Re:Password Strength
by karmakaze on Sunday January 16, 2022 @11:14AM
(#29957175)
There are two points to add: (1) If you use a randomly generated password of upper/lower/punctuation that's 63^length; (2) Four words is vocabulary^4.
One breakeven point is 63^10 (9.8e17) ~ 20000^4 (1.6e17).
This is being quite generous to the four words as the likelyhood of word selection from the vocabulary would be far from uniform. Also offset by the fact that shorter length passwords are brute-forced. If four-words became the dominant meta then we could instead have brute force applied to the word dictionary forms.
Re:Password Strength
by ncmncm on Sunday January 16, 2022 @01:07PM
(#29958299)
Use shuf(1) to pick words reliably randomly from /use/share/dict/ dictionaries. It is not a mistake to misspell the words proposed, but not necessary. Preserving capitalization is not necessary.
One breakeven point is 63^10 (9.8e17) ~ 20000^4 (1.6e17).
This is being quite generous to the four words as the likelyhood of word selection from the vocabulary would be far from uniform. Also offset by the fact that shorter length passwords are brute-forced. If four-words became the dominant meta then we could instead have brute force applied to the word dictionary forms.