On Wednesday 01 April 2009 19:29:54 jgw321 wrote:
> On Tue, Mar 31, 2009 at 3:17 PM, Shlomi Fish <shlomif_at_iglu.org.il> wrote:
> > > I am also going to download the solver now there is
> > > a Windows install version.
> >
> > Nice. Please let me know (either via the mailing list or via private
> > email) which problems you encounter and what you think about it.
>
> Will do. It will be a day or two before I find the free time to do it.
> Now I'm retired and in the U3A I seem to be busy all the time.
>
OK. :-)
> > > I was involved with a free cell solver many
> >
> > > years ago when I was a member of CIX.
> >
> > Is it this CIX: http://en.wikipedia.org/wiki/CIX ?
>
> Yes that's the one. It was the place for online information in those days
> when Alta Vista was the prime search engine, Yahoo was just getting going
> and Google hadn't been thought of.
>
I remember the Alta Vista days. And the Lycos days before that. Yahoo started
out as a web-directory (similar to
http://en.wikipedia.org/wiki/Open_Directory_Project ), and now no longer
really is.
> > Can you tell us more about
>
> > that solver you were involved with?
>
> We had a CIX conference devoted to proving (and finding) solutions to all
> the Freecell games. In those days the only Freecell (of note) was the free
> microsoft one with 32k games. We succeeded in solving all the games except
> the one impossible one. One of our members cracked the deal behind the
> Microsoft version and we developed a player and extended it to 64k games. I
> proved that all of the games in the 64K version were unique. Since this
> was in Win 3.11 days, pre Pentium, I coded the deals down to a linear
> string and wrote them out to a file which I then transmitted to an IBM
> mainframe, sorted in to alphabetic order on the code string and ran the
> result through a duplicate checker I wrote in Assembler (which is my lingua
> franca). Seems long winded now.
Cool. You could have kept a collection of the hash values of the states and/or
use something a database like Berkeley DB.
>
> As far as the solver went I was invloved in the beta testing of it, but not
> the coding. I will have to dig out the old cix conference and look up more
> detail. All I can remember now, without checking back, is that it was
> initially slow and could get stuck in shuffling loops moving from one stack
> to another and then back again.
This was solved in Freecell Solver by normalising positions by sorting their
columns lexicographically.
> To over come both problems hashing was
> introduced and storing of positions reached. The solver, initially stand
> alone, was later incorporated into the Freecell player. The player was
> quite good, as we all contributed ideas to its development and helped in
> the testing. It stored all the solved games and would restore new
> soloutions if better, either based on time or number of moves (it could do
> either based on a configuration option). The player was later extended to
> some huge number of games (I forget how many) but to my knowledge it has
> not been proven that they are all unique.
Interesting.
> The player was available on the
> web at one time but the chap who developed it had to give up due to lak of
> time. When I retired, I offered to take it over but he had lost the source
> by then.
Too bad.
Regards,
Shlomi Fish
--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/
"The Human Hacking Field Guide" - http://xrl.us/bjn8q
God gave us two eyes and ten fingers so we will type five times as much as we
read.
Received on Wed Apr 01 2009 - 13:18:37 IDT