all of the system routines from the hard disk.

The next problem will be software. I’m not sure how much comes with the basic system, but such things as languages, editors, “Visicalc,” database management systems, etc., may be expected to cost $1000 per machine. It is clearly possible (but illegal without special arrangements) to buy one and copy it. I not only wouldn’t plan on it, I’d specifically prohibit it.

Where will system support come from? Aerospace cannot continue to add machines in both type and number and expect them to be supported with a constant 2915 budget (the “computer techniques” JO). It might be an interesting exercise to compare 2915 growth with the growth in number of users and machines. … who is going to furnish the support given the existence of budget? I have my doubts that getting system support from “volunteers” is going to be satisfactory to the supporters, the supportees, or the management of either!

The cost of an adequate PC is $10–12K. The choice of a 16-bit system over an 8-bit system is not super clear. The biggest advantage of the 16-bit machines is the ability to directly address a larger memory. The 16-bit CPUs are not particularly fast although they will be when an arithmetic co-processor is added. At this time, the only advantage of the IBM PC is that I understand there is a TEMPEST model available (not from IBM) for roughly $8000.

Back to the Spring 2010 Table of Contents

Go to the main article:  Highlights of the Early Days of Computing at The Aerospace Corporation

To sidebar:  Aerospace-Developed Software