[Starbish]

i made 3 int variables just for replacing MAXPLAYERS+1 boolean, by using Bitwise operation.

but some guy told me that it's too slow to use BITWISE OPERATION in VM like javascript.

is it true that using such things make operation too slow?

sample code is in below this.

PHP Code:

#define MAX_PLAYER_ARRAY 3
#define MAX_AFFORDABLE_PER_VAR 31

new iVariable[MAX_PLAYER_ARRAY];

stock bool:IsPlayerSatisfied(client, iValue[MAX_PLAYER_ARRAY]){

    if(iValue[RoundToFloor( float(client) / float(MAX_AFFORDABLE_PER_VAR) )] & ( 1 << client-1 ))
        return true;

    return false;
}

stock SetPlayerCondition(client, iValue[MAX_PLAYER_ARRAY], bool:bValue = true)
    iValue[RoundToFloor( float(client) / float(MAX_AFFORDABLE_PER_VAR) )] |= ( 1 << client-1 );

stock ClearPlayerArray(client, iValue[MAX_PLAYER_ARRAY])
    for(new i; i < MAX_PLAYER_ARRAY; i++)
        iValue[i] = 0; 


[asherkin]

No.

[Starbish]

Originally Posted by asherkin No.

then, do you think is it worth to change this rather than using MAXPLAYERS+1 boolean vars?

and what do you think which is faster?

[friagram]

Dunno why it would be, bitwise is the lowest level of math, and should be extremely fast. Xor is used on a register against itself to zero it out faster than setting it to zero on x86 since it's a smaller instruction. There is even a special case/optimization for some cases like this.. Some people use it rather than other math operators in some situations because of its speed. Some examples http://lab.polygonal.de/?p=81

[asherkin]

Don't worry about micro-optimizations like this unless you've profiled it and found a bottleneck - you're wasting time you could spend actually building the meat of your plugin on premature optimization.