About distances

[pierremgi 4737]

Hi all,

Any suggestion about all commands referring to distance (between 2 objects)?

 

About distance calculation, I found few difference between the commands:

 

Example: player at [4142.04,4569.92,0] , object: o1 at [4165.75,4751.43,0]

 

player distance o1                 // 188.577   algorithm applied to center?

player distanceSqr o1           // 35561.1  (sqrt 188.577)

 

getPosASL player vectorDistance getPosASL o1   // 188.906 

same as:   getPosATL player Distance getposATL o1

same as: sqrt((getPosASL o1 #0 - getposASL player #0)^2 + (getPosASL o1 #1 - getposASL player #1)^2 + (getPosASL o1 #2 - getposASL player #2)^2)

 

As reminder: player distance2D o1       // 183.064  sqrt((getPosASL o1 #0 - getposASL player #0)^2 + (getPosASL o1 #1 - getposASL player #1)^2)

probably used in radius of trigger, distance 2D for nearestObjects...

 

About performance, if you don't need precision, just using a distance filter for objects in vicinity of player, not so much difference between:

player distance o1 < 200 (sphere)

player distanceSqr o1 < 40000 (sphere)  // perhaps a little better in performance

player distance2D o1 < 200 (cylinder)    // perhaps a little better in performance.

 

And you, how do you check distances?

[sarogahtyp 1104]

mostly I use distanceSqr. I thought bout if distance2D is faster but never tested it.

With nearestObjects I use 2d mode as well.

[panther42 40]

@pierremgi I usually just use distance2D for my calculations between flying helicopter and specific point I'm using on the map.

The link gives a good picture, which I'm sure you have seen. 

Difference?  Probably only minimal.

[sarogahtyp 1104]

I tried to measure the difference between distance, distanceSqr and distance2D in the debug console but without success. I could not see any difference in performance between these 3 commands.

[pierremgi 4737]

distance2D skips the elevation (as shown above). That's faster than vectorDistance.

The distanceSqr avoids the sqrt calculation, probably non significative for our CPUs.

Distance seems to have its own algorithm, i didn't find but i didn't test comparison with vector between centers of models.