Deep Blue--Part III
Keith Ian Price
Sorry for the delay. Here is Part III of my Deep Blue Report:
8.There was a rather long thread going on about a month ago on rgcc,
concerning whether DB was an example of Artificial Intelligence.
During his presentation, Hsu gave his opinion on the subject. He
stated that chess is considered a game of intelligent people, and DB
was able to play the game against the best player in the world, and so
therefore, it could be argued that DB had passed a Turing test of
sorts, albeit a chess-specific Turing Test. However, Hsu continued, he
did not think that this constituted intelligence. He did not directly
support his decision, but did show a cartoon that occurred just after
the match, in which Kasparov is playing Deep Blue and Kasparov's foot
has slipped under the plug for Deep Blue. Both have "thought balloons"
showing in the cartoon. DB's balloons show a bishop with diagonal
arrows; a rook with horizontal and vertical arrows; and a King with
short arrows going in all directions, etc. Kasparov's balloon shows
his foot lifted, and the plug out of the socket. He said this
represents the difference between a chess-specific intelligence, and
real intelligence. If DB were losing, it would have no way to think of
a different solution outside the bounds to which it had been
programmed.
9. One of the longest-running arguments on rgcc and CCC has been how
well micros might fare against Deep Blue. During the Deep Blue
excitement last year the news slipped out that there had been a match
between DB, Jr. and Rebel 8 and Genius. DB, Jr. was supposed to have
been slowed down to match the PC's speed somehow. I asked Hsu about
this 10-game match. He was quite familiar with the results. He
confirmed that there had been 5 games against each opponent. He stated
that there was only one chess processor used, and that it's clock
speed had been halved. He also said that several pruning algorithms
were turned off, with some selective extensions, in order to emulate
the performance of the micro hardware as much as possible. They did
this to see how well they did against the micros on an evaluation
specific level, keeping the speed advantage down to the difference
between what the micros could evaluate given their nps levels, and
what could be accomplished in the chess specific processor evaluation,
rather than how many nodes were searched. Since the speed of a single
chess processor is about 2-2.5 million nodes per second, and Hsu
estimated that the removal of the algorithms caused a 5-10 times
reduction in nodes searched, the probable nps level for DB, Jr. was
somewhere between 100,000 and 250,000 with the clock speed reduction
factored in. This is similar to the fast searchers, but is probably
2-5 times faster than Rebel 8 at the time. In any case, I asked how
the games went, and Hsu pulled no punches. He said that the
performance of the micros was much poorer than he had imagined they
would be. He said all 10 games were basically blowouts. When I asked
for specifics, he mentioned two examples against Rebel that had
surprised him as to how little understanding they had of endgames and
King safety. In the first example, the ending was with bishops of
opposite color and normally would have been a draw. Rebel allowed an
exchange which gave DB two widely separated passed pawns, and there
was no way to stop both. Rebel did not realize until a few more moves
that it was in trouble. Hsu said this was the kind of thing that is in
his evaluation routines, and he was surprised that it was not in
Rebel's. The second example was where DB sacrificed a Rook for a pawn
next to Rebel's King. After the exchange, Hsu reported, Rebel showed
2+ pawns advantage. DB showed a .5 pawn advantage. A couple moves
later, DB went up to a much higher advantage, and Rebel still showed
+2. After a few more moves, Rebel suddenly realized it was busted, and
dropped its eval way down. Hsu thought this was due to a minimal King
safety evaluation. He did state that even with this, he thought Rebel
had a much better understanding of positional play than Genius did. I
asked him if it were possible to get scores of these games. He said he
did not want to release them, as he did not want to give out any help
to future competitors. I mentioned that he had said the chance of Deep
Blue ever giving another match were almost nil, and so there should
not be any future competitors. He responded that if he got the rights
to the chess processors, Rebel and Genius would likely be the future
competitors, and he wanted to leave his options open. I stated that
even so, once released, there would be thousands of games available
rather quickly, and that these 10 would not make much difference. He
said that he wasn't even sure if the game scores had been saved. I
realized that he was not going to let them out, so I suggested that if
he found them, not to erase them, as there were a lot of people
interested in them, and I moved on.
10. Since we had been talking about evaluations and positional
understanding, I took the time at this point to bring up my current
favorite among the chess programs I have, Chess System Tal. I stated
that I was impressed with the amount it accomplished within 3000 nodes
per second searched. I said that if its evaluation was able to search
at a much higher rate, that I thought it would be much better than the
other micro programs. I was surprised by the enthusiasm Hsu showed
about the program. He mentioned that how it handled King Safety was
much more similar to Deep Blue than the other micro programs, although
perhaps a little bit more extreme, and that he, too, was impressed
with it. He said that many things in CSTal were implemented in Deep
Blue, which I found strange, since it wasn't released until after the
match. I didn't think of that until later, so I wasn't able to ask him
about this. Perhaps he meant implemented similarly, or perhaps ideas
from CSTal's style of play exhibited in games Thorsten posted. Or
maybe he had access to a beta version, or he was referring to Complete
chess system, I don't really know. I only mention this as it gives a
little insight into the approach used in the evaluation. During the
presentation Hsu stated that unlike other chess programs, DB's
evaluation in not just a matter of adding weights together with
bonuses to arrive at a score, but some functions were calculated
non-linearly, through multiplication, or other "second level" methods.
When asked about this, he said that some examples of the non-linear
evaluations were in the method of calcualating a pawn's value based on
it's advance, and its position relative to other pieces and pawns, and
King safety, which was an example of what he had referred to as
"second level" methods. This was a question and answer section at the
end of the presentation, and since it wasn't my question, I could not
ask him to expand on these generalisms.
11. There has been some question as to the endgame databases used
during the match. Hsu stated that there were 20 gigabytes of endgame
databases from Ken Thompson and Lewis Stiller on the hard drive. He
said that they were all of the five-man and down, plus selected
six-man endgame databases. To his knowledge, during the match, they
were never accessed, but he was not sure of this. He said that since
the chess processors have some of the engame databases built in (I
have read that these are the 3-man set), he figured that it never got
to the point where the SP2s would need to access the hard-disk-based
databases. He said that it was probably a good psychological weapon
for Kasparov to know that they were there, since, if he made one wrong
move during the endgame, he would know that he would quickly look
foolish in front of millions of people, and this would have to have an
effect. Other differing reports about how many processors DB used were
also answered. Deep Blue employed 30 SP2 Scalable Processors. The
frames were capable of holding sixteen each, and there were two
frames, but in each frame, two processors were tied together to form a
master processor, which meant a total of 30 instead of thirty-two.
