but there's no risk using trace is there? t.
The unsafe... functions are called that for a reason, and their use should be highly discouraged, except in cases where there is absolutely no other reasonable way.
"Yitzchak Gale"
computation is. And since it's all calls to map and filter et al., it's ...it's not immediately clear how to provide any feedback on how much longer there is to wait.
Andrew Coppin wrote:
It seems unsafePerformIO is the way to go here. ...unsafeInterleaveIO
I disagree. The unsafe... functions are called that for a reason, and their use should be highly discouraged, except in cases where there is absolutely no other reasonable way. I don't believe that this is one of those cases. A better approach is to use pure monads. Instead of f :: ... -> a write functions whose type is something like f :: ... -> MyMonad a where MyMonad is defined in one and only one place in your program. Or, better yet, polymorphic things like f :: Monad m => ... -> m a f :: MyMonadClass m => ... -> m a Then when you later need to add something like progress indicators, you just add that to the capabilities of your monad, and add some updateProgress calls to your functions. By using "do" notation wisely, you can keep your abstractions just as clear as they were, and cleanly separated from the progress logic. Other things you may need to add later in real-world programs are exception handling, logging, trace, etc. All of these are easy if you started out in a monad. -Yitz _______________________________________________ Haskell-Cafe mailing list Haskell-Cafe@haskell.org http://www.haskell.org/mailman/listinfo/haskell-cafe --- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.