The explanation why the macro is necessary seems tautological - certain syntax won't work without the macro or macro-equivalent sugar, because it requires the macro. But why!?
I presume the macro has been invented as a hack to prototype the feature, and now the implementation is stuck with it, but what's the theoretical type system problem that prevents a clean hack-less implementation?
And why isn't existing {} or const {} sufficient? I thought this syntax has been reserved for const-evaluated values in types.
My interpretation is that invoking the GCA machinery when it's not necessary will cause regressions at this stage, but mechanically detecting whether it's necessary isn't easy.
I think it's not "…because it requires the macro" but "…because it requires all the backend code the macro invokes, but it's nontrivial to decide whether to invoke it automatically or not"
this is a bit convoluted, but I think that gca_const_items is the only feature that would allow actually doing any calculation dependent on generic params? e.g. [S; T::N + U::N] mentioned in the sibling comment would require that, because otherwise you can’t do any calculation in gca!().
I’m a bit unsure about usecase for gca_min_const_items, it would be nice to have a more practical example.
kornel | 20 days ago
The explanation why the macro is necessary seems tautological - certain syntax won't work without the macro or macro-equivalent sugar, because it requires the macro. But why!?
I presume the macro has been invented as a hack to prototype the feature, and now the implementation is stuck with it, but what's the theoretical type system problem that prevents a clean hack-less implementation?
And why isn't existing
{}orconst {}sufficient? I thought this syntax has been reserved for const-evaluated values in types.ssokolow | 19 days ago
My interpretation is that invoking the GCA machinery when it's not necessary will cause regressions at this stage, but mechanically detecting whether it's necessary isn't easy.
nytpu | 19 days ago
I think it's not "…because it requires the macro" but "…because it requires all the backend code the macro invokes, but it's nontrivial to decide whether to invoke it automatically or not"
gspr | 20 days ago
I can't wait to write stuff like in the examples, especially things like
[S; T::N + U::N]. Been wanting this since the second const generics hit.goldstein | 20 days ago
this is a bit convoluted, but I think that
gca_const_itemsis the only feature that would allow actually doing any calculation dependent on generic params? e.g.[S; T::N + U::N]mentioned in the sibling comment would require that, because otherwise you can’t do any calculation ingca!().I’m a bit unsure about usecase for
gca_min_const_items, it would be nice to have a more practical example.