Each SP2 had 16 chess processors attached, so that meant a total of
480 chess processors. Up until this point I had only heard 256 or 512.
Hsu said that Deep Blue used "two-level parallelism" to process
positions. He described this as the method of the master processor
evaluating the first 4 moves, then sending the 1000 or so positions
involved to the other SP2s, which would carry it out for another 4
moves, and then turn over the positions to the chess processors, which
would go on for 4-5 more moves. He said that on average DB would reach
to 30 ply in considering a move, but in certain cases it had reached,
through selective extensions and pruning up to 70 ply, though this was
rare. It would on average process 200 million chess positions per
second, but that this reached as high as 400 million in certain cases.
The chess processors made for the rematch were capable of 2-2.5
million nodes per second processed, and with improved evaluation with
Joel Benjamin's help, and better selective search, the speed was
improved by 3-10 times over the 1996 version. I asked how many cycles
it took to evaluate a position, and was told that it varied. There was
a short evaluation used approximately 80% of the time which took only
one cycle, and there was a long evaluation used 20% of the time that
took 8 cycles. Move generation took 4 cycles. There were 8000
adjustable evaluation features, and these included such things as the
value of a rook on an unopened file which could later be forced open
with a pawn exchange or sacrifice. He said this was one that was added
through the help of GM Joel Benjamin, and he knew of one instance
during the match when it had an effect. (I have not looked over the
games to see where this would be, perhaps some helpful reader with
more time could find this out.) It would be very interesting to know
how these evaluations can be performed in hardware, but I am not sure
that this will ever be covered, especially if Hsu is really thinking
of a commercial version of the program. Since he also mentioned that
he would be interested to see if a single-chip chess machine could be
created to beat the world champion someday, he may not be forthcoming
on his research, as would be hoped.
12. Hsu evidently had difficulty in convincing the rest of the team to
switch to a redesigned chess processor between the match and rematch.
Since the lead time for a chess processor was normally a year for
design, testing, and debug, and since they only had 1 year and three
months until the rematch, they were more interested in tweaking the
program in the SPs, and leaving the chips alone. Hsu said he worked
for 6 months, 70-100 hours per week, redesigning the chess processors.
When he had them ready, and began the tests to see how well they
performed relative to the older chips, the difference was so great
that the rest of the team quickly agreed to switch to the new
processors and so continued on from there.
Well, there was more, but this concludes the report for this forum.
Most of the rest is anecdotal, and not so informative, so I will stop
here. I hope it was interesting.
-----------------------------------------------------------------
kp
Komputer Korner
If they were looking 30 ply ahead and if an opening was 30 ply then
the first 60 ply /30 moves would be evaluated completely. Not too far
down the line (especially considering 70 plies of extensions) where
chess would essentially be solved if Deep Blue ever made a comeback.
I consider chess essentially solved once you can reach 120 ply from
the first move (the opening book counts for part of the 120 ply). The
reason for this is almost all games are decided by 60 moves or if they
aren't decided by that time, the result is clear. 30 ply lookahead
is far greater than anybody thought. Note for you purists
"essentially solved" doesn't mean solved.
--
--
Komputer Korner
The inkompetent komputer
To send email take the 1 out of my address. My email address is
kor...@netcom.ca but take the 1 out before sending the email.
Keith Ian Price wrote in message ...
Chris Whittington
Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
<ihDRG0dUkHpV-p...@pdx03-pm2-02.teleport.com>...
> Sorry for the delay. Here is Part III of my Deep Blue Report:
>
< snipped some bits >
> 10. Since we had been talking about evaluations and positional
> understanding, I took the time at this point to bring up my current
> favorite among the chess programs I have, Chess System Tal. I stated
> that I was impressed with the amount it accomplished within 3000 nodes
> per second searched. I said that if its evaluation was able to search
> at a much higher rate, that I thought it would be much better than the
> other micro programs. I was surprised by the enthusiasm Hsu showed
> about the program.
Likewise :)
> He mentioned that how it handled King Safety was
> much more similar to Deep Blue than the other micro programs, although
> perhaps a little bit more extreme, and that he, too, was impressed
> with it.
Hmmm. Not sure I believe this. If DB had king-safety terms like CSTal has
king safety terms, it would have certainly seen the draw possibilities in
the famous missed draw game - very early.
> He said that many things in CSTal were implemented in Deep
> Blue, which I found strange, since it wasn't released until after the
> match. I didn't think of that until later, so I wasn't able to ask him
> about this. Perhaps he meant implemented similarly, or perhaps ideas
> from CSTal's style of play exhibited in games Thorsten posted. Or
> maybe he had access to a beta version, or he was referring to Complete
> chess system, I don't really know.
He could mean almost anything. But one thing I know for sure, that many
early, far earlier than pre-beta, versions of CSTal were doing the rounds
of freaks and programming teams. Despite non-disclosure-agreements,
personal assurances, and conditions of general morality; the level of
dishonest behaviour, right at the heart, and right amongst some of the
loudest copy-protectors of their own software, is really quite high.
Mostly however, and amusingly, these persons didn't recognise what was
staring them in the face.
I guess it is quite possible that somebody sent a version to Hsu. I
recollect at the time, that I was part of a vociferous campaign that DB
didn't stand a chance in hell since it was a materialistic bean-counter in
the Hyatt-Crafty mould.
However, Hsu's GMs wouldn't have needed CSTal to tell him that what was
lacking in the DB program, as in any materialistic-Hyatt style program, was
dynamic stuff to deal with that which the
quiescent-seeking-Hyattian-nonsense tried to ignore. Namely positions
leading to mating attacks. Duh.
