On Tue, 19 Sept 2023 at 19:42, Arthur O'Dwyer <arthur.j.odwyer_at_[hidden]>
wrote:

> On Tue, Sep 19, 2023 at 9:35 AM Patrice Roy via SG14 <
> sg14_at_[hidden]> wrote:
>
>> Thanks. The actual path to take will depend on the actual paper to come.
>> Right now, what we have is a request for a «feature» in the list compiled
>> by P2966, and until there is an actual paper written on that one
>>
>
> Patrice: Part of what Jonathan is saying is that the right approach to
> this specific entry is *not* a WG21 paper; it's a vendor bug report. But
> only if your vendor actually has that bug, and it sounds like nobody does.
> So there's Nothing To Do on this entry.
>
> But! PMR's wording *used to* contain dynamic_cast, until we removed it
> via LWG3000 in March 2018. The P2966 entry was probably brought up sometime
> before 2018, and should have been removed from P2966 as a result of
> LWG3000's resolution.
> https://cplusplus.github.io/LWG/issue3000
>

Ah yes, well spotted. Did anybody ever implement it like that though? When
did MS STL first implement those resources? The initial commit to github in
2019 already didn't have the dynamic_casts, but maybe they were present
before it was open sourced. The initial implementation of them in GCC
(2018) didn't have the casts. And the initial implementation in libc++
(2022) didn't either.

Anyway, it's only one item in the paper, which has plenty of other topics.
But I did see others that also seemed like QoI, this one just stood out.