
Is there any chance that the virtual function boost::detail::is_polymorphic_imp1<T>::d2::foo defined in boost/type_traits/is_polymorphic.hpp could be renamed to have a name less likely to cause conflicts? I appreciate that the unique struct prevents this function from overriding any virtual functions in the base class, but it still has the potential to hide base class functions, which, at suitable warning levels in many compilers, causes a lengthy warning to be emitted. Although foo is unlikely to be a common function name in production code (well, good production code, anyway), it is the sort of name that is frequently found within unit tests and so forth. (For example, the name ``foo'' is used 2657 times in the gcc testsuite.) Cheers, Richard Smith

"Richard Smith" <richard@ex-parrot.com> wrote in message news:Pine.LNX.4.61.0410141627050.16642@sphinx.mythic-beasts.com...
Is there any chance that the virtual function boost::detail::is_polymorphic_imp1<T>::d2::foo defined in boost/type_traits/is_polymorphic.hpp could be renamed to have a name less likely to cause conflicts?
Okay. I suggest 'bar'. ;-) Or, borrowing from libs/thread/src/once.hpp, foo2AC1A572DB6944B0A65C38C4140AF2F4 Jonathan

Although foo is unlikely to be a common function name in production code (well, good production code, anyway), it is the sort of name that is frequently found within unit tests and so forth. (For example, the name ``foo'' is used 2657 times in the gcc testsuite.)
Sigh, yes you're correct, I'll change it. John.
participants (3)
-
John Maddock
-
Jonathan Turkanis
-
Richard Smith