Designed to emulate the SNOCOM computer built in 1960 by David Wong from Sydney University for the Snowy Hydro.
8KiB (2048 32-bit words)
Plus accumulator and instruction pointer.
- Write a fully functional SNOCOM emulator.
- Write an assembler for SNOCOM.
- Get remaining programs from the NMA into ROM and tape files.
- Build a physical SNOCOM.
Runs a hello world program from rom.
Input instruction: need a way to differentiate between nibble reads and word reads. Perhaps add an operand: if 0 then read 1 byte, discard top nibble and store bottom nibble if 1 then read whole word. Use z-mode flag to determine how much to write.
Output = writes bytes. Bottom 8 of address are what to output. Uses Z-mode flag to determine which output device to use (Z set = typewriter, stdout)
These two changes mean that programs restored from punch tape will need some changes in order to work... but they needed changes to work anyway.
Also has the ability to load the machine's memory with a rom file, to avoid having to load via 'paper tape'.
The conditional transfer doesn't use the Z physical switch (because there isn't one), it asks via stdin what to do.
snocom [-r biosfile.rom] -i intape.txt -o outtype.txt -d
-r overwrites the tape autoloader with a custom memory image. Without a rom image specified, it will use Wong's tape autoloader as specified in his thesis.
-i is the input tape file (required)
-o is the output tape file (required)
-d specifies debugging mode, where a STOP instruction will enter the debugger rather than quitting.
-v shows version info
-h shows help info
typewriter goes to stdout
helloworld.bin prints "Hello world" to output tape.
helloworld_typewriter.bin does the same thing to the stdout typewriter.
Use https://github.com/hlorenzi/customasm
#ruledef
{
bring {mem: u11} => 0b00000000 @ 0b0001 @ 0b00000 @ mem @ 0b0000
add {mem: u11} => 0b00000000 @ 0b1110 @ 0b00000 @ mem @ 0b0000
sub {mem: u11} => 0b00000000 @ 0b1111 @ 0b00000 @ mem @ 0b0000
mulf {mem: u11} => 0b00000000 @ 0b0111 @ 0b00000 @ mem @ 0b0000
muli {mem: u11} => 0b00000000 @ 0b0110 @ 0b00000 @ mem @ 0b0000
div {mem: u11} => 0b00000000 @ 0b0101 @ 0b00000 @ mem @ 0b0000
extract {mem: u11} => 0b00000000 @ 0b1001 @ 0b00000 @ mem @ 0b0000
jump {mem: u11} => 0b00000000 @ 0b1010 @ 0b00000 @ mem @ 0b0000
jneg {mem: u11} => 0b00000000 @ 0b1011 @ 0b00000 @ mem @ 0b0000
jnegz {mem: u11} => 0b10000000 @ 0b1011 @ 0b00000 @ mem @ 0b0000
store {mem: u11} => 0b00000000 @ 0b1100 @ 0b00000 @ mem @ 0b0000
storec {mem: u11} => 0b00000000 @ 0b1101 @ 0b00000 @ mem @ 0b0000
storea {mem: u11} => 0b00000000 @ 0b0010 @ 0b00000 @ mem @ 0b0000
redadd {mem: u11} => 0b00000000 @ 0b0011 @ 0b00000 @ mem @ 0b0000
input nibble {mem: u11} => 0b00000000 @ 0b1110 @ 0b00000 @ mem @ 0b0000
input word {mem: u11} => 0b10000000 @ 0b1110 @ 0b00000 @ mem @ 0b0000
output tape {char: u8} => 0b00000000 @ 0b1000 @ 0b00000000 @ char @ 0b0000
output type {char: u8} => 0b10000000 @ 0b1000 @ 0b00000000 @ char @ 0b0000
stop => 0x00000000
}
You'll need a little endian computer, make and gcc.
make
ISC License
Pull requests welcome.
https://www.pearcey.org.au/blog/2020/snocom-and-cirrus/
https://ses.library.usyd.edu.au/handle/2123/16005
Not as fun as it sounds.
0 to 9 as they are
| Binary | Normal Hexadecimal | Wong's Sexadecimal |
|---|---|---|
| 1010 | A | + |
| 1011 | B | - |
| 1100 | C | N |
| 1101 | D | J |
| 1110 | E | F |
| 1111 | F | L |