On January 22, 2015 3:56:12 AM EST, John Maddock <boost.regex@virgin.net> wrote:

...

It's useful to have

#ifndef __has_feature #define __has_feature(x) 0 #endif

#ifndef __has_extension #define __has_extension __has_feature #endif

in suffix.hpp so that one can then use __has_extension(foo) everywhere without fear. Otherwise, even

#if defined(__clang__) && defined(__has_extension) && __has_extension(foo)

is a preprocessor error when __has_extension is not defined (which it isn't when not on Clang.)

Good point.

However, aren't those reserved symbols? If we're being pedantic....

I was going to point out the same. What about BOOST_HAS_FEATURE and BOOST_HAS_EXTENSION? ___ Rob (Sent from my portable computation engine)

On 22 Jan 2015 at 8:56, John Maddock wrote:

...

It's useful to have

#ifndef __has_feature #define __has_feature(x) 0 #endif

#ifndef __has_extension #define __has_extension __has_feature #endif

in suffix.hpp so that one can then use __has_extension(foo) everywhere without fear. Otherwise, even

#if defined(__clang__) && defined(__has_extension) && __has_extension(foo)

is a preprocessor error when __has_extension is not defined (which it isn't when not on Clang.)

I'd be opposed to that because the draft before SG10 allows patterns of using __has_extension which cannot be emulated reasonably in current compilers. Much safer is to use the #ifdef __cpp_rtti type method of detecting features.

Good point.

However, aren't those reserved symbols? If we're being pedantic.... John.

https://github.com/ned14/Boost.BindLib/blob/master/include/cpp_feature .h provides a header file which makes the SG10 feature macros standard across all the major compilers, including fixing up any compiler version specific abberations. I would view this emulation of SG10 on older compilers as safe and reasonable. We emulate new compiler stuff on older compilers all the time. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/