Neal Becker wrote: > I'd just like to point out that the problem I ran into was not at all an > unusual use case. I was trying out cython to interface to some hypothetical > C code. The first test was C code with this interface: > > void F (double* in, double* out) > > where 'out' was a numpy array allocated by cython. No problem with this use > case. > > The second test was > > double* F (double* in) > > where the output was allocated by the C code and needed to be adopted into a > numpy array. This is where trouble began. I would think this use case > would also be quite common. > Yes, I hope this is easier to deal with in a year. It *can* be made quite user-friendly just by creating some utility functions to deal with it though.
Dag Sverre _______________________________________________ Cython-dev mailing list [email protected] http://codespeak.net/mailman/listinfo/cython-dev
