pt., 25 lis 2022 o 11:30 Frederick Virchanza Gotham via Std-Proposals
<std-proposals_at_[hidden]> napisał(a):
>
> Guys, seriously, can you give me an actual compileable code snippet
> instead of an overly-simplistic English sentence? I have a great
> imagination, I'm actually writing a stage play at the moment, but I
> don't want to try imagine what you might mean in your posts here.
> Please be verbose and pedantic, please give me a code snippet that
> compiles so I know what you're talking about.
>
> My imagination right now is telling me that you might mean having two
> separate functions as follows:
>
> extern void Red( wxTextCtrl * );
> extern void Red( wxComboBox * );
>
> The problem with having these two overloads is that it's not generic
> enough. I want my function to work with *any* class that inherits from
> both wxControl and wxTextEntry. Of course I could write a template
> function but I want to have one simple concrete function.
>

Probaby Proxy is the correct solution for this problem?
https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2022/p0957r5.pdf
And even more it would allow unrelated types to pass to this function too.
As long as they can be mapped to the required interface.

> Gergely I saw your other post in which you use "if constexpr (requires
> { p.what(); }) p.what();". I found that post helpful and I'm currently
> testing code and composing my reply.
>
>
>
> On Fri, Nov 25, 2022 at 9:53 AM Gergely Nagy <gergely.nagy.alt_at_[hidden]> wrote:
> >
> > However I do not understand why simple overload is not good enough.
> >
> > On Fri, Nov 25, 2022, 10:52 Frederick Virchanza Gotham via Std-Proposals <std-proposals_at_[hidden]> wrote:
> >>
> >> On Fri, Nov 25, 2022 at 6:15 AM Sebastian Wittmeier
> >> <wittmeier_at_[hidden]> wrote:
> >> >
> >> > Hi Frederick,
> >> >
> >> > as I understand your paper a Chimeric Pointer is a pointer to a class inheriting from all
> >> > the listed base classes and supporting the -> operator for members of those base
> >> > classes without additional cast
> >>
> >>
> >> Yeah that's right. The chimeric pointer may be bigger than
> >> "sizeof(void*)" in cases of virtual inheritance.
> >>
> >>
> >> > What would be the advantages and disadvantages of a concept with those properties?
> >> > Is the main advantage that the Chimeric Pointer erases the actual type and the function
> >> > with the pointer parameter does not have to be a template?
> >>
> >>
> >> Please write the concept you have in mind, and I'll compare it to a
> >> chimeric pointer.
> >>
> >> The idea behind a chimeric pointer such as
> >> "chimeric_pointer<wxControl,wxTextEntry>" is that it can hold the
> >> address(es) of any object whose pointer can convert implicitly to a
> >> wxControl* and also wxTextEntry*.
> >>
> >> Template functions are an absolutely fantastic addition to the C++
> >> language, they have brilliant uses in particular in implementing
> >> containers, but sometimes I just want _one_ function. If I have three
> >> instantiations of a template function, then that's three unique
> >> functions with three unique addresses in memory, and also that's three
> >> different copies of any static objects defined inside the function.
> >> Sometimes I just want one simple function, in particular if I'm
> >> exporting a function from a dynamic shared library.
> >> --
> >> Std-Proposals mailing list
> >> Std-Proposals_at_[hidden]
> >> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals
> --
> Std-Proposals mailing list
> Std-Proposals_at_[hidden]
> https://lists.isocpp.org/mailman/listinfo.cgi/std-proposals