Subject: Re: [SG10] Updates to SD-6
From: Ed Smith-Rowland (3dw4rd_at_[hidden])
Date: 2015-01-06 11:29:26
On 01/05/2015 04:29 PM, Nelson, Clark wrote:
> Here's an updated document.
> Summarizing recent discussion:
> With respect to N4267 (u8 character literals), we have now had three macro
> name proposals:
> The latter two, of course, update the value of a macro from C++11. I'm
> inclined to guess that this is going to be considered a small enough tweak
> that introducing a new macro name (such as the first) would not be
> People have talked about how unfortunate it is that the C++11
> recommendations use two different macro names for such closely related
> features. So far, I have not interpreted this is an actual proposal to
> change SD-6 -- it seems like noodling up to this point. But if someone
> wants to produce a concrete proposal along these lines, we can certainly
> consider it.
> But I also want to point out that there are nine other C++17 papers for
> which no proposal has yet been made.
> For N4190 (removing old stuff) and N4258 (cleaning up noexcept), if no one
> from SG10 makes a proposal pretty soon, I'm going to contact the authors to
> see what kind of ideas they have.
> Clark Nelson Chair, PL22.16 (ANSI C++ standard committee)
> Intel Corporation Chair, SG10 (C++ SG for feature-testing)
> clark.nelson_at_[hidden] Chair, CPLEX (C SG for parallel language extensions)
> Features mailing list
Finally looking at
N4258: Cleaning-up noexcept in the Library
The only thing I can think of that would be helpful is
I think the actual application of noexcept to all the various containers
don't lend themselves to a macro in a way that would help migration. It
seems the customization of is_always_equal is used inside noexcept to
set the exception policy. If __cpp_lib_allocator_is_always_equal is
defined you know you have to provide and set non-default is_always_equal
for user-defined stateful allocators. Similarly, user-defined
containers could use the macro to migrate noexcept specification for
their container methods.
In my last email I was thinking out loud and maybe it looked chaotic to
me this morning.
I would like to summarize my favored approaches:
N4267 (u8 character literals)
N4190 (removing old stuff)
These would get date bumps if/when std::not_fun etc. or std::bind is
I am slightly annoyed that the fist two don't have _deprecated_ but they
look nice to me as they are.
N4279 - Improved insertion interface for unique-key maps
N4051Allow typename in a template template parameter
This may be too little to mess with.
N4268 - Allow constant evaluation for all non-type template arguments
N4230 - Nested namespace definition
I like this better than __cpp_nested_namespaces that I had before.
N3922 - New Rules for auto deduction from braced-init-list
__cpp_auto_deduction 200604 for C++11
__cpp_auto_deduction 201411 for C++17
This could be an alternate spelling of|__cpp_return_type_deduction|:
__cpp_auto_deduction 201304 for C++14
SG10 list run by herb.sutter at gmail.com