hi 3DGS.
i don`t know if it is a bug or some thing...
when i want to get hight of a triangle,i use the method in the manual:
VECTOR pointA;
VECTOR pointB;
VECTOR pointC;
//these 3 points define a triangle(of course their Z is all the same)
VECTOR vec_A_to_B;
VECTOR vec_A_to_C;
VECTOR vec_B_to_C;
//these 3 vectors means 3 lines of the triangle
var angle_A = acos(vec_dot(vec_A_to_B,vec_A_to_C)/(vec_length(vec_A_to_B)*vec_length(vec_A_to_C)));
var hight_B = vec_length(vec_A_to_B) * sinv(angle_A);
and i found that the value of hight_B is bigger than the right value
then i use Cosine Theorem:
var cos_A=(vec_A_to_B^2 + vec_A_to_C^2 - vec_B_to_C^2)/(2 * vec_A_to_B * vec_A_to_C);
angle_A = acos(A);
hight_B = vec_length(vec_A_to_B) * sinv(angle_A);
i found that the value of hight_B is smaller than the right value
only when i don`t use acos() function:
hight_B = sqrt(vec_A_to_B^2 - (vec_A_to_B * cosA)^2);
this value of hight_B is the right value...and i really don`t know why..
thanks for read this,hope to get your answers!

As far as I see, your formulae are correct. But for large or acute triangle you're possibly outside the precision range of var variables. Better use float for trigonometry.

If a formula does not give the desired result, split it in sub-steps and check in which of the steps the result becomes wrong. You must do this yourself - find the reason and fix it. This does not belong to the bug forum. Blaming the programming language for wrong results of your script won't help.

ok,thank you!
indeed,my triangle is relly big,around 1000 per line...maybe it`s the reason.
after cos() and sinv(),the hight`s value goes wrong i think