SCELBI

~ the obituary ~

Nat Wadsworth was an electronics engineer who sat through an Intel seminar in late 1972 about a new part called the 8008 and understood immediately what it meant. His employer did not, and would not be talked into it. He had a refurbished PDP-8 at home in a cabinet the size of a refrigerator, bought at a tenth of list price during a DEC clearance, and he had already written an 8008 cross-assembler on it. He offered Intel that cross-assembler in exchange for about two thousand dollars of chips. Intel countered with three processors and a handful of memory parts, and he turned them down.

So he and two engineering friends put in about two hundred dollars each and bought the chips themselves. With Bob Findley he founded SCELBI Computer Consulting in Milford, Connecticut in 1973. The name is an acronym whose last two letters stand for biology, which is a strange thing to find in a 1973 computer company. The Scelbi-8H was five boards carrying 1K of memory, with cassette storage and provision to hang a teletype or an oscilloscope off it, at $565 for the kit. It was announced in a small advertisement in the back of the march 1974 issue of QST, an amateur radio magazine, which is as close as anybody has to a birth certificate for the personal computer as a product you could order.

The timing was the problem, and so was Wadsworth's heart. He had two heart attacks while the first kits were going into production, at twenty-nine and then at thirty, and the second one in may 1974 put him out of the business for five months, which was exactly when the first real wave of interest arrived. The 8H was discontinued at the end of that year and replaced by an improved 8B. Then in january 1975 the Altair landed on the cover of Popular Electronics with an 8080 in it, and every 8008 machine became a footnote in a week. Fewer than 150 SCELBI board sets and assembled systems went out in total.

The 8H stopped being sold a decade before I was born and it was a mail order product from Connecticut, so there is no reaching for this one. What holds me is the pivot, which is the most honest business decision in this whole period of computing. SCELBI noticed that its customers wanted to understand the machine more than they wanted the machine, and it sold them that instead. The 8H sits in museums now. SCELBAL and the assembler listings are still readable and still work if you type them back in. Documentation outlasts hardware by decades and hardly anybody funds it that way.

~ the case file ~

~ the funny part ~

The company that shipped the first microcomputer kit made its living selling paperbacks about microcomputers. SCELBAL was a BASIC interpreter published as a book of source listings, with optional modules for strings and transcendental functions, and it ran on 8008 and 8080 machines that had nothing else. Buyers typed it in. An assembler and editor came out the same way, as printed code with the explanation attached, and so did a run of game programs.

This is not a fallback business, it is the same business seen clearly. In 1975 a machine without documentation was a pile of parts, and the scarce thing was not silicon. Microsoft's first product that year was a BASIC for the Altair sold as a tape you paid for and everybody copied. SCELBI sold the listing and let you type it, which meant nobody had to steal it, and it kept the company alive for seven more years than the hardware managed.

~ the long tail ~

Priority arguments in computing are always messy and this one is worth stating carefully. The French Micral N shipped in 1973 and predates the 8H, but only as a finished machine for industry rather than something a person assembled. The Mark-8 appeared in Radio-Electronics in july 1974 as construction plans, so you bought a board and sourced the rest yourself. The Scelbi-8H is the one you could order as a complete kit or as a working computer, four months before the Mark-8 article, which is a narrower claim than "first personal computer" and is the one that actually holds up.

The publishing business went in 1982 and the machines had been scarce from the start, so what survives is documents, which is fitting for this company in particular. There is a SCELBI Computer Museum online holding Wadsworth's own account of the founding alongside his lab notes, with the schematics and software kept beside them, and people have built working replicas out of that. Fewer than 150 units ever sold, and the design is more recoverable today than most hardware from twenty years later.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »