Good talk. The NaN issue already exist when putting floating point numbers in associative containers as key. Btw the TIE advanced is a TIE fighter as well if you follow the lore of the game 'TIE fighter'
18:20 I think you are wrong. The defaulted operator== will automatically have a noexcept-ness based on whether or not all of it's members had noexcept operator==.
Could you in theory use the same idea of swapping the order of arguments to create some kind of generic [[commutative]] directive that works for any operator? I'd love to not need to define both scalar * vector and vector * scalar, for example...
Good talk. The NaN issue already exist when putting floating point numbers in associative containers as key. Btw the TIE advanced is a TIE fighter as well if you follow the lore of the game 'TIE fighter'
18:20 I think you are wrong. The defaulted operator== will automatically have a noexcept-ness based on whether or not all of it's members had noexcept operator==.
Could you in theory use the same idea of swapping the order of arguments to create some kind of generic [[commutative]] directive that works for any operator? I'd love to not need to define both scalar * vector and vector * scalar, for example...
Seems to say partial ordering is "a bit weird" :) but that's almost all graph nodes ...
Spaceship operator 👍
Materials from presentation author: www.jonathanmueller.dev/talk/meetingcpp2019/
Really great talk with a lot of useful information. Thank you! I personally rate is as the best i've seen on .