Dlists maybe good it all the app is written using them. Probably not
good idea to switch to them in the middle of project...
I know it is lazy, but I don't think it is able to eliminate operations,
is it?
At least intuitively, the map f list takes n*C ticks (C is for
application of f and list "creation", n is the list length, f is of no
importance, it is always the same, but list probably must be created due
to ++).
Then, (++) take n*K ticks (K for list creation - I want to write out the
list at the end, so that it is created).
In my case (mapapp), it is n*CK, where CK stands for f and list
creation... the CK is very similar to C... Thus, I should save the n*K,
or at least its large portion... shouldn't I? If not, how the compiler
can eliminate the operations?
Dusan
Bulat Ziganshin wrote:
Hello Dusan,
Tuesday, August 18, 2009, 2:50:38 PM, you wrote:
but with less efficiency. Or am I wrong?
probably wrong. haskell is lazy language
also there is differential lists (dlist) implementation on hackage
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