> I only mention this as it gives a
> little insight into the approach used in the evaluation. During the
> presentation Hsu stated that unlike other chess programs, DB's
> evaluation in not just a matter of adding weights together with
> bonuses to arrive at a score, but some functions were calculated
> non-linearly, through multiplication, or other "second level" methods.
This is repeated again and again, in all kind of situations, and is just
woffle.
I guess if your program has moved beyond the stage of adding up 32 piece
square table entries, then you might want to start crowing about
'non-linear second-level differential equations packed with Greek letters,
and things to the power of'. It might even do as a ICCAJ article title. But
meaningless nevertheless.
> When asked about this, he said that some examples of the non-linear
> evaluations were in the method of calcualating a pawn's value based on
> it's advance, and its position relative to other pieces and pawns, and
> King safety, which was an example of what he had referred to as
> "second level" methods.
I'ld call it 'dynamic'. ie. evaluating with respect to what the pieces can
do in combination with each other. Very powerful.
The Hyattian quiescence paradigm can't do this. This can't evaluate
anything to do with chess, especially not those 'second-level' things. So
they try and extend to a 'quiet' position (defined as one with no
captures), so that they can then basically just add up the material, and
not worry about any dynamic features (because, by their definition, there
aren't any when quiet).
See the big hole in the logic ? They can't evaluate any dynamic king stuff.
So they don't know whether their 'quiet' position is quiet or not with
respect to the king. They just ignore it :)
The result is a materialistic bean-counter. Which, as Thorsten has shown in
many published test games, has no idea when it is on receiving end of a
vicious king attack. Or not until too late.
> This was a question and answer section at the
> end of the presentation, and since it wasn't my question, I could not
> ask him to expand on these generalisms.
>
Well, that leave me to expand, then ...... :)
1. As has been repeatedly said, the bean-counters will get nowhere until
they abandon their quiesence paradigm.
2. The reason the bean-counting programmers use the quiesence paradigm is
because they know zilch about chess. Fach-idiots. Experts at programming.
Stupid at chess.
Anyroads, if the GM's chess knowledge + Hsu's programming/hardware skills
were put together into DB, effectively, then no GM or super-GM would be
able to withstand.
My only doubt with regard to DB remains with the quiesence paradigm, which
must have been at the heart of the program. Did they throw the paradigm out
? Almost impossible in one year.
Did they implement dynamic king-safety/attack stuff and incorporate into
the evaluation ? Quite possible. But one of the features of king attack
situations is their unresolvability within any reasonable time frame. You
can exhaust captures fairly quickly, and then evaluate; but exhausting
checks and mate threat attacks ? No way.
Chris Whittington
mclane
kpr...@spamfree.teleport.com (Keith Ian Price) wrote:
>Sorry for the delay. Here is Part III of my Deep Blue Report:
First let me say thanks for your reports.
I have attacked this machine so often and projected my prejudices
towards it, that I am always glad to get information that is not
arrogant and snobistic.
When I was a teen I dived deep into computerchess. I read Opfermann's
strange books about the chess-players all over the world (his style is
very associative, he build links between chess and other related
topics), followed each of David Levy's games with my Mephisto II and
ate any PIONIER article about AI and Botwinnik.
In germany Helmut-Pfleger and Vlastimil Hort commented TV-world-cups.
I saw Seiravan, very smart clothes and I am lucky that he has not
developed negatively since now. I read Frederic's globetrotter
articles about: Fred visits chess-programmers all over the world, says
hello to Dave and Kathe and and and.
In my dreams when I went to sleep in nights dedicated chess computers
and Super-Constellations and names like HITECH and Berliner circeled
arround.
This was like drugs.
We thought: ONE day the AI machine will be developed. ONE day. This
was my dream. I was very passive, only a consumer. Not able to change
or help or do anything. I was forced to read the magazines, and that
was all.
Now I am 31 and many things have changed.
When I read about Deep Blue and Kasparov and their "match" i thought
materialist fight in an advertising-campaign. He gets money, gives his
name, and they pay and get the games and the fame, if they have a
change. This was all very materialistic to me.
After I saw game 2 anything was clear to me. This match is not a real
match, it is manipulated. This is not the right way to play against a
machine, like he has done in the first game.
Later they attacked each other, Kasparov, always in bad mood when he
loses, and Deep Blue, developed into a nice to misuse myth.
The enormous NPS rates made it impossible to me to see any kind of
INTELLIGENCE in Deep Blue. It looked more like a gigantic calculator,
moving chess-moves, but not playing it.
An emulation, and a fast one. But a strong one ?
I always doubted this. Also I always doubted these 10-0 scores.
Your report gives new material to me.
I have more the feeling that some people are materialists because they
refuted or failed to stand idealism. And I can live with this much
better, since I understand that not anybody has the chance to hold his
ideals until growing old.
> However, Hsu continued, he
>did not think that this constituted intelligence.
I agree completely.
The main strength of intelligence happens when you go to sleep. You
forget the unimportant stuff. You dream strange dreams. How can a
computer ever FORGET the unimportant stuff ??? No - I have seen less
chess programs DREAM so far.
And the few I saw, I tried to work with... :-)))
So that MY dreams come through !
> He said he
>did not want to release them, as he did not want to give out any help
>to future competitors. I mentioned that he had said the chance of Deep
>Blue ever giving another match were almost nil, and so there should
>not be any future competitors. He responded that if he got the rights
>to the chess processors, Rebel and Genius would likely be the future
>competitors, and he wanted to leave his options open. I stated that
>even so, once released, there would be thousands of games available
>rather quickly, and that these 10 would not make much difference. He
>said that he wasn't even sure if the game scores had been saved. I
>realized that he was not going to let them out, so I suggested that if
>he found them, not to erase them, as there were a lot of people
>interested in them, and I moved on.
I don't believe in the above passage. When I read the above it sound
very suspicious to me. I guess somebody is cheating here.
Not saving the games, although: not want to give them in the hand of
competitors, although they would have no chance. In the above
paragraph we have so many contradicting conclusions that I do better
not comment on it, since my language would not be polite.
>10. Since we had been talking about evaluations and positional
>understanding, I took the time at this point to bring up my current
>favorite among the chess programs I have, Chess System Tal.
You did this really ??!!
Amazing.
