Not really a surprise considering all the ADL shenanigans going on and the C++20 additions, but now we've got tough issues to fix. The problem is evidenced when trying to compile the test schwartz_adapter_every_sorter_reversed where std::reverse_iterator overloads of iter_swap become ambiguous.
Interestingly the issue wasn't introduced with the latest change to iter_move/iter_swap in cpp-sort, but also occurs with the old model.
Not really a surprise considering all the ADL shenanigans going on and the C++20 additions, but now we've got tough issues to fix. The problem is evidenced when trying to compile the test
schwartz_adapter_every_sorter_reversedwherestd::reverse_iteratoroverloads ofiter_swapbecome ambiguous.Interestingly the issue wasn't introduced with the latest change to
iter_move/iter_swapin cpp-sort, but also occurs with the old model.