C++ Logo

sg10

Advanced search

Re: [SG10] First draft of SD-6 with changes from Oulu

From: Jonathan Wakely <cxx_at_[hidden]>
Date: Wed, 20 Jul 2016 20:19:34 +0100
On 20 July 2016 at 19:22, Richard Smith <richard_at_[hidden]> wrote:

> On Tue, Jul 19, 2016 at 11:12 AM, Nelson, Clark <clark.nelson_at_[hidden]>
> wrote:
>
>> I have incorporated all of the changes approved at the last meeting into
>> the
>> table for C++17. The draft can be found at:
>>
>> http://wiki.edg.com/pub/Wg21oulu/SG10/sd-6.html
>>
>> Very few of those proposals had their own macro recommendations, but I
>> have
>> taken them into account. And as yet I haven't done anything about filling
>> in
>> the rationale, even for the cases for which I made my own recommendation.
>> So
>> we definitely have work to do.
>>
>
> Some suggestions:
>
> Forward progress guarantees: no macro (these papers really just add
> definitions)
> Inline variables: __cpp_inline_variables
> Guaranteed copy elision: no macro (portable code should avoid or cope with
> copies)
> Expression evaluation order: no macro (portable code should not rely on
> the order)
> Constexpr if: either __cpp_constexpr_if (matching the paper name) or
> __cpp_if_constexpr (matching the syntax)
> Selection statement with init: no macro (portable code can perform a
> simple rewrite to avoid the feature)
>

All agreed.


> Structured bindings: __cpp_structured_bindings
>
>
I know that's the paper name, but it's not a term in the standard. Would
__cpp_decomp_decl or something based on "decomposition declaration" make
more sense?



> All the variant changes from Oulu should be covered by
> __has_include(<variant>); I don't think we have a need to track them
> separately unless someone chooses to produce a <variant> header that
> doesn't match the contents of any working draft.
>
>

Agreed.




> Up until now I have been updating SD-6 on isocpp.org basically around the
>> holidays. But we might want to try to publish an update before the
>> Issaquah
>> meeting, to cover the C++17 CD.
>>
>> I'd like to schedule a telecon to make some progress on this. August 1 and
>> August 15 look like plausible candidates. If anyone has any definite
>> preference for one over the other, please let me know.
>>
>> --
>> 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
>> Features_at_[hidden]
>> http://www.open-std.org/mailman/listinfo/features
>>
>
>
> _______________________________________________
> Features mailing list
> Features_at_[hidden]
> http://www.open-std.org/mailman/listinfo/features
>
>

Received on 2016-07-20 21:19:55