> I stated
>that I was impressed with the amount it accomplished within 3000 nodes
>per second searched.
on a 200 Mhz machine, yes.
>I said that if its evaluation was able to search
>at a much higher rate, that I thought it would be much better than the
>other micro programs.
Could be...
> I was surprised by the enthusiasm Hsu showed
>about the program.
I am surprised to read this too. Is this materialistic-HSU maybe a
resigned idealist ?
If somebody gives up his dreams, or has to give up them, he is maybe
so depressed that he turnes arround and becomes a materialist.
> He mentioned that how it handled King Safety was
>much more similar to Deep Blue than the other micro programs, although
>perhaps a little bit more extreme, and that he, too, was impressed
>with it.
!!! :-))))
Although we never designed it for Deep-Blue engineers... :-)))
> He said that many things in CSTal were implemented in Deep
>Blue, which I found strange, since it wasn't released until after the
>match.
You forgot that we worked for arround 6 years on it and that I
published many games...
>I didn't think of that until later, so I wasn't able to ask him
>about this. Perhaps he meant implemented similarly,
I guess I know what he means.... :-))
> or perhaps ideas
>from CSTal's style of play exhibited in games Thorsten posted.
How shall he know the games I post ? I don't think people like him
have time to read rgcc ?!
>Or
>maybe he had access to a beta version,
Clear No !
>or he was referring to Complete
>chess system, I don't really know.
I don't think you can be enthusiastic by relating to Complete Chess
system, or ?
:-)
> I only mention this as it gives a
>little insight into the approach used in the evaluation. During the
>presentation Hsu stated that unlike other chess programs, DB's
>evaluation in not just a matter of adding weights together with
>bonuses to arrive at a score, but some functions were calculated
>non-linearly, through multiplication, or other "second level" methods.
>When asked about this, he said that some examples of the non-linear
>evaluations were in the method of calcualating a pawn's value based on
>it's advance, and its position relative to other pieces and pawns, and
>King safety, which was an example of what he had referred to as
>"second level" methods. This was a question and answer section at the
>end of the presentation, and since it wasn't my question, I could not
>ask him to expand on these generalisms.
Understand.
Thanks for your little report. I laughed much about this. I am sure we
will kill Deep Blue one day, no matter how many NPS it makes. :-))
The slowest beats the fastest ? My dreams come through ?!
> I hope it was interesting.
>-----------------------------------------------------------------
>kp
Thanks !
best wishes
mclane
Robert Hyatt
Keith Ian Price <kpr...@spamfree.teleport.com> wrote:
: Sorry for the delay. Here is Part III of my Deep Blue Report:
: 9. One of the longest-running arguments on rgcc and CCC has been how
: well micros might fare against Deep Blue. During the Deep Blue
: excitement last year the news slipped out that there had been a match
<snip>
This is basically a replay of what I gave here last year... the details
sound exactly the same. I was more careful about what Hsu thought about
the micros, but since he's now being "open" I think that's progress...
: 10. Since we had been talking about evaluations and positional
: understanding, I took the time at this point to bring up my current
: favorite among the chess programs I have, Chess System Tal. I stated
: that I was impressed with the amount it accomplished within 3000 nodes
: per second searched. I said that if its evaluation was able to search
: at a much higher rate, that I thought it would be much better than the
: other micro programs. I was surprised by the enthusiasm Hsu showed
: about the program. He mentioned that how it handled King Safety was
: much more similar to Deep Blue than the other micro programs, although
: perhaps a little bit more extreme, and that he, too, was impressed
: with it. He said that many things in CSTal were implemented in Deep
: Blue, which I found strange, since it wasn't released until after the
: match. I didn't think of that until later, so I wasn't able to ask him
: about this. Perhaps he meant implemented similarly, or perhaps ideas
: from CSTal's style of play exhibited in games Thorsten posted. Or
I think you misinterpreted this. He most likely meant that "CSTal does
a lot of things like we do them"... *not* that he copied CSTal ideas,
since there are no details of what CSTal is doing that are available.
As I had mentioned, they are able to do a bunch of things that the rest of
us can't... attack analysis where you look at which pieces bear on the
squares around the king, which pieces are in "battery" with those pieces
directly bearing on the king, and so forth. Things that are expensive for
us are "el cheap" for them, thanks to being done in hardware. If you look
at a bunch of programs, you probably find that they can easily "have the
eval of CSTal's king safety, the positional knowledge of Hiarcs and all the
rest, the search extensions of a Cray Blitz, and so forth... and they can
do it at 250M nodes per second to boot..."
: maybe he had access to a beta version, or he was referring to Complete
: chess system, I don't really know. I only mention this as it gives a
: little insight into the approach used in the evaluation. During the
: presentation Hsu stated that unlike other chess programs, DB's
: evaluation in not just a matter of adding weights together with
: bonuses to arrive at a score, but some functions were calculated
: non-linearly, through multiplication, or other "second level" methods.
: When asked about this, he said that some examples of the non-linear
: evaluations were in the method of calcualating a pawn's value based on
: it's advance, and its position relative to other pieces and pawns, and
: King safety, which was an example of what he had referred to as
: "second level" methods. This was a question and answer section at the
: end of the presentation, and since it wasn't my question, I could not
: ask him to expand on these generalisms.
<snip>
: how these evaluations can be performed in hardware, but I am not sure
: that this will ever be covered, especially if Hsu is really thinking
: of a commercial version of the program. Since he also mentioned that
: he would be interested to see if a single-chip chess machine could be
: created to beat the world champion someday, he may not be forthcoming
: on his research, as would be hoped.
How they are done in hardware is easy... just like they are done in software,
only orders of magnitude faster. :) But for a better answer, look at Crafty's
evaluation. Imagine every And, Or, Xor, shift, comparison, etc replaced by a
simple silicon gate... and you get the idea... And then imagine each "feature"
evaluated in parallel by multiple sets of these "gates" that get their results
added together in a "tree" style adder later on...
: 12. Hsu evidently had difficulty in convincing the rest of the team to
: switch to a redesigned chess processor between the match and rematch.
: Since the lead time for a chess processor was normally a year for
: design, testing, and debug, and since they only had 1 year and three
: months until the rematch, they were more interested in tweaking the
: program in the SPs, and leaving the chips alone. Hsu said he worked
: for 6 months, 70-100 hours per week, redesigning the chess processors.
