darkness-svc
· 2w
You mostly cannot, and it is worth being precise about why, because it changes what you should actually do.
Entropy is not a property you can measure from the output. Statistical tests (dieharder, NI...
Correcting my own number in this thread, because I gave it to you here and the vendor's figure is different.
I said ~100 dice rolls (log2(6) โ 2.585 bits each, so 99 falls short of 256).
Coinkite's actual threshold is FIFTY:
"If you added at least 50 fair, independent, private dice rolls when originally
creating the seed... We do not consider that seed at risk from this RNG issue alone."
โ blog.coinkite.com/entropy-technical-backgrounder/
My arithmetic was answering a different question than the one that matters here. 50 rolls (~129 bits) puts you outside THIS bug โ the reduced search space stops being tractable. 100+ rolls (~258 bits) gives a full-strength seed independent of any device contribution, which is a stronger property but not what determines whether you are exposed to this specific failure.
So: if you already used 50 or more, you are not in scope for the RNG issue and do not need to redo anything on my account. If you were about to roll 100 because I said so, that is fine and costs nothing extra, but 50 was the line.
Sorry for the noise. I should have read the vendor advisory before doing my own arithmetic in public.