Re:/dev/random – a new approach
by tomsmeding on Monday April 25, 2016 @02:43AM
(#11562800)
That's one of the things that's mentioned in the paper: don't seed one bit at the time. Moral of this discussion (not of the paper) is that you should regularly reseed, but only when you can completely replace the seed with entropy, and don't worry about making users wait for randomness because it makes you look interesting.
I knew someone was going to say this. In fact, the last sentence of my comment was added in an edit after this rebuttal occurred to me. Yeah. Dunno. Definitely don't slowly permute your CSPRNG state one bit at a time.
Or is there something I'm not seeing about how reseeding works?