7 Feb
2003
7 Feb
'03
1:41 p.m.
On Friday, 2003-02-07, 12:31, CET, Arjan van IJzendoorn wrote:
[...] Yes, that does mean adding type classes, but not the whole machinery. If we support Eq, Ord, Show and Num with a limited number of instances, chapters 1 to 11 of Hudak's book can be used without modification.
I just have had a look at Helium and think that at least a deriving mechanism is absolutely needed. The need to write down obvious implementations for equality is very disturbing, IMHO. Furthermore, I think, the need to state equality functions explicitely as paramters of certain functions makes learning harder and not easier. But it's an interesting project. Keep going on it.
[...]
Wolfgang