Tristan Brindle

> Hi Tristan, > > Thanks for the feedback, I've moved the majority of the code shared between `cartesian_product`, `cartesian_product_with`, and the new `cartesian_power` (I feel it differentiates better with...

Something else that has just occurred to me: for the "power" versions, the cursor type could be `std::array` rather than `tuple`... I don't know if this would make things easier...

~Hi @isaacy2012, after saying I hadn't tried it I thought I'd better have a go! This is a very rough prototype of what I had in mind: https://flux.godbolt.org/z/54Wrs3Wha~ ~As you...

Hi again @isaacy2012, sorry for bombarding you with messages! I've been thinking about this some more and I think I have a cleaner approach. Rather than using different base classes...

> Hi @tcbrindle > > > Something else that has just occurred to me: for the "power" versions, the cursor type could be `std::array` rather than `tuple`... I don't know...

> Hi @tcbrindle > > > Yeah, from a pure design perspective bundling everything into one big class that contains every code path is certainly not the recommended way! But...

> > You could rename this to `repeated_invocable` and re-use it in both places. > > I moved it to `flux/core/concepts.hpp`, which I've just realised is probably the wrong place,...

I think `min()` itself [wouldn't need to be changed](https://github.com/tcbrindle/flux/blob/87b50c52b74123442e981ff1394119af9db7cf84/include/flux/op/minmax.hpp#L23), as it just calls `fold_first()`. But we could implement *that* as something like: ```cpp template auto fold_first(Seq&& seq, Func func) ->...

> I think too many customization points can confuse users. I agree in general. In this case though I think the argument in favour of adding one more -- in...

For anyone else who finds themselves here because Doctest no longer works with CMake 4, I've created a temporary fork of the master branch that fixes the issue. Feel free...