On Mon, Apr 30, 2012 at 8:56 AM, Aaron Ballman <[email protected]> wrote: > On Mon, Apr 30, 2012 at 10:51 AM, Douglas Gregor <[email protected]> wrote: >> >> On Apr 28, 2012, at 2:04 PM, Francois Pichet wrote: >> >>> On Sat, Apr 28, 2012 at 4:09 PM, Ahmed Charles <[email protected]> >>> wrote: >>>> qualified name) >>>> MIME-Version: 1.0 >>>> Content-Type: text/plain; charset="utf-8" >>>> Content-Transfer-Encoding: 7bit >>>> >>>> Has there been any consideration around having MS imply C++11? Given >>>> that VC doesn't support modes for standard versions, it wouldn't >>>> breaking users expectations as much as requiring that their code is >>>> compatible with the most recent version of VC. >>> >>> >>> Yes that would make sense if we want to emulate the latest MSVC. >>> (clang can't the MSVC 2010 standard header files without -std=c++11) >>> >>> I haven't pushed that because I always use -cc1 and explicitly pass >>> the options I want. >> >> >> I'd be fine with ms-compatibility mode enabling C++11. There's no point in >> trying to support older MSVC versions. > > I think it's reasonable as well.
Note that the msvc c++ library headers expect stddef.h to define std::nullptr_t when building in c++11 mode. Without this, they won't parse. (That's not an argument against the proposed change, doing it will just require other changes too.) Nico _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
