Dear Haskell implementors, Please, always try to provide the working profiling. For it appears indeed very helpful. I had made many experiments with certain large program. Spent many days in this. And it was very hard to find out the practically most expensive parts of the program. Finally, I gave up and waited for some admissible profiler to come. Now, as it was applied, one day of experiments revealed a couple of unexpected earlier main expense centers in the program. I realized then that these parts may indeed occur the most expensive. Then, tested this independently, and it confirmed that the profiler was right. ------------------ Sergey Mechveliani [EMAIL PROTECTED]
