Very strange!!!
I did some more tests on this problem.
First I note that the solution of Ross doesn't work.
And now the surprising thing:
- if I parent the ball to the player with NO previous rotation of the ball --> the ball is not scaled, everything is OK
- if I do a rotation first on the ball, and if after I parent the ball to the player --> the ball is scaled, and there is no simple way to anti-scale it (as proposed by Ross)
So, in fact, the child is scaled as soon as its axis are not any more aligned with the one of the player (y axis not any more parallel). That sounds very strange! I don't see any way to unscale the ball easily: I should unscale it after a rotation in its space, but there that's very difficult, and I am not sure it's possible within blitz...
I did some more tests on this problem.
First I note that the solution of Ross doesn't work.
And now the surprising thing:
- if I parent the ball to the player with NO previous rotation of the ball --> the ball is not scaled, everything is OK
- if I do a rotation first on the ball, and if after I parent the ball to the player --> the ball is scaled, and there is no simple way to anti-scale it (as proposed by Ross)
So, in fact, the child is scaled as soon as its axis are not any more aligned with the one of the player (y axis not any more parallel). That sounds very strange! I don't see any way to unscale the ball easily: I should unscale it after a rotation in its space, but there that's very difficult, and I am not sure it's possible within blitz...