fixed_study

multiply! it produces the same values as C itself for all the multiplies of signed 16-bit values I throw at it, except that its representing each 16-bit value as two 8-bit values. hopefully this will extend reasonably to two 32-bit values, and then to four 32-bit values.

subtract

signed 16-bit ints based on 8-bit ints: conversion int16_t <--> 's2'. comparison. addition.

Initial commit