Date: Sat, 5 Dec 2020 09:51:56 -0500
On Sat, Dec 5, 2020 at 9:05 AM darkdragon via Std-Proposals <
std-proposals_at_[hidden]> wrote:
> It would be awesome if a constexpr version of hash_code() and
> preferably name() would be included in the standard. Is there anything
> more constexpr than type_info?
>
Sadly, yes; `type_info` is intimately tied up with RTTI, and many
implementations implement type_info::operator== as a string-comparison of
the names of the types, so that it'll keep working even in the presence of
dynamically linked libraries (DLLs).
So your goal might not be literally *impossible*, but it won't be trivial,
either.
–Arthur
std-proposals_at_[hidden]> wrote:
> It would be awesome if a constexpr version of hash_code() and
> preferably name() would be included in the standard. Is there anything
> more constexpr than type_info?
>
Sadly, yes; `type_info` is intimately tied up with RTTI, and many
implementations implement type_info::operator== as a string-comparison of
the names of the types, so that it'll keep working even in the presence of
dynamically linked libraries (DLLs).
So your goal might not be literally *impossible*, but it won't be trivial,
either.
–Arthur
Received on 2020-12-05 08:52:09