begin quoting Carl Lowenstein as of Tue, Jan 08, 2008 at 03:11:46PM -0800: > On Jan 8, 2008 2:08 AM, SJS <[EMAIL PROTECTED]> wrote: > > > > I skimmed thru the code and wrote down my initial reactions, and > > then followed 'em up with the mitigation or rebuttal of the criticism. > > Not much rebutting, and darn little mitigation going on here. > > > > Most of it's pretty general, as I didn't have the patience to do a > > line-referencing full on code review. I'd've ended up rewriting it. > > Looking at it from the puzzle point of view rather than programming, > there is one fatal flaw. A simple parity argument shows that 1/2 of > the plausible solutions can never be reached. Just swap two tiles in > the target and you can't get there from here.
I believe you, but I'm afraid I'm not seeing it. > Starting with '2468135 7' and trying for '12345687 ' > > After running for 60 seconds it has expanded to occupy 2 GB of RAM and > is still trying. I killed it. Well, yah, it's going to brute-force all possible solutions if it can't find the proper one. Starting with 'abcdefgh ' and looking for 'ABCDEFGH ' will cause all manner of fun. > This is of course a simplified version of the "Fifteen Puzzle", > attributed to Sam Loyd, circa 1890. Yup. It's not a new puzzle at all. -- There are configurations you can arrange on a Rubick's cube That can only be achieved with a screwdriver and graphite lube. Stewart Stremler -- [email protected] http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-lpsg
