Hi all,
someone on Freenode referred me to this link:
http://codegolf.stackexchange.com/questions/43150/build-a-fewest-moves-freecell-solver
It is a challenge to build a Freecell solver that generates short solutions for
the first 32,000 MS Freecell boards, where the tie breaker is the solver having
the shortest code. It has yet to receive a solution , but received some
comments with questions and answers.
My personal opinion is that this challenge ignores the fundamental trade-off of
Freecell solvers, which is of average short solution length vs. speed.
Normally, one can have fast heuristics that generate long solutions, or
relatively slow heuristics that generate short solutions on average, or
heuristics that are somewhere in between on both counts. So the
codegolf.stackexchange.com challenge can easily result in excessively long
runtimes, unless possibly it is a cheating solver which emits prerecorded
solutions which will increase the code size significantly.
I also think that writing a good Freecell solver is too complex a job for code
golf (see
https://en.wikipedia.org/wiki/Code_golf ) and so not a very
interesting problem for that.
Regards,
Shlomi Fish
--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/
http://is.gd/KNvczZ - The FSF Announces New Versions of the GPL
There are only two hard things in Computer Science: cache invalidation and
naming things. — Phil Karlton
Please reply to list if it's a mailing list post - http://shlom.in/reply .
Received on Wed Nov 18 2015 - 16:16:34 IST