#10365: Implement Semigroup as a superclass of Monoid Proposal (Phase 1) -------------------------------------+------------------------------------- Reporter: gidyn | Owner: quchen Type: feature request | Status: patch Priority: high | Milestone: 8.0.1 Component: libraries/base | Version: 7.10.1 Resolution: | Keywords: report-impact Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: None/Unknown | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Phab:D1284 -------------------------------------+------------------------------------- Comment (by bgamari): I realize that this has [[https://mail.haskell.org/pipermail/libraries/2015-May/025648.html|all been discussed before]] but I do wonder whether we really ''need'' to fold in `Data.List.NonEmpty` at this juncture? It does not appear to be a widely used type in practice and the interface has a few quirks. What do we gain by it being in `base`? The reason I ask is that I've noticed that `NonEmpty` introduces yet more partial functions to `base`. For instance, why must `words` have type `NonEmpty Char -> NonEmpty String`? Why not `NonEmpty Char -> [String]`? After all it seems quite reasonable for a non-empty string to nevertheless contain no words. Given that these are design decisions that reasonable people may disagree over, might it be best just to defer them packages outside of `base`? Sorry bringing this up so long after the initial consideration but I just wanted to make sure we have thoroughly considered this angle. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/10365#comment:20 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler