Question may be off topic, but I was just wondering if this wouldn't be a nice and simple random number function. It's supposed to give a random Number from 0 to $max:
function randomNumber($max){
((int)(1000000000000 / microtime())) % ($max+1);
}
If I'm not missing something pretty obvious here this should work, shouldn't it? But, the next question would be: Are the numbers created here really random? When I simulate 1'000'000 dice rolls, I get something like the following:
Array
(
[1] => 166520
[2] => 166619
[3] => 166522
[4] => 167001
[5] => 166512
[6] => 166826
)
Why I'm asking this question is: I've read that it's not possible to create a truly random number, so I started thinking and this function seems to give quite random results, without being complicated at all. So am I missing something here?
That's not pseudo-random enough to be considered acceptably random.
Please keep in mind that even numbers generated with a random number generator aren't truly random. They're just pseudo-random enough that they can be used when a random number is required. They still follow a pattern, albeit a really complicated one.
While it may be a simple solution, and it may appear to give you random numbers, you might find oddities in aggregated data generated specific to your application. For example, if you were generating rolls for a game like Monopoly where you generate two dice rolls at the same time, you might find a higher incidence rate for rolling doubles than is typical. I can't say that your specific algorithm would cause this specific problem. This is just an example of the type of anomalies you could encounter using your own random number generation routine.
One thing I can say specifically is that your generated values are based on the current number of microseconds at generation time. If you generate two values within a couple of milliseconds, the results will be nearby. A typical random number generator seeds the generator with a value (usually from the current time), but from that point on each call to generate another number simply provides the next value in the random sequence started by the seed. The number you get is not dependent on the time unless you reseed the generator with each call. This element is present in your routine, though, and may also be the source of unwanted side-effects.