Now that I can disassemble it, I've been digging into it. It is a BIG program, uses all of bank 0. First, the instructions in kahla_lab.txt are totally wrong. Using its settings will immediately crash.
I can get it to work a bit, but it's really not functional, still looking into it.
You can see it play itself, though.
The real (so far) instructions:
TW switches 0-5 specify the number of pieces in each pit to start with. Don't try using more than a small number, too many will crash the program and corrupt memory.
TW switches 15-17 determine the lookahead depth the program uses. Bigger numbers will take MUCH longer per move as will more pieces. MUCH longer.
A good demo setting is:
TW 003001
The sense switches are heavily used. They are only scanned when the program starts or is restarted at location 02000.
SS1 - just halts
SS2 - same as SS3
SS3 - show a detailed numeric representation of the moves being played
SS4 -
no display, but show game decision making as it plays itself
SS5 - show a numeric representation of the moves being made
SS6 - show a typed board updating as the moves are being made
SS2,3,5,6 are exclusive. If more than one is set, the lowered number one takes precedence
So, what's wrong?
It NEVER asks for input in any mode.
If SS4 is off, it just displays the initial board setup and does nothing else.
If SS4 is on, it plays itself.
Clearly something is broken, I'll poke around some more until I get bored.
Bill