On Sat, Nov 21, 2009 at 3:09 PM, Gwern Branwen
This doesn't surprise me; but how much slower are we talking?
If it's not at the point that a browser of a Gitit wiki could notice the difference, then it seems to me that the dep ought to be loosened: the parsec/quickcheck/base diamond dependency problem is one of the worst ones a user can run into, the hardest to resolve, and one that can arise in the course of ordinary safe use of Haskell.
Running 'pandoc --strict' over the Markdown readme.text takes: ~0.09s with pandoc built against parsec-2 ~0.19s with pandoc built against parsec-3 on my machine. I have a branch of parsec-3 which seems to brings us back to parsec-2 numbers, but also fails the rst-reader test-case in the pandoc testing suite: http://community.haskell.org/~aslatter/code/parsec/cps Antoine