I noticed Andrew Kuchling's METH_VARARGS submission: Use METH_VARARGS instead of numeric constant 1 in method def. tables While METH_VARARGS is obviously a lot better than a hardcoded 1, shouldn't METH_VARARGS be something like Py_METH_VARARGS or PY_METH_VARARGS to avoid potential conflicts with other packages? Skip