From: G DOT DegliEsposti AT ads DOT it To: "Salvador Eduardo Tropea (SET)" cc: djgpp AT delorie DOT com Message-ID: Date: Mon, 23 Mar 1998 17:45:13 +0100 Subject: Re: Re: real random numbers Mime-Version: 1.0 Content-type: text/plain; charset=us-ascii Precedence: bulk >Hmmm... That isn't true, you can easilly create a random number generator using >a diode and an A/D, or just sampling the enviroment sound ;-). There are a lot You are right! I would like to see a copy of libc with not only rand() and random(), but also ADrandom() !!! :-) >of cases like that, not all involves an A/D converter, some signals (not >properlly grounded, etc) have random patterns. You can do it, but you need some >hard. Of course. But I meant software solutions alone. I am not sure on this, but you don't need complicated hw: the system clock is often a good rng. >> This makes the functions look like if they generate random numbers, but >> they actually don't! > >And that's a feature because in this way you can test programs with random >values more than ones and with the same values, or use an encripting methode >using "random" numbers. Sure, and also a little mistery for newbies: "Why do I get always the same sequence of numbers every time? Aren't they supposed to be random?" ;-) ciao Giacomo