Date: Tue, 19 Aug 2014 16:56:24 +0000
The meeting was attended by:
Clark Nelson (chair)
John Spicer
Aaron Ballman
Walter Brown
The draft of SD-6 that was posted on 08-15 was reviewed. All of the name
proposals were approved, including the new proposal of
__cpp_lib_shared_timed_mutex. The consensus was to remove the
originally-proposed __cpp_lib_shared_mutex.
Aaron volunteered to provide introductory and rationale text for
__has_cpp_attribute. (Thank you, Aaron.)
It was proposed and agreed that the table of features removed from C++14
should be presented separately from the actual features of C++14.
Aaron's proposal to change the result of __has_cpp_attribute from a pure
Boolean to include version/date information was discussed and accepted. This
functionality is considered especially important for vendor-specific
attributes. For the sake of consistency, it was decided that the original
standard attributes should have year+month values specified in the C++11
feature table.
We discussed the approach used by the library groups to prevent interface
lock-in when a TS is eventually incorporated into the standard. There was no
SG10 consensus to make waves.
Pending action items:
Clark to check with Richard about rationale wording for __cpp_lib_is_final.
Clark to prepare and post a new draft of SD-6.
Clark Nelson (chair)
John Spicer
Aaron Ballman
Walter Brown
The draft of SD-6 that was posted on 08-15 was reviewed. All of the name
proposals were approved, including the new proposal of
__cpp_lib_shared_timed_mutex. The consensus was to remove the
originally-proposed __cpp_lib_shared_mutex.
Aaron volunteered to provide introductory and rationale text for
__has_cpp_attribute. (Thank you, Aaron.)
It was proposed and agreed that the table of features removed from C++14
should be presented separately from the actual features of C++14.
Aaron's proposal to change the result of __has_cpp_attribute from a pure
Boolean to include version/date information was discussed and accepted. This
functionality is considered especially important for vendor-specific
attributes. For the sake of consistency, it was decided that the original
standard attributes should have year+month values specified in the C++11
feature table.
We discussed the approach used by the library groups to prevent interface
lock-in when a TS is eventually incorporated into the standard. There was no
SG10 consensus to make waves.
Pending action items:
Clark to check with Richard about rationale wording for __cpp_lib_is_final.
Clark to prepare and post a new draft of SD-6.
-- Clark Nelson Vice 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)
Received on 2014-08-19 19:06:27