On Sun, 7 Jul 2002, Adrian Ettlinger wrote:
> Hi Shlomi,
>
> This is a separate reply to your other (second?) message.
>
> <<Adrian, your mailer does not seem to have cut the long lines in this
> message. Please fix its configuration for next time.>>
> I don't understand what you mean by this.
>
It seems to be OK now, but I'll explain. Sometimes messages are sent with
no line-feeds separating long paragraphs into separate lines. Then, some
UNIX mailers display them all on one replied to line:
> A really really long line......................
Which is quite annoying because I have to format such letters manually.
> <<It was not particularily weird - I just forgot to initialize a few
> things.>>
> I guess any bug that takes a long time to run down seems weird to me.
> It had to me a very weird appearance until, first, I was able to define the
> circumstances in which it was occurring, then, after all, you at first
> didn't think it was there because you couldn't replicate it. Maybe a better
> word than "weird" would have been "difficult". I'll point out, also, that
> you didn't know it was there until I started doing extensive testing with
> intermediate positions.
>
Agreed.
> <<Actually, version 2.6 can run either with atomic moves, or without them.>>
> Yes, I understood that.
>
> <<Use of atomic moves seems to make it considerably slower.>>
> That's not so good. I might want to do something to it so that it will
> only run with an atomic move preset following an "impossible" verdict with
> non-atomic moves. But for zero and one freecell, where a majority of the
> verdicts will be "impossible", it should probably run with atomic all the
> time.
>
Whatever, it all depends on the amount of logic you'd like to put there.
> <<I suggest we give the users a choice between several presets some of them
> would be based on atomic moves and some of them would be based on
> meta-moves.>>
> I agree, and also I hope we'll find strategies tailored for the number
> of freecells down to zero.
>
So far I only benchmarked the scans for four freecells. I suppose I can do
the same for a lesser number of Freecells.
> <<Are you sure you checked it against Patsolve's speedy mode? >>
> Yes, Patsolve's speedy mode is very close in speed to FcPro's. But
> Patsolve is still superior because it's (so far as we know) free from false
> impossibles. FCS is twice as fast as them, but it's not really a fair
> comparison, because FCS is presently false-impossible-prone. The real test
> will be the comparison in FCS's best atomic move preset.
>
I'm still working on the Atomic Moves Preset, so stay tuned. I'm also
considering integrating Patsolve's logic into Freecell Solver. I looked at
the patsolve's code and was able to tell what all the rating numbers mean,
so I might implement such a mechanism in Freecell Solver too. (only with
user-definable ratings)
> <<In any case, the development version of Freecell Solver, with the computer
> generated solving presets, can solve
> the MS 32,000 in roughly three times the speed of the best preset I could
> find for 2.4.x>>
> Oh, then, that's very good news. It means you have a factor of six
> improvement over Patsolve's speedy mode. So if atomic moves slow it down by
> less than a factor of six, you should still be ahead.
>
Unfortunately, they seem to be a great deal worse.
> <<Like I said, I'll put it on my web-site. However, I do suggest that
> Michael Keller's site would carry the most up-to-date version.>>
> Mike Keller has been very silent lately. I think he's been having some
> personal problems. You might recall that several months ago he raised
> vehement objections to releasing FcPro on an open-source basis. I don't
> think he realized, however, that the distribution of the source can be
> handled in the way I am handling it. I've asked him a number of times
> whether he'd object to what I'm doing, but he's never answered me. So I
> decided to go ahead without his explicit approval. He has acknowledged that
> FcPro is my own property (or mine and Wilson Callan's, but Wilson really no
> longer has any interest), and I can do anything I want with it. But
> previously he had threatened that if I released it open-source, he'd stop
> providing it on his website. So I don't want to provoke him by urging him
> to carry it. --- But furthermore, I don't consider the Solver Evaluation
> Edition of FcPro to necessarily be a later "up-to-date version", It is a
> different version for a different purpose, and not necessarily intended for
> the general Freecell player's use. Depending on how the solver situation
> resolves itself, I can see the possibility at some point in the future of
> substituting FCS for the present solver, and having the standard release of
> FcPro be with FCS as its only solver. -- But I don't know if Mike Keller
> would necessarily go along with that if we had to make that version also
> open-source.
>
> Also, I don't know if Mike is regularly checking your forum. If he is,
> we'll see whether or not he reacts to the posting of FcPro V6.5. I have no
> objection to your writing directly to Mike if you want to, but I think it
> would be better to wait to see if we get any reaction from him.
>
Whatever. We can always distribute Freecell Pro on a different web-site.
Regards,
Shlomi Fish
> Best regards, --------------------Adrian
>
>
>
>
> To unsubscribe from this group, send an email to:
> fc-solve-discuss-unsubscribe_at_yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
----------------------------------------------------------------------
Shlomi Fish shlomif_at_vipe.technion.ac.il
Home Page:
http://t2.technion.ac.il/~shlomif/
Home E-mail: shlomif_at_iglu.org.il
He who re-invents the wheel, understands much better how a wheel works.
Received on Sun Jul 07 2002 - 05:37:23 IDT