#14389: Improved results in GHCi during basic calculations -------------------------------------+------------------------------------- Reporter: vanto | Owner: (none) Type: feature request | Status: closed Priority: normal | Milestone: Component: Compiler | Version: 8.2.1 Resolution: invalid | Keywords: Operating System: Unknown/Multiple | Architecture: | Unknown/Multiple Type of failure: Other | Test Case: Blocked By: | Blocking: Related Tickets: | Differential Rev(s): Wiki Page: | -------------------------------------+------------------------------------- Changes (by svenpanne): * status: new => closed * resolution: => invalid Comment: Sorry to say, but this ticket is basically the same as your previous tickets #14377 and #14384. The bottom line is: The floating point numbers/operations available on normal HW are fundamentally different from the usual numbers/operations from math. If you want to understand the details, please have a look at ''the'' paper in this area "What Every Computer Scientist Should Know About Floating-Point Arithmetic" (available at e.g. http://www.validlab.com/goldberg/paper.pdf). Your `exp`/`log` result is therefore fully expected, and the same holds for the last example, although there you can calculate in a different domain by adding a `:: Rational` after it. Before opening more tickets in this area, please have a look at that paper first and try to understand Haskell's defaulting mechanism, too. -- Ticket URL: http://ghc.haskell.org/trac/ghc/ticket/14389#comment:1 GHC http://www.haskell.org/ghc/ The Glasgow Haskell Compiler