Random.. not actually random

[Xav]

How about the current time, but just the millisecond part of it? I usually use DateTime.Now.Millisecond in .NET, which is the number of milliseconds in the current time.

[TcM]

Would it be possible to take more than one seed, combine them and make a more 'unique' and 'random' output?

[MeTh0Dz|Reb0rn]

Yes of course it is possible. Just write your own algorithm that can take multiple seeds run them through an algorithm then produce a seed to actually be use.

About polymorphism...

A polymorphic code is basically a code that can change itself or mutate itself but still accomplish the same thing. Basically it can patch itself, move parts of its code around, do things in different orders, rename things, etc.

A random number is necessary to decide in which way the algorithm should change itself. So a polymorphic code will use different types of seeds to help ensure its randomness.

Here are some links on polymorphism...

Polymorphic code - Wikipedia, the free encyclopedia

Polymorphism

Check those out they'll explain it better...

[marwex89]

And one could always use a random number as a seed too

[MeTh0Dz|Reb0rn]

How do you get that random number to begin with unless you have a seed?

[marwex89]

What kind of question is that? You use another random number of course

[MeTh0Dz|Reb0rn]

LOL, yeah you can't really say that you will use a random number as a seed. Because you have to have a seed to get that random number, and if you use a random number for that seed you'll need another for that seed, etc.

Your best bet is just finding something that varies quite often either between systems, or is just changing constantly. This will of course depend on the type of program that it is.

[marwex89]

But it might work if you use different random generators, or would it?

[MeTh0Dz|Reb0rn]

What do you mean by different random generators? You mean algorithms? And even so you are going to need different seeds otherwise your algorithms will end up always producing the same thing.

For example, let's say that you were using time as your seed and your first algorithm will start one second past midnight.

Algo 1) Takes the Seed of 1 and produces 5 (Multiplies by 5)
Algo 2) Takes the return of Algo 1 as a seed, 5, and produces 25 (Squares the Value)
Algo 3) Takes the return of Algo 2 as a seed, 25, and produces 18 (Subtracts 7)
Algo 4) Takes the return of Algo 3, as a seed, 18, and produces 9 (Divides by 2)

Okay do you finally end up with 2.

Now you might be saying okay that is more random because it went through more algorithms, but it's not because each algorithm depended on the first seed which is always going to be dependent on time. So basically everytime that it is one second past midnight you are going to get 2, just as if you had used the regular random function the value will always be the same.

That being said, you are best off finding a good seed. And if that is not good enough writing your own algorithm that can produce psuedo-random integers. But using multiple algorithms is pointless because they depend on eachother. Which will all lead back to the original seed.

[marwex89]

Yeah, you're right when one's thinking about it.. Man, it's late (yawn)