Date: Tue, 10 Jun 2025 12:59:18 +0100
Things aren't added to the standard because they can be done. If it's
possible to do, just ship a library on GitHub. No proposal needed.
Things are added to the standard because they are impossible to do, or
because they are vocabulary types for interop between libraries. There are
also some things that are there for convenience, but there hasn't been an
addition like that in a while that I can recall.
This proposal seems to fit none of those niches. Ship a library, get users,
if the community starts raising a stink about this needing to be in the
standard then you will have a glowing "feature gap analysis" section in
your paper. Until then, I'm afraid it's very unlikely to get in.
On Tue, Jun 10, 2025, 11:49 Frederick Virchanza Gotham via Std-Proposals <
std-proposals_at_[hidden]> wrote:
> On Tue, Jun 10, 2025 at 11:40 AM Ville Voutilainen wrote:
> >
> > Well, that's not *entirely* accurate, there are people in WG21 who
> > will tell a proposal author "go home, and come back when you have
> > prototyped it" for unimplemented proposals. :)
>
>
> The purpose of my implementations is to show that it can be achieved
> without an ABI break.
>
> 'std::polyhandle' would not get off the ground if it meant that
> compiler vendors had to change the size of their vtables (e.g. add
> more info to their vtable).
>
> My implementations prove that there's enough info already there in the
> the vtables, type_info's, RTTI, and so on. No alterations needed. No
> ABI break.
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>
possible to do, just ship a library on GitHub. No proposal needed.
Things are added to the standard because they are impossible to do, or
because they are vocabulary types for interop between libraries. There are
also some things that are there for convenience, but there hasn't been an
addition like that in a while that I can recall.
This proposal seems to fit none of those niches. Ship a library, get users,
if the community starts raising a stink about this needing to be in the
standard then you will have a glowing "feature gap analysis" section in
your paper. Until then, I'm afraid it's very unlikely to get in.
On Tue, Jun 10, 2025, 11:49 Frederick Virchanza Gotham via Std-Proposals <
std-proposals_at_[hidden]> wrote:
> On Tue, Jun 10, 2025 at 11:40 AM Ville Voutilainen wrote:
> >
> > Well, that's not *entirely* accurate, there are people in WG21 who
> > will tell a proposal author "go home, and come back when you have
> > prototyped it" for unimplemented proposals. :)
>
>
> The purpose of my implementations is to show that it can be achieved
> without an ABI break.
>
> 'std::polyhandle' would not get off the ground if it meant that
> compiler vendors had to change the size of their vtables (e.g. add
> more info to their vtable).
>
> My implementations prove that there's enough info already there in the
> the vtables, type_info's, RTTI, and so on. No alterations needed. No
> ABI break.
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
>
Received on 2025-06-10 11:59:33