|Posted: 2011-December-27 at 12:02am | IP Logged
OK, I'd read years ago that the Analytical Engine itself could be programmed by setting levers, similar to the way the first electronic computers were programmed with switches. Seems that he never got that far. I had read that the first programmer was Ms. Lovelace, and that she actually wrote programs, but that they were never used.
Incidentally, the first computer I worked on, in 1978, was an ancient Burroughs B55, from around 1960, and it was programmed by pushing the neon lights on the front panel. The preferred method was to read programs from cards, but it could be programmed, one bit at a time, by finger poking. It was rumored to be the first computer series, (B200, 300, 500), Borroughs built that didn't have any vacuum tubes. Thrid National Bank of Ashland bought it new in the early 60s, and processed work for several other local banks. It was slow, ran one program at a time, and had only about 20K of magnetic core memory, but was almost indestructible. Even power failures didn't bother it. The program would die, but the last data was still in memory, and could be recovered, when it came back on.
We had learned about "sense switches" in college, which i had just finished, but the computer there was newer, and the "switches" were read from punch cards or entered by console typewriter. The B500 had 6 pushbutton switches, numbered 1 through 6, and you physically turned them on and off.
That was the mainfrome running a multibank processing center. Talk about starting at the bottom! Never mind Cobol or Fortran, this thing was programmed in machine language... The operator output was also in machine language. If it halted, you read the address and halt number, in binary, from those same neon lights. We finally got a new computer, a modern (1980) mainframe, a few years later, but all of us operators certainly had a good understanding of what REALLY went on in a computer!
The original Space Shuttle computers used core memory, although a little newer version. After Challenger's computer was retrieved from the Atlantic much of its memory was still readable.
Given the radiation problems in space, your mechanical computer might have advantages. With the current work on nanoscale mechanical systems I wonder if Babbage's design could be miniaturized?