Thanks Manuel for the heads up.
Don't worry, it had nothing to do with type families. I just was a
bit sloppy yesterday (I would swear I had amended that patch..).
I guess I could have avoided this if I were running validate in a
separate darcs repository, apart from the development one.
Che
Manuel M T Chakravarty wrote,
I just ran validate on the current head.
Unexpected failures:
break003(ghci)
break006(ghci)
break012(ghci)
break013(ghci)
break018(ghci)
These have some slight permutation of the output. I don't think any
of that changes the meaning or quality of
I just ran validate on the current head.
Unexpected failures:
break003(ghci)
break006(ghci)
break012(ghci)
break013(ghci)
break018(ghci)
break021(ghci)
I guess this is due to simultaneous changes in both the type
families patch and ghci debugger code. (Ie, both changes have b
