You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There exists a footgun in using raw RNG output when in reality you need a number between x and y with evenly distributed probability. This is called "modulo bias", and a good explanation can be found e.g. over at Kudelski.
Should the PSA Crypto API, which already provides a function to get whole bytes with uniform distribution, be extended with an API that gives an unbiased uniformly-distributed number between x and y?
The text was updated successfully, but these errors were encountered:
There exists a footgun in using raw RNG output when in reality you need a number between x and y with evenly distributed probability. This is called "modulo bias", and a good explanation can be found e.g. over at Kudelski.
Should the PSA Crypto API, which already provides a function to get whole bytes with uniform distribution, be extended with an API that gives an unbiased uniformly-distributed number between x and y?
The text was updated successfully, but these errors were encountered: