"John Maddock" <[EMAIL PROTECTED]> wrote in message 010201c2a5c9$c1feb560$43e7193e@1016031671">news:010201c2a5c9$c1feb560$43e7193e@1016031671... > given the Standard's wording in 3.9/1: > > "... There are two kinds of types: fundamental types and compound types. > ..." > > isn't the current implementation of is_compound overly complex? Can't we > just change it to basically match: > > is_compound = !is_fundamental > > given a working version of is_fundamental of course, but this seems > easier than providing an always safe and portable is_class or other > stuff required for the current version. > > Regards, Daniel > > ~~~~~~~~~~~~~~~~~~~ > > I think you may be on to something, I'll try it out when I have a chance, > thanks! > In the same fashion - I think that if type is not cv void, not reference and not convertible to bool (via standard conversion) then it's class type. It might be useful implementation for compliers on which the SFIAE technique to detect class type doesn't work.
Rani _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost