Re: jacko

On Feb 27, 12:27 pm, Willem <wil...@xxxxxxxx> wrote:
industrial_...@xxxxxxxxxxx wrote:

) On Feb 27, 10:32 am, Willem <wil...@xxxxxxxx> wrote:
)> Nope. The question is, and always has been, what is the chance of the
)> output of a codec that compresses normal data, to be RAD.
)
) I just answered that: the chance is 100%
)
)> And here's another question: Suppose I have a true random generator (for
)> example, one that measures decaying atoms or something), and I generate a
)> file from that, what is the chance of *that* file being RAD ?
)
) About 99.99% i'd say. If it's a really good random number generator
) then you will almost always end up with a file of true randomness
) (0.900+) I've generated random files a thousand times and only ONCE
) did I end up with one that could be compressed by at least 1 bit with
) WinZIP (0.875)

A while ago, you claimed that less than half of all possible files are RAD.
Do you still think that ?

I think there are more non-RAD files than RAD but like i said: math
isn't my thing and I cannot give an accurate answer.

P.S. I said a *true* random generator. Not a really good one. A true
random generator would be the same as flipping a fair coin once for each
bit in the file. Your experience with *pseudo* random number generators is
quite useless here.

Well DUH. *True* RNG *IS* what I meant. My point was that *IF* it was
a good/true RNG then it will 99.99% of the time generate RAD. The
reason it's not 100% is because the data it generates is technically
random, so theoretically it IS also possible to end up with a file of
english text.

By the way, true/pseudo RNG makes no difference, they both generate
random-appearing data.

.