On a lark, I've worked up a hashing algorithm designed for my retro
computing platform of choice, the Apple II. Using emulation and real
hardware, I've tested to the best of my ability for randomness of
output and "avalanche" bits.
As such, I've made some improvements to version 1.0. (source published
here: https://github.com/option8/sg32 with a more in-depth
presentation here: https://youtu.be/HA0IC7Sh5aQ). I'd like to properly
evaluate the 1.1 for collisions and other weaknesses but I lack the
necessary knowledge. Combine that with the algorithm being native to a
40 year old hardware platform, and even automated tools are lacking.
Where would you start?