Can we stop writing all the extra 0x0's?

I might suggest having the parser notice when 3 aren't provided, and fill in 0x0 for all missing arguments.
(Or the parser could be smarter)

```
LOOP 0xF 0x0 0x0
SET8 0x1 #R1 0x1
GET8 0x2 0x2 0x0
IF 0x1 #R2 0xF
PASSFAIL 0x0 0x0 0x0
ENDTEST 0x0 0x0 0x0
ENDIF 0x0 0x0 0x0
DELMS 0x1 0x0 0x0
REGINC 0x1 0x0 0x0
ENDLOOP 0x0 0x0 0x0
```

Becomes
```
LOOP 0xF
SET8 0x1 #R1 0x1
GET8 0x2 0x2
IF 0x1 #R2 0xF
PASSFAIL 0x0
ENDTEST
ENDIF
DELMS 0x1
REGINC 0x1
ENDLOOP
```

Notes:
- I expect the tokenizer would fill these in, so no firmware updates are necessary.
- I don't like exceptions, so a rule like "any trailing arguments of `0x0` may be left out and are assumed to be `0x0`" should apply to all opcodes to prevent confusion.
- In cases where 0x0 is a real value, such as `REGLOD 0x1 0x0`, I personally don't think it's a problem to let people use `REGLOD 0x1` shorthand. In C and C#, we know that if you write `int a`, the variable defaults to 0.
- This might be a confusing one: `PASSFAIL`. That would set the test status to fail. The test would start with `PASSFAIL 0x1` and on a failed chip validation (sub test), we'd write `PASSFAIL; ENDTEST` or if we have "constant variables", define "PASS 0x1; FAIL 0x0" so we can write `PASSFAIL FAIL; ENDTEST`. Come to think of it, it would also be strange to see `REGLOD` which would load value 0x0 into register 0x0 (#R0).
- The other option around this is have your tokenizer know how many arguments each opcode needs, and just have the reserved/unused parameters be ones we can omit... and it fills in. So just having `PASSFAIL` or just `REGLOD` wouldn't "compile" and must include the required parameters.

I really don't want to type or see this anymore (Only 3 of those 18 args are real):
```
PASSFAIL 0x0 0x0 0x0
ENDTEST 0x0 0x0 0x0
ENDIF 0x0 0x0 0x0
DELMS 0x1 0x0 0x0
REGINC 0x1 0x0 0x0
ENDLOOP 0x0 0x0 0x0
```

Thoughts?