
18 Jul
2009
18 Jul
'09
8:49 a.m.
Am Samstag, 18. Juli 2009 08:58 schrieb Miguel Mitrofanov:
Oops... Sorry, wrong line. Should be
isAB :: forall p. p A -> p B -> p x
Is this a well-known approach for closing classes? I came across the same idea and implemented a generic framework for closing classes in this way. I did this to simulate algebraic data kinds and kind polymorphism over such kinds. I needed this for the record system of Grapefruit. The code is here: http://code.haskell.org/grapefruit/main/grapefruit-records/src/ Explaination of the techniques used in this code will probably follow as part of an IFL 2009 paper. Now I wonder which of my ideas are actually new and which are just old hat. Could you maybe answer this question? Best wishes, Wolfgang