Hacker News new | ask | show | jobs
by gbonc2 1473 days ago
Hi, I'm the author of this one ( I lost the other account password). I'm happy I found someone else doing this stuff to be able to talk about it. I discarded time related algorithms because I thought they could introduce sneaky bugs related to event dependencies. So, having had to work in the gamble industry, I opted for a simpler algorithm: roulette. I have a loop increasing a register from 0 to 15, when a particle is detected the value present in that registry is returned and it's used as an half of a 8 bits ( 1 byte ) random number and, after every extraction, the counter is restarted from zero. If anyone think the algorithm is unreliable, any suggestion to improve it's welcome.
1 comments

It's pretty similar to mine, you could improve yours by relying on an hardware timer/counter that can go faster and thus provide more entropy.

Also you should not reset the counter at every event, this would lead to a poisson distribution if event ever became of comparable time scale to the timer