[Kde-games-devel] Re: Kpat solvers

[Shlomi Fish]

Hi Stephan,

On Thursday 27 Jan 2011 11:46:03 Stephan Kulow wrote:
> Hi,
> 
> Just so you know: I subscribed now and I don't like the approach the patch
> is taking.
> 

Thanks for subscribing. It's fine that you don't like the approach it took, as 
it was intended as a crude first version and a proof-of-concept. I didn't 
intend it to be used in production as is.

> I'm fine with having a 2nd implementation of freecell and simsimon solvers,
> but patching the fc-solve library into the patsolve algorithm is very, very
> ugly.

I guess you're right.

> 
> So from my perspective you should send individual patches to
> - add the library functions you need to dump the board

Do you mean «QString Simon::solverFormat() const» and 
«QString Freecell::solverFormat() const» ?

> - change the rules of simple simon - but I find the current way of simply
>   allowing meta moves without any animations very confusing. So I would
> reject that patch if it's just a snippet from the current patch. 

OK, should I make heterogeneous sequence moves animated? Is it handled in 
freecell.cpp in a similar way?

> I would
> prefer solving the original game

What do you mean? If I understood you correctly, then PySolFC now implements 
heterogeneous sequence moves in Simple Simon as one move for Simple Simon that 
both the user and the automated solver can use (without intermdiate step 
animations that are not present for Freecell as well.).

> - make the interface to the solver abstract to allow plugging in different
>   solvers in there (note that Parker discussed that independent of you and
> we have goals that will be harder to achieve with your current work, so it
> would have been better if you contacted me earlier about your plans) - add
> solvers for freecell that will replace the patsolve ones if fc-solve is
> found as library. I don't want fc-solve as hard requirement. fc-solve is
> surely very fast and more correct, but if it takes 2 or 800ms to solve the
> patience game is of little interest in the context of kpat.

OK.

> 
> Beside that: I tested your patch and simple simon game 177 kills my
> computer while solving. 

Yes, I see that too.

> It looks like you didn't implement limits.
> patsolve is configured to give up solving after using 30MB.

It shouldn't be hard to do it for my solver as well, as it has 
freecell_solver_limit_iterations and other mechanisms for that. I'll work on 
it.

Regards,

	Shlomi Fish

> 
> Greetings, Stephan
> _______________________________________________
> kde-games-devel mailing list
> kde-games-devel at kde.org
> https://mail.kde.org/mailman/listinfo/kde-games-devel

-- 
-----------------------------------------------------------------
Shlomi Fish       http://www.shlomifish.org/
First stop for Perl beginners - http://perl-begin.org/

Chuck Norris can make the statement "This statement is false" a true one.

Please reply to list if it's a mailing list post - http://shlom.in/reply .