------- Comment #5 from pcarlini at suse dot de 2007-05-25 17:12 -------
(In reply to comment #2)
> I don't think support for C++0x precludes support for TR1. They coexist very
> well, especially because TR1 was designed to be compatible with C++0x. For
> example, C++0x-conforming implementations of the TR1 facilities also meet the
> requirements of TR1.
Hi Doug. Upon Benjamin' invitation, I'm working on this issue, with the initial
goal of making progress on the C++0x version of type_traits (for example, it
will exploit more front-end builtins).
Now, however, I do not agree completely with your statement above: certainly,
for example, a C++0x-conforming type_traits doesn't meet the requirements of
TR1. That, in turn, sheds doubts to the very point that TR1 facilities must be
available in C++0x mode: if the user does an using from namespace tr1 to
namespace std of a tr1 facility which finds a same-named, but incompatible
facility (i.e., implementing different semantics) in namespace std, which one
is supposed to "survive"?
--
pcarlini at suse dot de changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |pcarlini at suse dot de
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31426
