Wednesday, June 4, 2008

A Personal History of Computing, Part 5

Leaving Little Town on the Prairie

It wasn’t just the turn down of my proposal to invest in computer technology (see Part 4) that led to my leaving Little Town on the Prairie in the late 1970s. It was a collapsing rural economy.

My cousin Roger, a good small town jack-of-all-trades lawyer and County Attorney, thought our small burg might thrive as a retirement community. On the frozen tundra of the northern plains? I thought differently. I sold the real estate and property management business to the bank that refused the loan; our house was the last residential sale in about four years.

That was the summer I ran for State Senate. One foot in the district and one foot out. Not a profile likely to be successful...but still surprisingly close.

We ended up in St. Paul, I took a few courses at the University of Minnesota and made some money trading stocks. I might have been the first “day trader,” except the trading cycle wasn’t “day,” because – wait for it – there wasn’t (personal) computer access to markets.

Big Oil Productivity

After some time, I went to work for a “management consulting” company, Keith-Stevens, Inc., that specialized in maintenance productivity. Work orders, reports, backlog management, FTE’s (that’s Full Time Equivalents). We could usually cut some fat out – there was a lot to cut – but the changes weren’t often lasting. Short term improvement that faded after time – a pattern repeated through many other companies and many more years to come.

The Killer Ap

My first attempt at “automation” was designing work control forms – work orders and backlog reports – using early spread sheet functionality on a Compaq “portable.” (Had Compaq gotten a deal on the molds for Singer sewing machine plastic cases?) I started that at Shell Oil in McCamey, Texas. Did more of it at Mobil Oil in Bakersfield, California.

By then I was the Account Manager for Texaco, Shell, Unocal, Gulf, Chevron. For the next few years I was in most of the major oil fields and refineries in North America, overseeing consulting gigs.

RIS and VAX

One of those was RIS – Refinery Information System – for Gulf Oil’s refinery in Port Arthur, Texas. It would be the prototype for all of Gulf’s refining and chemical production facilities.

There were two teams at Gulf; one focused on work practices (what people actually did in their jobs) and the other on coding a system that would support those practices, providing on-line tools to initiate work requests, and to plan, schedule, and then summarize the completion of work. I was the overall manager. We were writing in INFO, a prototypical non-procedural 4GL (fourth generation language) on a DEC VAX 750. When the system went live, it would reside on multiple 750s, clustered with DEC’s HSC 50 clustering hardware.

We finished that, and implemented it, but INFO turned out to be a dog. Most of the system was rewritten in Fortran. (I suspect many young computer geniuses have no idea what Fortran is.)

That was around 1982. We were on the cusp of such systems. They’ve since become multimillion dollar investments. Hell, I guess they were then, too. More on that later.

I became an Assistant Vice President (sort of like a bank, wasn’t it) at Kieth-Stevens. They dumped me Christmas 1984; in that I followed a string of predecessors. By January 1985, I was up to my eyeballs in the commercialization a maintenance management-inventory control-purchasing software product at EMA, Inc., in St, Paul.

That’s Part 6.

No comments: