Over the past quarter century I seem to have had something of a knack for betting on the wrong PC pony. Maybe I'm a contrary indicator? Perhaps I should be paid by manufacturers not to buy their systems? Or are my needs and interests and criteria too atypical for the market to fill?

Regardless and nevertheless, here's a brief chronology of my twisted path through the thickets of home computer systems:

• ~1977 — Commodore PET. With its built-in cassette recorder for program storage, 25x40 character monochrome screen, a big 8kB of static RAM, and a mighty 1 MHz 6502 processor, the PET at $800 was a decent machine for its time. Microsoft BASIC in ROM left room for quite a bit of fun programming, with gateways into assembly language and other coding environments. More practical applications were, alas, scant. So I played around with my PET (= "Personal Electronic Transactor"), did some science-fair-like projects, wrote them up, and sold a few articles to hobby computing magazines to help pay for the box. (see PetBibli1 (23 May 2000) and PetBibli2 (14 Jul 2001) One day, a year or two after I got the PET, I started seeing intermittent errors in some of my calculations. I ran a bunch of tests and finally tracked it down to a failing memory chip: a particular bit was fading out about every tenth of a second. It cost me $20 to buy a replacement — a lot of money for a starving grad student — but when the component arrived I took my clothes off (to avoid static electricity), installed the new chip, and got my bit back. Whew! Out of a naïve sense of corporate loyalty I bought a few shares of Commodore stock; a couple of years later I sold them to protest Commodore's lack of support for programmers and users. No one noticed.
• ~1982 — Atari 400/800. Mainly a game platform, the Atari 400 (with its flat-panel pseudo-keyboard; the 800, its more expensive and bigger brother, had real keys) gave me opportunities to explore some new programming ideas. In other directions I tried to use it for a bit of writing and dial-up bulletin-board work, but nothing of significance came out of those activities. Like Commodore, Atari as a company never managed to figure out how to move its machine from hobbyist toy to mainstream system.
• ~1984 — Apple Macintosh "Classic" (et seq.). Here were some really useful personal computers. My family and I did writing, art, music, and animation on our Macs. I programmed them in FORTH and C, and created my free-text indexer/browser software on them. I learned Hypercard and the associated Hypertalk scripting language, which let me do rapid prototyping of nice graphical user interfaces. Apple was wise enough to encourage a broad spectrum of developers via its "Library of Tomorrow" initiative and other methods. Over the years I acquired a variety of Macintosh models, most noteworthy being upgrades of the Classic Mac toaster, luggable Mac Portables, and a UNIX-equipped Mac IIcx. Fine machines all, each in an idiosyncratic way. (see ThanksAlot (27 Nov 1999), WorldTexasHistory and NorthAmericanTexasHistory (15 May 2000), FreeTextFriends (25 Aug 2000), etc.)
• ~1987 — Toshiba T1000. This slow but lightweight and reliable laptop was, for over half a decade, my workhorse for computing on the road and for abortive forays into C coding in the IBM PC-compatible cosmos. The Toshiba served well in textual dial-up Internet and BBSing. But I was already spoiled by UNIX and Mac; the MS-DOS poor excuse for an interface made this and other conventional PCs feel like (insert your favorite metaphor of an unaesthetic activity here) for all but the most pedestrian applications. Nevertheless, I held my nose and used the T1000. Interoperability with a quasi-monopoly hath its benefits. In addition to the Toshiba, for the bulk of the 1990s I worked with an uninteresting and quite forgettable list of other DOS- and Windows-based systems.
• ~1998 — Apple Newton eMate. This ultraportable combination of touch-screen PDA and keyboard in a curvy turtle-shell lived in an alternative dimension of its own, beyond human reality. Although it let me compose ^zhurnal items, read email messages, and produce Web pages, the eMate was limited by its architecture to small projects (<20kB or so in practical terms). Worse, Apple never managed to sell a critical mass of Newton systems, and so development of software was limited to labors of love by genius aficionados such as Steve Weyer (the Newt's Cape browser), David Benn (the Little LISP language), and Simon Bell (the Simple Mail email client). I abase myself before these and other underappreciated Newton programmers; our world is unworthy of your labors. But when my eMate died in 2001 (mostly, his son RadRob still plays Daleks and Yahtzee on it /* or did until the power cord was lost; the battery doesn't hold a charge - RadRob */) , I had to move back to this mundane plane of existence.

My favorite Iron today? I'm almost afraid to reveal the answer, since so far things are going well ... maybe too well, as the horror movie cliché goes. For most of the past year I've been using an Apple iBook. It's a constant delight — the perfect combination of affordability, portability, reliability, processing power, and buck-naked beauty. The UNIX under the hood of Mac OS-X is a real operating system; no more need be said.

TopicPersonalHistory - TopicProgramming - 2002-02-25

(correlates: KillTheProject, FireFighting, DalaiLamaBirthdayGift, ...)