Well, it looks like testing for the signature (using your code or presumably something from the upcoming for review function type library) will not preclude testing for actual accuracy... Though one can wonder at this stage if we are still dealing with QOI or a broken library.
Looks like it might be a broken implementation: looking at the output from config_info at http://www.meta-comm.com/engineering/boost-regression/cvs-head/developer/out... it seems that _RWSTD_NO_OVERLOAD_C_POW is defined, and the Rogue Waves headers say: /* If your 'C' library does not provide overloads for the pow function * (i.e. pow(float,float), and pow(long double, long double) if appropriate), * then uncommment the following. */ /* #define _RWSTD_NO_OVERLOAD_C_POW 1 */ So I guess at least one API is broken. John.