On Mon, Aug 4, 2008 at 8:14 AM, Stéfan van der Walt <[EMAIL PROTECTED]> wrote:
> > I agree -- seeing the NumPy-related build warnings would be very useful. I did not anticipate people would disagree on the idea of removing warnings, so I guess thare are two questions: - why / why not removing bogus warnings - if removing is considered OK, what's the best method: with the compiler-specific method, it could be done automatically and safely with some regex. > Do you know how to address these? > > numpy/core/src/arrayobject.c:3092: warning: format '%d' expects type > 'int', but argument 4 has type 'long int' Yes, at least on some platforms: this is addressed by using the inttypes.h header, proof of concept here: http://projects.scipy.org/scipy/numpy/ticket/847 But it would need a configuration stage to test we have the required formats in inttypes (the patch relies on how configuration is done in the python header, which I don't like and is not really robust). cheers, David _______________________________________________ Numpy-discussion mailing list [email protected] http://projects.scipy.org/mailman/listinfo/numpy-discussion
