Try A Little Randomness
summary
Explains what counts as real randomness versus deterministic pseudorandomness and shows concrete sources and failures: random.org uses atmospheric static and hotbits uses a cesium 137 Geiger counter, CPUs expose RD-RAND, Python’s default is the Mersenne Twister with period 2^19937−1 and 624-word state, which is fast but reversible if 624 outputs are seen so it fails CSPRNG tests; old LCGs like RANDU and Visual Studio rand (multiplier 214013) produce visible patterns and 15 hyperplanes or vertical streaks in visual tests.
AI-written summary · about
chapters (17)
- 0:00Start
- 0:42what are random numbers ?
- 1:12what use are they?
- 2:20how do i get one?
- 2:45example sources of noise
- 3:46entropy pool
- 5:22what are pseudorandom numbers ?
- 5:41why use them?
- 7:43prng concepts
- 8:43generating a prng
- 9:28good prng
- 9:55merseinne twister
- 11:58cryptographically secure prng
- 14:14bad prng
- 14:54linear congruent generator
- 19:34randu in 3D
- 20:00summary
related talks
28mRandomness in Python: Controlled Chaos in an Ordered Machine
12mConstrainedrandom - a Python package that does what it says on the tin
30mAll in the timing: How side channel attacks work
01h 30mAn Introduction to Probability and Statistics
29mI wish I learnt that earlier!
01h 11mMay 18, 5pm
59mTesting in Python
01h 48mSimple Games for Python Beginners