Why CMatrix4<T> uses vector3df

[scippie]

I am using doubles in my internal calculations to afterwards convert them from absolute doubles to relative 3D floats for the Irrlicht Engine.

I use the Irrlicht classes for these calculations and atm. I use the CMatrix4<double> template class.

Now I need to transform a vector3d<double> with that matrix and although CMatrix<T> has a transformVect function, I can't use it because it uses fixed vector3df (which is vector3d<float>) as the io-parameter where I believe it should be better if it were a vector3d<T>.

Is there a reason for this or is this just a simple (understandable) forgetfulness?

Should I change it in my version of the engine or should I keep my hands of and use a custom made function for this?

[hybrid]

Well, I started to change the matrix to a template only some time ago. It gets some more template features on request, so this will probably go into the class some time soon. But until then you should be fine changing your local copy.