The full Monte

years ago 1). So, a word of self-introduction is in order. Back in the summer of 1960, while interning at Grumman Aircraft, I was assigned to learn the new language Fortran so that I could program the company's brand-new IBM 704. The rules of engagement were arcane—you wrote out the program on an official IBM programmer's pad, submitted it to the keypunch operators, and (if the Force were with you) got back a punched deck several days later. Then the fun began. Imagine debug-ging on a system with three-day job turnaround! Luckily, the next summer I got to program a Burroughs 220 (in machine language) and an IBM 7090 (in Fortran II) in a university setting. We were actually allowed to touch the key-punch machines ourselves (or in the Burroughs's case, the tape punch). And—oh, joy!—the turnaround time was " only " 24 hours! If you want some idea of what prehistoric computing was like, read Fred Hoyle's sci-fi novels, The Black Cloud (Lightyear, 1998) and Ossian's Ride (Harper, 1959)—both have chapters whose " action " takes place in computer centers. The personal computer revolution began in the late 1970s. I soon discovered that getting a program to run using interactive Basic on a desktop machine (such as the HP-1000 or the NorthStar) was faster than on the big CDC-6600 mainframe using Fortran. I even published a paper 2 whose computations were all done on a Sinclair ZX-81 computer that ran overnight. It had a 4-MHz, 8-bit Z80 chip, with an interpretive ROM Basic and 16 Kbytes of RAM. It displayed 16 lines of 64 characters on a TV set and used a tape cassette player for storage. The personal computer was lots slower than a mainframe, but the answers came out much faster—a paradox we all live with now. When I learned to program, computer science departments lay far in the future. For good or ill, I learned about programming through reading and practice. The upside is that I acquired no linguistic prejudices along the way and felt free to try anything. The downside is that I am still ignorant of things for which I did not have an immediate use. In 1985, mainly through force of circumstance, I began to use Forth for most programming, preferring it to the various dialects of Fortran and Basic that had served me until then. Although I still occasionally dabble in C …