Mistakes in 6.0 SDK?


    Oct 19 2011 | 8:41 am
    In the updated "simplemsp~" example, the "float" message method declaration accepts a double, not a float as one might expect. Is this a typo, or does Max 6 now secretly pass around doubles, with unsupporting objects just taking the MSB?
    void simplemsp_float(t_simplemsp *x, double f);
    Also, take a look at this (32-bit perform method):
    t_simplemsp *x = (t_simplemsp *)(w[1]);
    t_float *inL = (t_float *)(w[2]);
    t_float *outL = (t_float *)(w[3]);
    int n = (int)w[4];
    Compared to this (64-bit perform method):
    t_double *inL = ins[0];
    t_double *inR = ins[1];
    t_double *outL = outs[0];
    The 32 bit method has only 1 input and 1 output, whereas the 64-bit method has 2 inputs and 1 output. Am I missing something, or is this just a mistake?

    • Oct 19 2011 | 11:35 am
      For as long as I can remember it has been necessary to declare typed float arguments as doubles for windows compatibility at least. The A_FLOAT type is still an actual float though....
      A.
    • Oct 19 2011 | 11:46 am
      Ah, thanks. C'est un peu bizarre, non?