: When he had them ready, and began the tests to see how well they
: performed relative to the older chips, the difference was so great
: that the rest of the team quickly agreed to switch to the new
: processors and so continued on from there.
He had told me this as well. Until hearing this from him, I was not aware
that DBII was different from DB, other than more processors. He told me
that they needed different evaluation hardware to take some of the suggestions
from Joel and implement them.
: Well, there was more, but this concludes the report for this forum.
: Most of the rest is anecdotal, and not so informative, so I will stop
: here. I hope it was interesting.
: -----------------------------------------------------------------
: kp
yep... good work...
--
Robert Hyatt Computer and Information Sciences
hy...@cis.uab.edu University of Alabama at Birmingham
(205) 934-2213 115A Campbell Hall, UAB Station
(205) 934-5473 FAX Birmingham, AL 35294-1170
Robert Hyatt
Komputer Korner <kor...@netcom.ca> wrote:
: If they were looking 30 ply ahead and if an opening was 30 ply then
: the first 60 ply /30 moves would be evaluated completely. Not too far
: down the line (especially considering 70 plies of extensions) where
: chess would essentially be solved if Deep Blue ever made a comeback.
: I consider chess essentially solved once you can reach 120 ply from
: the first move (the opening book counts for part of the 120 ply). The
: reason for this is almost all games are decided by 60 moves or if they
: aren't decided by that time, the result is clear. 30 ply lookahead
: is far greater than anybody thought. Note for you purists
: "essentially solved" doesn't mean solved.
"30 plies" means 30 plies *selectively*... *not* full-width. They do
this by extending what looks interesting, not extending what doesn't.
There's no guarantee that what they think is uninteresting really is,
so "solved" is way too big a stretch..
Robert Hyatt
Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
: Hmmm. Not sure I believe this. If DB had king-safety terms like CSTal has
: king safety terms, it would have certainly seen the draw possibilities in
: the famous missed draw game - very early.
Not necessarily. In fact, in game 6, DB was at +3 on the first non-book
move, according to the operator. It was a *long* time before it really
was able to win anything.. Hsu had said their king safety terms were
*very* significant. This simply verifies that...
: However, Hsu's GMs wouldn't have needed CSTal to tell him that what was
: lacking in the DB program, as in any materialistic-Hyatt style program, was
: dynamic stuff to deal with that which the
: quiescent-seeking-Hyattian-nonsense tried to ignore. Namely positions
: leading to mating attacks. Duh.
You keep saying that. How do you do against this "nonsense"? IE results
count, elegance doesn't.. there is an *easy* way to measure results...
:> I only mention this as it gives a
Interesting, since I am already doing some of it. You want to harp on
"quiescence".. I want to harp on "basic search". DB's q-search is also
quite simple. And quite effective, it would seem. Ferret and I are doing
ok as well with our "simple qsearch"...
: See the big hole in the logic ? They can't evaluate any dynamic king stuff.
: So they don't know whether their 'quiet' position is quiet or not with
: respect to the king. They just ignore it :)
This is utter nonsense. We simply evaluate it a little earlier and choose
whether to proceed, take care of piece exchanges, or just evaluate. I see
nothing that says a simple qsearch + a clever basic search can not play
strong chess. I have ample evidence that says otherwise, in fact..
: The result is a materialistic bean-counter. Which, as Thorsten has shown in
: many published test games, has no idea when it is on receiving end of a
: vicious king attack. Or not until too late.
And the over-evaluated speculative program generally comes out on the short
end of the stick in these speculative attacks, according to Thorsten, also?
So you win a game by smoke and mirrors, then lose 5 by the same methodology,
and this methodology is "superior"??
:> This was a question and answer section at the
:> end of the presentation, and since it wasn't my question, I could not
:> ask him to expand on these generalisms.
:>
: Well, that leave me to expand, then ...... :)
: 1. As has been repeatedly said, the bean-counters will get nowhere until
: they abandon their quiesence paradigm.
Nowhere except to the "top". IE witness deep blue. capture-only q-search
although they do recognize that one side can be mated by a capture, which I
don't currently do..
: 2. The reason the bean-counting programmers use the quiesence paradigm is
: because they know zilch about chess. Fach-idiots. Experts at programming.
: Stupid at chess.
Now we break out the broad brush and paint with eyes closed?
: Anyroads, if the GM's chess knowledge + Hsu's programming/hardware skills
: were put together into DB, effectively, then no GM or super-GM would be
: able to withstand.
: My only doubt with regard to DB remains with the quiesence paradigm, which
: must have been at the heart of the program. Did they throw the paradigm out
: ? Almost impossible in one year.
Nope, not at all. They extend their stuff *before* reaching the q-search.
That's the whole point.. and it *does* work. I can point you to a couple
of programs that are quite "stiff", even with a simple qsearch. No law I
have seen says the "qsearch" has to find mating attacks. *I* don't search
that way. Humans don't search that way...
: Did they implement dynamic king-safety/attack stuff and incorporate into
: the evaluation ? Quite possible. But one of the features of king attack
: situations is their unresolvability within any reasonable time frame. You
: can exhaust captures fairly quickly, and then evaluate; but exhausting
: checks and mate threat attacks ? No way.
not just "possible" but "actually done that way". Hsu was clear here...
it is all *evaluation*... although before quiescence is reached, they have
extensions that drive the search deeper. IE a large king safety term will
definitely trigger singular extensions to follow that line deeper, which is
what is needed...
: Chris Whittington
Chris Whittington
Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
<6j4enh$2mr$1...@juniper.cis.uab.edu>...
> Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
>
>
> : Hmmm. Not sure I believe this. If DB had king-safety terms like CSTal
has
> : king safety terms, it would have certainly seen the draw possibilities
in
> : the famous missed draw game - very early.
>
>
> Not necessarily. In fact, in game 6, DB was at +3 on the first non-book
> move, according to the operator. It was a *long* time before it really
> was able to win anything.. Hsu had said their king safety terms were
> *very* significant. This simply verifies that...
