Daan
Daan
Interesting -- make sure to share your results :-) The compiler can only prove a direct recursive function can terminate if you match explicitly on an inductive datatype; I think...
Fixed in #104 :-)
Ouch -- that is no good :-( I am going to look into this to see what is the cause -- in principle there should be no such performance surprises....
Great to see your nice code! You are a koka expert already :-) I have been playing with the sample and made it about 3x faster (see next post). When...
With regard to the Koka code; I tried: - koka is not yet a good optimizing compiler (except for refcounting optimization). For example, it does not do CSE (yet) and...
Very nice code btw. Surprised to see such large complex structure implemented in Koka :-) It worries me too -- fingertrees seem quite complex :-) Some things I noticed: -...
Ouch, confusing indeed! Thanks for finding this -- I'll fix it :-) Changed to: * `(handler{ fun emit(msg){ println(msg) } })( fn(){ hello() } )` * `finally(fn(){ println(...) }, fn(){...
Ah good remark! I had to think twice when I saw your response :-) `handler{ a ) -> e a` for an effect `l`. And the desugaring of `with handler...
Yes -- I think so; give the `--console=raw` to suppress color output
Nice you are working on collections; Did you read about size-balanced trees as I used in the Haskell Data.Map implementation? or the integer Patricia trees in Haskell's Data.IntMap ?