Date: Fri, 21 Mar 2025 11:47:22 +0100
It is the link time choices I have mind, possibly requiring the linker to be smarter.
> On 21 Mar 2025, at 11:37, Robin Savonen Söderholm <robinsavonensoderholm_at_[hidden]> wrote:
>
> You have several depending on context. You have templates with constraints and overload resolution that you can use at compile time to choose and customise what to call. And you have "weak linked functions" that you can override at link-time (like the terminate handler in GCC).
>
> On Fri, Mar 21, 2025, 11:10 Hans Åberg via Std-Proposals <std-proposals_at_[hidden]> wrote:
>
> > On 21 Mar 2025, at 10:56, Ville Voutilainen <ville.voutilainen_at_[hidden]> wrote:
> >
> > On Fri, 21 Mar 2025 at 11:48, Hans Åberg <haberg_1_at_[hidden]> wrote:
> >>>> No, it is for compiler writers to mark what is not implemented, instead of doing nothing or using “delete”. See the example I gave upthread.
> >>>
> >>> If a compiler writer wanted a facility like this, they would already
> >>> be using it.
> >>
> >> Indeed, we haven't heard anything from any actual C++ compiler writers in this thread I would think.
> >
> > You just did hear from one.
>
> Great. There are some others I noticed didn't appear in this thread.
>
> I think though interesting, at least from the theoretical point of view, having a feature that the one definition rule means that one can by some policy choose which one to use, a user extension to what is already supplied for global operator new, and “inline”.
>
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
> On 21 Mar 2025, at 11:37, Robin Savonen Söderholm <robinsavonensoderholm_at_[hidden]> wrote:
>
> You have several depending on context. You have templates with constraints and overload resolution that you can use at compile time to choose and customise what to call. And you have "weak linked functions" that you can override at link-time (like the terminate handler in GCC).
>
> On Fri, Mar 21, 2025, 11:10 Hans Åberg via Std-Proposals <std-proposals_at_[hidden]> wrote:
>
> > On 21 Mar 2025, at 10:56, Ville Voutilainen <ville.voutilainen_at_[hidden]> wrote:
> >
> > On Fri, 21 Mar 2025 at 11:48, Hans Åberg <haberg_1_at_[hidden]> wrote:
> >>>> No, it is for compiler writers to mark what is not implemented, instead of doing nothing or using “delete”. See the example I gave upthread.
> >>>
> >>> If a compiler writer wanted a facility like this, they would already
> >>> be using it.
> >>
> >> Indeed, we haven't heard anything from any actual C++ compiler writers in this thread I would think.
> >
> > You just did hear from one.
>
> Great. There are some others I noticed didn't appear in this thread.
>
> I think though interesting, at least from the theoretical point of view, having a feature that the one definition rule means that one can by some policy choose which one to use, a user extension to what is already supplied for global operator new, and “inline”.
>
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
Received on 2025-03-21 10:47:53