Bob Hyatt also wrote elsewhere:
(2) There is another option. For this to work, *every* person here will
have
to subscribe to this plan. If you are interested, read on:
(a) immediately add rolf, chris and anonymous to your kill files.
(b) after doing this, *absolutely* do not respond to anything they write,
for *any* reason, on *any* topic. Simply behave as if they don't
exist,
except for posts that are "on topic" and reasonably "polite" and also
"to the point." This is the critical step. They must get *no*
responses
no matter *what* they say to provoke. Even one reply will keep the
problem alive.
And then he didn't follow his own advice ......
Now, Mr Hyatt, if you're going to call on the massed ranks of the
petit-bourgoisie to assemble for the counter-revolution, marching behind
the banner "Crafty has a king-safety function, honest"; then it would be as
well if you maintained your promise to give me the great honour and
privilege of being in your kill-file.
Thank you.
Chris Whittington
Don Getkey
In article <6j3o1d$34m$1...@tor-nn1.netcom.ca>, "Komputer Korner"
<kor...@netcom.ca> writes:
>almost all games are decided by 60 moves or if they
>aren't decided by that time, the result is clear. 30 ply lookahead
>is far greater than anybody thought. Note for you purists
>"essentially solved" doesn't mean solved.
I'll even give you "almost all games are solved by 40 moves", but even at this
lower threshhold few would insert or associate the word "solved", or even
"essentailly solved", to describe THE "chess solution". The relative numerical
abundance of possible moves in the middle game is well beyond our comprehensive
grasp, and theirs (i.e.computers).
yours in chess,
Don
Coon Rapids MN USA
Robert Hyatt
Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
: Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
: <6j4enh$2mr$1...@juniper.cis.uab.edu>...
:> Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
:>
: (b) after doing this, *absolutely* do not respond to anything they write,
: for *any* reason, on *any* topic. Simply behave as if they don't
: exist,
: except for posts that are "on topic" and reasonably "polite" and also
: "to the point." This is the critical step. They must get *no*
: responses
: no matter *what* they say to provoke. Even one reply will keep the
: problem alive.
: And then he didn't follow his own advice ......
: Now, Mr Hyatt, if you're going to call on the massed ranks of the
: petit-bourgoisie to assemble for the counter-revolution, marching behind
: the banner "Crafty has a king-safety function, honest"; then it would be as
: well if you maintained your promise to give me the great honour and
: privilege of being in your kill-file.
Why don't you stop acting like a jerk, and *carefully* read (b) above.
Let me give you a hint:
"except for posts that are "on topic" and reasonably "polite" and
also "to the point"...
Get it now? Maybe? Your post was reasonable, was about computer chess,
and I responded. If you'd simply read first, then knee-jerk, your knee
would not be needed nearly as often. That post was reasonably written.
I don't see the need for the personal type "Hyattism" since there are plenty
of other names that could be substituted for "Hyatt" (IE moreland, Heinz,
Slate, Thompson, Hsu, Campbell, etc...)
But I *did* follow my own advice, thank you very much. Now if you would do
the same... we *might* have interesting discussions here. It's up to you...
bruce moreland
>Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
>> He said that many things in CSTal were implemented in Deep
>> Blue, which I found strange, since it wasn't released until after the
>> match. I didn't think of that until later, so I wasn't able to ask him
>> about this. Perhaps he meant implemented similarly, or perhaps ideas
>> from CSTal's style of play exhibited in games Thorsten posted. Or
>> maybe he had access to a beta version, or he was referring to Complete
>> chess system, I don't really know.
>
>He could mean almost anything. But one thing I know for sure, that many
>early, far earlier than pre-beta, versions of CSTal were doing the rounds
>of freaks and programming teams. Despite non-disclosure-agreements,
>personal assurances, and conditions of general morality; the level of
>dishonest behaviour, right at the heart, and right amongst some of the
>loudest copy-protectors of their own software, is really quite high.
Either that or he could have a vague clue about what you do, based
upon what he's read here, or what the questioner told him, and
answered that he does some of the same stuff.
It's not like anyone really knows what you do.
The rest of your post is mostly about hatred of other people,
specifically Bob.
You've stated in the past that you are willing to evaluate
non-quiescent positions. I don't see this as being very important, it
sounds like a speed optimization, since you don't have to recurse and
call eval again after you find a promising capture, and this is all
happening way out at the tips anyway.
Do you also direct the search in the higher nodes? If so, that seems
more likely to resolve attacks.
bruce
bruce moreland
>Now, Mr Hyatt, if you're going to call on the massed ranks of the
>petit-bourgoisie to assemble for the counter-revolution, marching behind
>the banner "Crafty has a king-safety function, honest"; then it would be as
>well if you maintained your promise to give me the great honour and
>privilege of being in your kill-file.
I would love to see another match between CST and Crafty.
bruce
Keith Ian Price
On Sun, 10 May 1998 14:34:42, Robert Hyatt <hy...@crafty.cis.uab.edu>
wrote:
> Keith Ian Price <kpr...@spamfree.teleport.com> wrote:
>: He mentioned that how it handled King Safety was
> : much more similar to Deep Blue than the other micro programs, although
> : perhaps a little bit more extreme, and that he, too, was impressed
> : with it. He said that many things in CSTal were implemented in Deep
> : Blue, which I found strange, since it wasn't released until after the
> : match. I didn't think of that until later, so I wasn't able to ask him
> : about this. Perhaps he meant implemented similarly, or perhaps ideas
> : from CSTal's style of play exhibited in games Thorsten posted.
> I think you misinterpreted this. He most likely meant that "CSTal does
> a lot of things like we do them"... *not* that he copied CSTal ideas,
> since there are no details of what CSTal is doing that are available.
That may well be. I thought long and hard afterwards about what he
said. His words were as described but at the time I felt he meant
"implemented similarly" and separately. But since he had redone the
chess chips in the interim between the two matches, and since he had
beefed up the king safety in that revision, and since that is what he
said, I felt I should put it in, not just the way I originally
accepted it. I would suspect that you are right, though. I don't
think he meant that he "copied" anything.
