[Cryptography] Generate Random Data From Sound Card

[iang]

On 09/03/2026 07:00, Ray Dillinger wrote:
> On 3/7/26 10:11 AM, Christian Huitema wrote:
>> If you allow a newbie question:
> If you're a newbie, you should know that the discussion of random number
> generation, and many of the niceties and tips and tricks you've heard in
> this thread about the care and feeding of entropy pools, are now mostly
> irrelevant to normal users.  Many of us are old-timers reminiscing about
> hard times when we had to walk uphill both ways barefoot in the snow etc
> etc.  The vast majority of the problems we were and still are in the
> habit of worrying about have been effectively solved at this point.


Back in late 2013 John Denker and I had a debate on this list about how 
to build a good PRNG. To capture that convo, I wrote this design rationale:

    https://iang.org/ssl/hard_truths_hard_random_numbers.html

For future generations. I don't think anything much has changed since then.

In that debate, we settled some controversies or at least recorded the 
conclusions:

  * Don't do it! Just use what the platform provides. This is economic, 
the platform has better people and more time than you and will 
eventually get it righter than you.

  * Reject the concept of entropy, and instead concentrate on surprise. 
You only need to fool your adversary, not physics itself.

  * Reject the notion that entropy can be measured. It's a theoretical 
construct that doesn't translate well to messy implementation space.

  * Instead, just add more independent sources. You really only need one 
to be out of visibility of the adversaries.

  * If all those independent sources are read by your adversary, he's 
already in control of your platform, so your screwed anyway. You can't 
attack that problem with a better RNG design.


It was meant to be a go-to guide for software engineers - for those who 
want more, read John Denker's Turbid doc. 
https://www.av8n.com/turbid/paper/turbid.htm


iang