On Tuesday 15 May 2012, Darin Adler wrote: > On May 15, 2012, at 10:04 AM, Allan Sandfeld Jensen wrote: > > The conflict is between isinf, isnan and std::isinf and std::isnan, but > > the conflict only exists in C++11 when constexpr versions are > > introduced. > > We should try harder to come up with a solution in MathExtras.h. Even one > that uses macros. > > If I understand correctly, the conflict here is not between some WebKit > namespace and the std namespace, it’s in the C++ library itself between > the global namespace and the std namespace. I think that’s a bug in the > C++ library, and MathExtras.h is the perfect place for a workaround for > that bug. > True. If we can find a method that would work, that would be the optimal solution, but it is a difficult case to work around. It is not that I haven't tried, but modifying those ~10 files was just a lot easier in the end.
> > Since the problem exists in both GCC 4.6 and 4.7, I see no alternative > > than ensuring any files that uses isinf or isnan does not open the std > > namespace. This is only around 10 files in total as you can see in my > > patch. > > Sure, but that assumes we won’t use more things from the std namespace in > the future, and I am not sure that’s a good assumption. > You don't have to open the namespace to use the methods within. I prefer to use the 'std::' prefix. I only kept them in this dcase to keep the changes to a minimum. Best regards `Allan _______________________________________________ webkit-dev mailing list [email protected] http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