> As I had mentioned, they are able to do a bunch of things that the rest of
> us can't... attack analysis where you look at which pieces bear on the
> squares around the king, which pieces are in "battery" with those pieces
> directly bearing on the king, and so forth. Things that are expensive for
> us are "el cheap" for them, thanks to being done in hardware. If you look
> at a bunch of programs, you probably find that they can easily "have the
> eval of CSTal's king safety, the positional knowledge of Hiarcs and all the
> rest, the search extensions of a Cray Blitz, and so forth... and they can
> do it at 250M nodes per second to boot..."
I'd settle for a one-chip PC model at 2--2.5 million nodes per second.
Even better a 4-chip "Pro" model. Heck, I may just call an IBM
salesman, and ask how much would an order of 100 DB, Jrs. cost. If
there was interest, IBM might decide to sell it as an application for
their SP2 scalable processors. I could probably find 90 businesses who
would like to add one to their web site to attract the public to their
site, and mark them all up 11% in order to keep 10 for free. Hmmm.
Maybe I'll go over to Intel tomorrow. Would UAB need one for their
Computer Science Dept.? 8-)
> <snip>
> : how these evaluations can be performed in hardware, but I am not sure
> : that this will ever be covered, especially if Hsu is really thinking
> : of a commercial version of the program. Since he also mentioned that
> : he would be interested to see if a single-chip chess machine could be
> : created to beat the world champion someday, he may not be forthcoming
> : on his research, as would be hoped.
>
> How they are done in hardware is easy... just like they are done in software,
> only orders of magnitude faster. :) But for a better answer, look at Crafty's
> evaluation. Imagine every And, Or, Xor, shift, comparison, etc replaced by a
> simple silicon gate... and you get the idea... And then imagine each "feature"
> evaluated in parallel by multiple sets of these "gates" that get their results
> added together in a "tree" style adder later on...
I can understand that, but the evaluations he described were more
complex than "Rook on open file = +.3", and there are 8000 features
taking up approx 1/3 of the chip. The chip is a .6 micron design, so
it doesn't pack 10 million transistors on it, so how he implemented
the evaluations in hardware would be of interest to me. Also, exactly
what is done in evaluation would also be of interest. If he is
touring, perhaps you really could get him to come to UAB, and delve
deeper.
-----------------------------------------------------------------
kp
Robert Hyatt
bruce moreland <bru...@seanet.com> wrote:
: On Sun, 10 May 1998 10:07:58 GMT, "Chris Whittington"
: <chr...@cpsoft.demon.co.uk> wrote:
:>Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
:>> He said that many things in CSTal were implemented in Deep
:>> Blue, which I found strange, since it wasn't released until after the
:>> match. I didn't think of that until later, so I wasn't able to ask him
:>> about this. Perhaps he meant implemented similarly, or perhaps ideas
:>> from CSTal's style of play exhibited in games Thorsten posted. Or
:>> maybe he had access to a beta version, or he was referring to Complete
:>> chess system, I don't really know.
:>
:>He could mean almost anything. But one thing I know for sure, that many
:>early, far earlier than pre-beta, versions of CSTal were doing the rounds
:>of freaks and programming teams. Despite non-disclosure-agreements,
:>personal assurances, and conditions of general morality; the level of
:>dishonest behaviour, right at the heart, and right amongst some of the
:>loudest copy-protectors of their own software, is really quite high.
: Either that or he could have a vague clue about what you do, based
: upon what he's read here, or what the questioner told him, and
: answered that he does some of the same stuff.
: It's not like anyone really knows what you do.
: The rest of your post is mostly about hatred of other people,
: specifically Bob.
: You've stated in the past that you are willing to evaluate
: non-quiescent positions. I don't see this as being very important, it
: sounds like a speed optimization, since you don't have to recurse and
: call eval again after you find a promising capture, and this is all
: happening way out at the tips anyway.
: Do you also direct the search in the higher nodes? If so, that seems
: more likely to resolve attacks.
: bruce
It is just the typical nonsense about "stereotypes" that Chris just can't
get beyond. As I have said a hundred times here, there is *nothing* in the
design of Crafty that precludes intelligence. It is about speed. But it is
slower. In fact, it is almost 1/2 the speed of Rebel now, where 2 years ago
it was faster. I can guarantee you that I haven't intentionally made the
code less efficient. Just that I have added to the eval, and added to the
search extensions. And I'm continuing to do that.
Chris believes that you have to go all the way to the other extreme, and
do *all* the unsound analysis that can be imagined. And this also leads to
great difficulties. And, slowly, the "junk" gets deleted from his program,
while things get "added" to mine (or yours) and we are still closing in on
the same goal, whether we are willing to admit it or not...
But, at the same time, success can certainly be measured. And it is not
measured by the "quality" of the chess a program plays, it is measured by
counting the number of 1's and comparing that to the number of 0's in the
result column. Based on that, I am still satisfied with the Crafty/Ferret/
Chess 4.x/Belle/Hitech/Deep Blue/fritz/nimzo/etc approach to chess.
to paraphrase a frequently used quote: "results talk, bullcrap walks."
Everyone I know is always on the lookout for something that improves the
results of their program. Taking pot-shots at *other* programs/authors is
*not* a way to improve results, however.
But since I know where I stand, result-wise, in relation to CST, I'm not
unhappy with what I've done... Lonnie could enlighten everyone about the
results he got...
Against many other programs...
Bob
mclane
bru...@seanet.com (bruce moreland) wrote:
>I would love to see another match between CST and Crafty.
>bruce
If you (or Moritz) tell me how to install crafty accurate (when I
tried it 1/2 a year ago, or even earlier, it was complicate. I had to
do this, take care of that, and and and).
But i will do anything you tell me what is needed !
Tell me what to do and I can do 40/120.
We can bet about the result :-)))
best wishes
mclane
Chris Whittington
bruce moreland <bru...@seanet.com> wrote in article
<3556dd7a....@news.seanet.com>...
> On Sun, 10 May 1998 15:46:34 GMT, "Chris Whittington"
> <chr...@cpsoft.demon.co.uk> wrote:
>
> >Now, Mr Hyatt, if you're going to call on the massed ranks of the
> >petit-bourgoisie to assemble for the counter-revolution, marching behind
> >the banner "Crafty has a king-safety function, honest"; then it would be
as
> >well if you maintained your promise to give me the great honour and
> >privilege of being in your kill-file.
