TR1 Math Special Functions

[Paolo Carlini]

Benjamin Kosnik wrote:

>> Then, the next issue is that the various check_nan.cc tests must be 
>> adapted to not rely on isnan being always available in std::tr1.
>
> How do you want to proceed on this stuff?

Nothing special. I pointed out already some fixes which are badly needed 
otherwise we have massive failures / bootstrap failures on rather common 
platforms and all non-C99 platforms.

More generally, very frankly speaking,  given the uncertain status of 
these features wrt C++0x, I'd like to make sure the starting point is 
pretty good and the maintainers don't have to suffer excessive pain over 
the next months at maintaining it. In my opinion, there is not much 
point in having an half-coked version of these features.

Paolo.