Re: Constructed Computer Architectures (Concomps?)

On Mon, 9 Feb 2009 18:07:31 -0500, Michael Potter <mhpotter@...>
wrote:

>A few months ago, I started designing a computer that used balanced
>ternary logic (which uses -1 as the third logical value, instead of 2).
>I did this as a thought experiment, but I have actually written a
>simulator that (mostly) works. I haven't tried making a language for it,
>though, and there's a lot that would still need to be done to make it
>look "real".
IA while back I played around with ideas for a balanced ternary computer.
Purely thought experiment, though; I have no chip design experience. It
mainly consisted of playing with a character encoding designed for 9-trit
trytes, which could be subdivided into 3-trit "tribbles". The first tribble
carried character behavior metadata (such as line break, word break,
hyphenation, and collation), while the remaining two were the character code
proper. Diacritics and ligatures used a stack, where the "combining"
meta-code meant that the character was to be pushed onto the stack; the last
character in the complex would not be marked as combining. Since the final
character could be marked with any non-combining metacode, this allowed for
proper hyphenation of the eszett and diacritics that did or did not affect
collation.

I also had an idea for a digitally switched analog computer. Opcodes took
the form of a pair of addresses: the first was an emitter (an oscillator,
steady voltage source, system input such as joystick axis, sample-and-hold,
or operator output) and the second a sink (operator input, sample-and-hold,
oscillator control, or system output). Address 0 in both cases pointed to
the ground. It assumed a vector monitor like an oscilloscope, and direct
audio output (since it was basically just an analog synth with addressing in
place of patch cords—which is what the idea grew out of in the first place).