https://gcc.gnu.org/bugzilla/show_bug.cgi?id=124890
--- Comment #5 from Lénárd Szolnoki <leni536 at gmail dot com> --- (In reply to Patrick Palka from comment #4) > Perhaps we could just note in LWG 4218 that std::expected and std::indirect > are also affected, rather than creating new issues for them. Possibly. The only slight concern I have with that is that the operators affected by LWG 4218 have "Constraints" or requires clauses, while the affected indirect::operator== overload only has "Mandates" in the current draft. It could be argued that this should be required to be SFINAE-friendly too, but that seems to be an extra thing compared to the existing cases in LWG 4218, where this isn't in question. The circular constraint only manifests if an implementation tries to make the overload SFINAE-friendly.
