Horst Schirra has pointed out to me the problem of name changes to the user functions and the potential for user confusion when changing versions. One example is the function 'diffusion_tensor.set()' which was renamed to 'diffusion_tensor.init()'. In the future, functions may be merged, shifted, deleted, or renamed. However I don't know the best way to document this. One solution could be to include dummy functions which throw an error like "RelaxError: Depreciated function, use 'diffusion_tensor.init()' instead". If anyone has any other ideas, these would be appreciated. Thanks, Edward