Hi Simon, --- Simon Peyton-Jones <simonpj@microsoft.com> wrote:
I don't know why you would possibly want this.
yes, in this example there isn't any obvious reason to motivate us writing such instance, but I am interested in how GHC now handles coinduction in type class. It will be neat if we can express coinductive algorithm in terms of Haskell type class. Currently I am working a project with Dr Martin Sulzmann in which we think such an extension is neccessary.
So far as I can see from your message, the recursive dictionary you are trying to build is exactly the same as what GHC will build automatically from the "conventional" instance declaration. Yes, exactly, I realized that after I compiled the code with -fext-core flag. Is it because it tries to avoid building infinite evidence constructors?
Thanks for pointing out. Regards, Kenny LU Zhuo Ming Research Assistant School of Computing National University of Singapore __________________________________________________ Do You Yahoo!? Faster. Easier. Bingo. http://sg.search.yahoo.com