As can Danny Jones and Gary Campbell’s solvers which Shlomi seldom mentions.
From: mailto:fc-solve-discuss_at_yahoogroups.com
Sent: Monday, May 19, 2014 8:25 AM
To: Freecell Solving Discussions
Subject: Scientific Paper about a Freecell solver: “GA-FreeCell: Evolving Solvers for the Game of FreeCell”
Hi all,
I discovered the following Scientific paper about writing a solver for Freecell
in a Google search:
*
http://www.genetic-programming.org/hc2011/06-Elyasaf-Hauptmann-Sipper/Elyasaf-Hauptmann-Sipper-Paper.pdf
*
http://www.genetic-programming.org/hc2011/06-Elyasaf-Hauptmann-Sipper/Elyasaf-Hauptmann-Sipper-Slides.pdf
Found in this search and others:
https://www.google.com/search?q=freecell%20solver%20ga
They describe a Genetic Algorithm approach to creating a solver for Freecell,
which in turn can solve 98% of the MS 32,000 games (which is the best performing
solver in the Scientific literature), with other improvements on the previous
most performant solver, such as solution length, time, etc. .
I found the description of the method there somewhat hard to follow due to the
proliferation of jargon, but still read the entire article.
Just for the record:
1. On my Mageia Linux x86-64 v5 system running on a Intel
x86-64 Core i3 machine, Freecell Solver can solve all the Microsoft 32,000
games (except for the unsolvable #11,982 which it reports as such) in under 20
seconds. That's with the default install of the program under the /usr/bin
directory. After applying all aggressive compiler and system optimisations, the
solver solves them in between 9 seconds and 10 seconds.
2. In the default configuration, the solutions are neither outputted nor
verified to be correct. Nevertheless, I've written a program using Freecell
Solver and using
http://fc-solve.shlomifish.org/verify-code/ that did all that
and took about 10 minutes to run.
3. Other open source solvers such as the Woods-Callan-Ettlinger solver and Tom
Holroyd's Patsolve can also solve all the relevant MS 32,000 deals at a
relatively short time.
----------------
So it seems like the solver that is discussed in the paper is underperformant.
Regards,
Shlomi Fish
--
----------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/
List of Text Editors and IDEs - http://shlom.in/IDEs
If the mountain does not come to Muhammad, then Chuck Norris will bring the
mountain over.
— http://www.shlomifish.org/humour/bits/facts/Chuck-Norris/
Please reply to list if it's a mailing list post - http://shlom.in/reply .
Received on Mon May 19 2014 - 09:30:03 IDT