At 04:36 PM 4/2/2010, Horace Heffner wrote:
The original Mits Altair 8800 we built had only 256 bytes (yes bytes
not k) of RAM, and we later upgraded it to 1k.
Same as me.
We programmed it in
machine language by toggling in the instructions using a bay of
switches on the front control panel.
Yeah. Flipped a lot of switches.
We added a "tape" card, which
wrote and read programs from an ordinary Radio Shack tape recorder.
I wrote an article for Byte magazine, they published it and paid me,
called the "Impossible Dream Cassette Interface." It was simply a
short program for writing data from memory to an ordinary cassette
recorder, through a direct connection with a data line, using pulse
width modulation (both transitions were used, which made the data
rate variable with the data, but it was terminally simple to
implement). The data was the read by the cassette recorder output,
clipped to +5 and ground, being fed into the interrupt line, counted
to discriminate between the "0" width and the "1 width"; again the
program was very short, which was necessary: it had to be entered
with the switches.
Materials needed: wire, capacitor, two diodes, a cassette jack, and
an ordinary cassette recorder.
It worked. Faster data rate than the interfaces you could buy for a
whole lot more.
Results were displayed in lights in binary.
I later added a video card, and expanded the memory by a collossal 8
kilobytes. And added a parallel I/O card and an EEPROM board,
programmed with the cassette interface programs as well as other
stuff, like, I think, BASIC. And that's as far as it went. I
eventually sold the Altair for $1 to a good home, sometime around
1992. The ribbon cables to the I/O card were beginning to tear.
This was the beginning
of my wife's career in computing. Prior she was somewhat anti-
technology anti-computer. Later Basic was available, but we
eventually decided to buy an Ohio Scientific Challenger computer
instead of upgrading. Both companies were soon defunct. We moved on
to Apple.
I went through a Vic-20 and a Commodore 128, before eventually
breaking down and buying a 286 PC.
That was an exciting time in computing, just as this is an exciting
time in energy transformation.
Yes. However, I'd become really good at working efficiently with
penciled printed circuit design, which was then taped on mylar. I
held out until the printed circuit design programs became more
efficient and cost-effective; but before that I was using the
Commodores to do wire-list analysis for checking of complex
multilayer designs against schematics. It was quickly obsolete, for
me, as soon as I had the 286 and the software, sometime around 1988
or so. Sorry, Chartpak and Bishop Graphics.
