Re: Proposal: Num instance for (a -> b)

On Sat, 10 Nov 2018, Daniel Cartwright wrote:

relevant reddit comment thread:https://www.reddit.com/r/haskell/comments/9vtis5/the_universe_of_discourse_i... urce=reddit-android

https://wiki.haskell.org/index.php?title=Num_instance_for_functions&oldid=36632 In short: It would make 2(x+y) no longer a type error but equivalent to 2. We would lose a lot of type safety for little syntactic gain.