On Monday 30 March 2009 21:25:15 Gary Campbell wrote:
> ----- Original Message -----
> From: "Shlomi Fish" <shlomif_at_iglu.org.il>
> To: <fc-solve-discuss_at_yahoogroups.com>
> Sent: Monday, March 30, 2009 10:38 AM
> Subject: Re: Release 2.20.0 is available now with an installer for Windows
>
> > My solver gives an API (as it is compilable as a dynamic or static
> > library), and a command-line interface that are both available for use by
> > any interested party. It exposes a lot of functionality that is not
> > directly available with its FC-Pro integration, and many people may find
> > it more convenient. The Windows package is just the binaries of the
> > source, that is also downloadable from http://fc-solve.berlios.de/, and
> > can be studied, modified, re- distributed, and almost any other
> > acceptable use.
> >
> > I should note that Freecell-Solver-2.20.0 has progressed beyond its
> > release that was integrated into the latest version Freecell Pro, and so
> > it is advisable to use it instead.
> >
> >> Is it merely an example of your programming
> >> standards and practices, or is it a free cell solver
> >> with significant added value in its own right?
> >
> > See above. Its API and CLI may allow things that the FC-Pro integration
> > (or any other embedding of it) may not allow, and so may prove useful to
> > interested users or developers.
>
> I'm afraid your explanation is too vague for me.
> I have no idea what your solver's solution times
> for hard games would be,
Which hard games are you interested in? And I should note that FCS has many
possible run-time configurations, each one potentially performing better on
some games than others. I've constructed some aggregate scans, that on average
perform better than others. See the "-l" flag for more details.
> its average solution
> length,
I haven't measured this, though I can easily. The question is with which
preset are you interested. Would "-l gi" - good-intentions be good enough for
you?
I should note that Freecell Solvers's solutions explicitly includes many moves
that are implicit in MS-Freecell/Freecell-Pro. As a result, the solution may
be artificially longer than a strictly FC-Pro compatible solver that
automatically moves some cards into the foundations. Which of the two options
are you interested in?
> its speed over a range of games,
To quote
http://tech.groups.yahoo.com/group/fc-solve-discuss/message/887 :
{{{{{{{{{{{{{{{{{
Added the FCS_FREECELL_ONLY compile-time flag to hard-code the settings
for Freecell and thus allow faster run-time. On a Pentium 4-2.4GHz machine
running Mandriva Linux Cooker, this allows to solve the Microsoft 32,000 in
194.56353 seconds ( 164 deals / second ) instead of
228.84 seconds for the generic version ( 140 deals / second ).
}}}}}}}}}}}}}}}}}
I only tested it against the first 32,000 deals.
> and its
> percentage of false negatives or intractables.
Freecell Solver supports atomic moves, so assuming there are no bugs in the
code, solving configurations that end up using atomic moves, will have no
false negatives.
I'm a bit more hazy about intractables. Freecell Solver has an arbitrary
(~2**32) limit on its resources so it may go on searching forever given enough
time and resources. If you want me to determine the number of intractables for
a certain iterations' limit - that would be doable, though may be somewhat
time consuming.
> A few more concrete reasons for why anyone
> would want to work with it would be helpful.
I've enumerated its major features here:
http://fc-solve.berlios.de/features.html
One highlight is that it has run-time-configurable support for solving several
different variants of Solitaire, similar to Freecell. Can your solver play by
any other rules aside from Freecell?
>
> Also, does your solver only work standalone?
> Or is it integrated with a free cell player, so that
> games can be interactively reviewed and layouts
> input and output?
My solver is essentially an ANSI C library, and some standalone command line
clients, and both options can be integrated into GUI Solitaire players for
whatever use one may find appropriate (reviewing a solution, providing hints,
play-against-the-computer, etc.).
FCS has been integrated into 4 GUI front-ends so far:
http://fc-solve.berlios.de/links.html#front_ends
Other people have done the actual integration for me, but they are still
calling the Freecell Solver code.
Regards,
Shlomi Fish
--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/
Parody on "The Fountainhead" - http://xrl.us/bjria
God gave us two eyes and ten fingers so we will type five times as much as we
read.
Received on Mon Mar 30 2009 - 12:42:42 IDT