Useless multiplication?

[ghley]

I might be wrong here as I haven't worked through all the code so far and I'm assuming inverseMass is always 1/m but isn't in DynamicsSystem.cpp line 152

mRigidBodyComponents.mConstrainedLinearVelocities[i] = mRigidBodyComponents.mConstrainedLinearVelocities[i] + timeStep * mRigidBodyComponents.mInverseMasses[i] * mRigidBodyComponents.mLinearLockAxisFactors[i] * mRigidBodyComponents.mMasses[i] * mGravity;

the inversMass * Mass always 1 and thus irrelevant?

[DanielChappuis]

Hello,

Yes, it seems to be useless but the only thing is that the 'inverseMass' of a body can also be zero that is useful to simulate a static body that cannot move. Therefore, if we remove the inverseMass * mass multiplication we cannot simulate a static body (that has zero inverse mass).

[ghley]

That makes sense, so instead of checking for the body to be static it's all just integrated in the calculation.

Thanks!

[DanielChappuis]

Yes exactly. Note that it is also possible for a user to set the inverse mass of a body to be zero even for a body that is not static.