FWIW, here's my experience with various C++11 compilers: - GCC >= 4.8 is C++11 feature complete (core language, not necessarily library-wise) and quite usable (this includes MinGW); - clang >= 3.3 is C++11 feature complete and in general slightly better for standard-conformance and stability than GCC; - Intel's compiler is reasonably feature-complete as of version 15 (at least on Unix-like archs), but it's an ICE trainwreck when one tries to use even slightly complex metaprogramming. I would expect it to choke on many modern C++ libraries; - MSVC is still lagging a lot behind.
My suggestion would be to require GCC 4.8 or greater and do not bother with anything earlier. Features will be missing and the likelihood of bugs in implemented features is much higher before 4.8. Another random thing to keep in mind is that OSX clang disables some C++11 features because "Apple". E.g., C++11 thread-local storage is still disabled pending some ABI changes in Apple's ecosystem: http://stackoverflow.com/questions/23791060/c-thread-local-storage-clang-503-0-40-mac-osx http://stackoverflow.com/questions/28094794/why-does-apple-clang-disallow-c11-thread-local-when-official-clang-supports (this is a real drag for parallel programming) Cheers, Francesco. On 23 September 2015 at 22:09, Clement Pernet <clement.per...@gmail.com> wrote: > The Givaro-LinBox-fflas-ffpack ecosystem now requires C++11 support. > I'm hitting this problem, while working on #17635, upgrading the spkgs of > these 3 libraries. > > Therefore: > > > 1) Do we want to mandate c++11 support > Y[X] > N[ ] > > > > 2) if yes what c++11 features do we want? > Feature complete [X] > > as many features are used. > > > List of features [ ] (list needed features) > Tentative list of what's currently being used (I may have missed some) > * using > * enable_if > * auto > * fixed width types and related constant suffixes (ex uint64_t x=2_ui64;) > * ... > > Clément > > * override > > > > François > > > >> On 22/09/2015, at 15:27, Ralf Stephan <gtrw...@gmail.com> wrote: > >> > >> On Monday, September 21, 2015 at 6:38:50 PM UTC+2, Volker Braun wrote: > >>> Afaik we already require C++11 support to compile Pynac > >> > >> Yes, Pynac git master requires it but we're still installing backported > versions (0.3.9.x vs 0.4.x). > >> > >> The ticket that never got finished was > >> http://trac.sagemath.org/ticket/18323 > >> > >> Regards, > >> > >> -- > >> You received this message because you are subscribed to the Google > Groups "sage-devel" group. > >> To unsubscribe from this group and stop receiving emails from it, send > an email to sage-devel+unsubscr...@googlegroups.com. > >> To post to this group, send email to sage-devel@googlegroups.com. > >> Visit this group at http://groups.google.com/group/sage-devel. > >> For more options, visit https://groups.google.com/d/optout. > > > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel. > For more options, visit https://groups.google.com/d/optout. > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
