[JASS] Does GetRandomInt accept zero?

[Bribe]

I've stuck with GetRandomInt between 1 and X for some time now. If I plug in "between 0 and X" does it count 0 as a random integer? This is helpful for setting arrays...

 

[Dynasti]

Yes.


Why shouldn't it?

 

[Bribe]

I have no idea how randomization works with computers... the last question I had was why getting a random 'real' number didn't cause the game to crash... logically, I pictured real numbers to have infinite decimal places...

All of the randomization triggers in blizzard.j start with 1 and not 0, so I thought they had a reason for that

 

[Adiktuz]

I have no idea how randomization works with computers... the last question I had was why getting a random 'real' number didn't cause the game to crash... logically, I pictured real numbers to have infinite decimal places...

All of the randomization triggers in blizzard.j start with 1 and not 0, so I thought they had a reason for that

you're getting a real number while using GetRandomInt??

if not,
reals can have infinite decimal places but why would they cause a crash, and if they would then I think the makers of warcraft won't even put the action GetRandomReal()...

 

[JeffreyQ]

Yes it does accept 0.
Also, there is no such thing as infinite decimals in computer. Even a super computer has it limits.

 

[Teelo]

All of the randomization triggers in blizzard.j start with 1 and not 0, so I thought they had a reason for that

Probably because most functions in blizzard.j, including those that use arrays, are 1-based.

 

[Dr Super Good]

Reals are 32 bit floating points I believe. Thus they only have a finite accuracy and so can not crash. Basically a floating point is an approximation of a number, meaning it actually is not exact.

0.1 for example could become 0.100025 so it is truncated for accuracy.

I seriously doubt WC3 uses IEEE complient floating points as speed is more important than precision thus accuracy probably is only a 32 bit float so 4294967296 different numbers capable of being represented. I could be wrong.

Larger numbers lose accuracy in the decimal places, thus how the spread is provided.