Hi,
Throwing away Yhc would be a bit depressing - I think we can come up with a different front end with different goals and different priorities. I think we can structure it differently, and contribute something fresh in some places. It is also important for GHC to have a competitor, just to keep Haskell "Clean" - a standard with one implementation is like a democracy without voting :)
Please don't throw away YHC I think it's important to have various implementations of Haskell around. We don't want the language to become stale.
However in the meantime having some way to use GHC as a front end will be a definate benefit, therefore I propose:
1) Add a GHC <-> Yhc Core translator (me, unless someone else volunteers!)
What I would find very useful is a standard interface to all Haskell front ends. Would it not be a better idea to make the API for GHC and YHC appear the same? That way one could plug a ghc or yhc front end onto their system and it would just work with the minimum of change. I say this because porting HaRe to GHC (from programatica) proved to be an enormous job. It would have been made so much easier if both front ends at least some common structures remaining the same. Regards, Chris.