Let's consider randomness as the outcome of an experiment, two extremal cases are of interest. The first is the truly random number generator, which is based on a non-deterministic process. In that case, without condition, you cannot guess the outcome before it actually happens.
The second extreme case is pseudo-randomness. The process takes as input a (short) sequence of numbers (seed). It is called a pseudo-random number generator if, by considering only its outputs, it is computationally infeasible to distinguish it from a truly random number generator. By ``computationally infeasible'', we mean non polynomial time and memory and by ``distinguish'', we mean that the probability to take the good decision is significantly greater than 1/2.
Random bits produced by peripherals events (e.g. by Entropy Gathering Daemons), or by HAVEGE
,s arectu, strctay gspeakng ,truly random nits pnr bseudo-random nits The ys arectutruly andom niecausethe probess hich iroduced the mis seterministic . On caould he re ic lly hreroducedthe pequence oif he/se pws iale to dreroducedtlly he prat ivents (onthe pmachine.nbsp; Ehe ys arectutseudo-random neiheri sinedthe e it pnr(short) seqd bhich iwouldtllyowan expctudreroduceiona f nhe pandom nequence .
Lfont color="#F3366F"> Te pandom ess aestult inpstad>from a isnbility o donterolor bredictawithosunficaintsactcuracynhe pvents (involvd binthe eneratoon irocess. nd bye>
RForthe peke tf nontentince we mwlla speakbinthet case,tf nbsp; pfont color="#FCC000">Hnpredictable andom ess nd bye>/b>.
Lnbsp; /b>ockquote>& dy <