On Wed, 12 Jun 2019 at 21:50, John Spicer <jhs_at_[hidden]> wrote:

> The process for fixing things like this is not well defined right now.
>
> In general, the feature test stuff is supposed to be part of the CWG/LWG
> process.
>
> For fixing things that were missed earlier, I did a paper for a few of
> those recently.
>
> How would folks like this handled? It could be an LWG issue, or I can
> write a short paper with this proposed resolution.
>
>
My preference is an LWG issue. It can probably get approved by email, and
be on the "Ready" list going into Cologne. That should take up less LWG
time during the meeting, and less of your time :-)





> John.
>
>
> > On Jun 12, 2019, at 3:58 PM, Jonathan Wakely <cxx_at_[hidden]> wrote:
> >
> > There are parallel overloads of algorithms in <memory>, so it should
> define the macro. Also, <execution> defines the exec policies for use with
> the algos, so that should define the macro too.
> >
> > Proposed resolution:
> >
> > In [support.limits.general] add <memory> and <execution> to the table
> row for __cpp_lib_parallel_algorithm.
> >
> >
> > | __cpp_lib_parallel_algorithm |
> > | 201603L |
> > | <algorithm> <INS><execution> <memory</INS> <numeric> |
> >
> > _______________________________________________
> > Features mailing list
> > Features_at_[hidden]
> > http://www.open-std.org/mailman/listinfo/features
>
>