[Cryptography] quantum computers & crypto

[cherry]

On 10/31/21 8:15 AM, Ray Dillinger wrote:
> So where fixing those things has a significant cost, that cost is a sign
> that they needed fixed anyway.

The major cost of "fixing" things is that broken crypto tends to get 
even more broken.  Our institutional process does not appear to be 
capable of fixing things.

The state of our Wifi protocols is merely a particular example of 
general brokenness everywhere.

There are a vast multitude of problems and issues just as serious in 
widely used cryptography, as the defect in Wifi, which allows passive 
sniffers to perform an offline dictionary attack, which will break a 
majority of human memorable passwords.


The fix for the Wifi problem (PAKE and OPAKE) has been known for over 
two decades, but instead of it being fixed, recent Wifi protocol changes 
have made it a great deal worse, by broadcasting the hash all the time, 
instead of only during logins.