naming a company

[Xander]

You only need to grow these arrays to a few items each (4 * 4 will do) to never have a naming clash again

yack yack yack academic theory yack

Run the code. Academically I can see your point, but in reality this is not even remotely an issue. JAMI doesn't have a problem and uses an almost identical algorithm (she uses a 3*3*3)

[Dustin]

You only need to grow these arrays to a few items each (4 * 4 will do) to never have a naming clash again

yack yack yack academic theory yack

Run the code. Academically I can see your point, but in reality this is not even remotely an issue. JAMI doesn't have a problem and uses an almost identical algorithm (she uses a 3*3*3)

LOL.

It's not as Academic as you might think. My undergrad work in in psychology and communication, so I don't know most of the academic stuff other programmers do. What I do know is that code exactly like you describe can become a horrible performance bottleneck in the right circumstance and it's very anti-intuitive that this is so.

[Zuu]

With 3*3*3 you have 3^3 = 27 possibilities. And with maximum 15 companies, that is not much of a problem. As even when you pick the 15:th name you have a possibility of (27-14)/27 = 0.48 of picking a valid name. This is a very high possibility considering the computer power we have available.

But if you only have 16 possibilities the probability when picking the name for the 15th company becomes (16-14)/16 = 0.125, which is still high enough that it will not be a practical problem in our case.

But there are cases when the probability of getting it becomes so low that given the time constraints you have it becomes a problem to choose by random until you find something that works. If you want to dig into this you can read on the RSA-algorithm.