Date: Tue, 10 Jun 2025 03:32:48 -0700
Your email does not do anything that has been *repeatedly* been asked of you.
No one cares about an implementation - the people who develop C++ compilers and the supporting runtimes can do the the work to support any feature that the standard requires.
When you are suggesting a change or addition to the standard you need to provide a justification for that change. If you can provide such a justification, a proof of concept implementation is a *nice to have*, but that is a minor part of the proposal, and is irrelevant if you have not done any of the other work in for a proposal.
In my last email I said what I believe you need to do:
> stop sending emails, and instead write an actual proposal. You can see many examples of proposals
> that you can use to model you proposal on at https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2025
> (and prior years/meetings as well)
Your refusal to do this means you do not have any interest in making a good faith proposal, and your constant refusal to do any of the work to actually justify your desired feature means that I don’t think it is fair for you to expect us to continue to give you the benefit of the doubt. As long as you continue to refuse to write an actual proposal that justifies your proposed feature, and why the existing options are insufficient (after all your first actual example use case is already handled by std::variant).
Until you do that work all you are doing is wasting the time of everyone else on this list. Your behavior implies that you believe you have a right to arbitrary amounts of time from other people on this list, and that you don’t consider our time to have any value.
I am serious.
Write a serious paper.
Provide concrete use cases.
Explain why existing features are insufficient.
Explain how your proposal is consistent with the rest of the language definition. Not “I can write code to do this”. Explain how it is consistent with the rest of the standard, and how it is consistent with the type system.
Explain how it can work when given a pointer to a subobject. Explain how it can work when provided a pointer to non-primary base classes.
Explain how it works given arbitrary implementations of polymorphism that don’t match your one preferred implementation.
If you are not willing to do this, please stop wasting everyone's time.
—Oliver
No one cares about an implementation - the people who develop C++ compilers and the supporting runtimes can do the the work to support any feature that the standard requires.
When you are suggesting a change or addition to the standard you need to provide a justification for that change. If you can provide such a justification, a proof of concept implementation is a *nice to have*, but that is a minor part of the proposal, and is irrelevant if you have not done any of the other work in for a proposal.
In my last email I said what I believe you need to do:
> stop sending emails, and instead write an actual proposal. You can see many examples of proposals
> that you can use to model you proposal on at https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2025
> (and prior years/meetings as well)
Your refusal to do this means you do not have any interest in making a good faith proposal, and your constant refusal to do any of the work to actually justify your desired feature means that I don’t think it is fair for you to expect us to continue to give you the benefit of the doubt. As long as you continue to refuse to write an actual proposal that justifies your proposed feature, and why the existing options are insufficient (after all your first actual example use case is already handled by std::variant).
Until you do that work all you are doing is wasting the time of everyone else on this list. Your behavior implies that you believe you have a right to arbitrary amounts of time from other people on this list, and that you don’t consider our time to have any value.
I am serious.
Write a serious paper.
Provide concrete use cases.
Explain why existing features are insufficient.
Explain how your proposal is consistent with the rest of the language definition. Not “I can write code to do this”. Explain how it is consistent with the rest of the standard, and how it is consistent with the type system.
Explain how it can work when given a pointer to a subobject. Explain how it can work when provided a pointer to non-primary base classes.
Explain how it works given arbitrary implementations of polymorphism that don’t match your one preferred implementation.
If you are not willing to do this, please stop wasting everyone's time.
—Oliver
Received on 2025-06-10 10:33:00