RE: GHC not able to detect impossible GADT pattern

2014-09-03 Thread Simon Peyton Jones
I believe this is probably an instance of https://ghc.haskell.org/trac/ghc/ticket/3927 There are numerous other similar tickets, about GHC's inadequate/misleading warnings for non-exhaustive patterns. A selection is #595, #5728, #3927, #5724, #5762, #4139, #6124, #7669, #322, #8016,

Re: GHC not able to detect impossible GADT pattern

2014-09-03 Thread Merijn Verstraaten
Hi Simon, Thanks for the clarification, I attached the code to that ticket. Does this mean improvements in this area are supposed to land in 7.10? Semi-relatedly, before attempting the current code (which at least compiles), I had a few other attempts and I was hoping you (or other readers)

RE: GHC not able to detect impossible GADT pattern

2014-09-03 Thread Simon Peyton Jones
This is a bug in 7.8.3: https://ghc.haskell.org/trac/ghc/ticket/9433 You should get this error (and do in HEAD): Sock.hs:28:5: Type family 'Readable' should have 1 argument, but has been given none In the equations for closed type family 'Foo' So your program is indeed wrong, but the

[Haskell] Haskell Weekly News: Issue 304

2014-09-03 Thread Daniel Santa Cruz
Welcome to issue 304 of the HWN, an issue covering crowd-sourced bits of information about Haskell from around the web. This issue covers from August 24 to 30, 2014 Quotes of the Week * pjdelport: Haskell: Lazy evaluation, eager debugging. * trap_exit: isn't hoogle awesome? the first