>
> I would love to see another match between CST and Crafty.
Binaries at dawn, eh Bruce ?
I suppose if Crafty wins then it's allowed to dismember drug dealers by
pulling them apart with buses ?
Chris Whittington
>
> bruce
>
>
Chris Whittington
bruce moreland <bru...@seanet.com> wrote in article
<3555d8bb....@news.seanet.com>...
> On Sun, 10 May 1998 10:07:58 GMT, "Chris Whittington"
> <chr...@cpsoft.demon.co.uk> wrote:
>
> >Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
>
> >> He said that many things in CSTal were implemented in Deep
> >> Blue, which I found strange, since it wasn't released until after the
> >> match. I didn't think of that until later, so I wasn't able to ask him
> >> about this. Perhaps he meant implemented similarly, or perhaps ideas
> >> from CSTal's style of play exhibited in games Thorsten posted. Or
> >> maybe he had access to a beta version, or he was referring to Complete
> >> chess system, I don't really know.
> >
> >He could mean almost anything. But one thing I know for sure, that many
> >early, far earlier than pre-beta, versions of CSTal were doing the
rounds
> >of freaks and programming teams. Despite non-disclosure-agreements,
> >personal assurances, and conditions of general morality; the level of
> >dishonest behaviour, right at the heart, and right amongst some of the
> >loudest copy-protectors of their own software, is really quite high.
>
> Either that or he could have a vague clue about what you do, based
> upon what he's read here, or what the questioner told him, and
> answered that he does some of the same stuff.
>
> It's not like anyone really knows what you do.
>
> The rest of your post is mostly about hatred of other people,
> specifically Bob.
No.
The hatred is actually from you to me.
Since Paris, and probably before.
Chris Whittington
Rolf Tueschen
Robert Hyatt <hy...@crafty.cis.uab.edu> wrote:
>But I *did* follow my own advice, thank you very much. Now if you would do
>the same... we *might* have interesting discussions here. It's up to you...
Excuse me if I ask a simple question.
Just to understand. "It's up to you..."
I ask how such a amoral, indecent, ignorant fascist Fach-Idiot had the
right to make such a pretentious remark ...
Look, it's not my verdict alone. That the guy I talk about is way
outside the civilized world ...
It was Enrique, another prominent censor of the chess groups, who
judged about the disgustful Hyatt.
Read yourself:
==============
[quote begin]
>And I suspect they'd much have preferred for the state to tie his hands
>to a northbound bus, and his feet to a southbound bus, and let them take
>off *slowly*. But since it wasn't *my* child that was killed in such a
>terrible way, I can only imagine how they felt.
>
>But I'd gladly have been one of the bus drivers...
>
>And if you had stepped in front of *my* bus... :) two problems with
>one bus eliminated. :)
Even without considering why this should be posted in a computer chess
newsgroup or to what extent your statement is morally repugnant and
alien to developed forms of civilization, it is quite extraordinary to
notice such a degree of cultural differentiation. In any country of
the European Community, a public statement like yours would be
prosecuted by law for “instigation to violence” and “apology of
torture”. Both cases lead invariably to prison sentences.
Enrique
[quote end]
====================
So, Hyatt is a repugnant, alien, uncivilized species.
Not my words.
And such a beast should be allowed to determine the way how civilized
adults talk with one another here on rgcc???
We won't go that low, no??
Chris?
Chris Whittington
nob...@nsm.htp.org wrote in article
<1998051019450...@nsm.htp.org>...
> On Sun, 10 May 1998 18:16:23 GMT mcl...@prima.ruhr.de (mclane) wrote:
>
> >bru...@seanet.com (bruce moreland) wrote:
> >
> >>I would love to see another match between CST and Crafty.
> >
> >>bruce
> >
> >If you (or Moritz) tell me how to install crafty accurate (when I
> >tried it 1/2 a year ago, or even earlier, it was complicate. I had to
> >do this, take care of that, and and and).
> >But i will do anything you tell me what is needed !
> >Tell me what to do and I can do 40/120.
> >
> >We can bet about the result :-)))
> >
> >best wishes
> >
> >mclane
>
> TheDoDo says
>
>
> If under windoze, do this:
>
> instructions w/o gui - you wont need gui to play cstal, since cstal will
> give you the interface.
>
> 1. download wcrafty at ftp://ftp.cis.uab.edu/pub/hyatt
> {windoze version has "w" at front of executable.
>
> 2. download winboard - i.e. only if you want gui
>
> 3. make crafty directory - "md\crafty" or put it where you like
>
> 4. create "crafty.rc" file and put the following in it"
>
> log on
> resign 3
> ponder off
> hash 10M
> hashp 2M
> book random 1
> book width 2
> exit
> <eof>
>
> 5. download "small" book from hyatt ftp site - and put in crafty
directory
> ftp://ftp.cis.uab.edu/pub/hyatt/common/small.txt
>
> 6. download "learn.dat" from same place - and put in same place
> ftp://ftp.cis.uab.edu/pub/hyatt/v14/learn.zip
>
> 6a. download start.zip and put in crafty directory
> ftp://ftp.cis.uab.edu/pub/hyatt/common/start.zip
>
> 7. you should have wcrafty-xxx.exe, small.txt, learn.dat, and crafty.rc
> all in same directory now. Make sure it's all unzipped.
>
> 8. execute crafty - you're now at crafty command prompt
>
> 9. You need to create opening book now and import learned data:
> at crafty command prompt type:
>
> "book create small.txt 60 30" {this creates opening book from pgn
games}
> "books create start.pgn 60" {you now have a small opening book}
>
> 10. Import learned data with following command:
> "import learn.dat"
>
> 11. For a 40/120 game, type: "time 40/120"
>
> 12. at the "white" prompt, enter your move w/ standard algebraic notation
> e4 <enter>
>
> 13. Crafty will reply with simple algebraic response: "e5" etc.
> If out of book, it will start displaying PV.
>
> 14. the game is automatically stored in last "game.x" file.
>
> 15. to resign, type "resign"
>
> 16. to offer draw, type "draw"
>
> 17. to exit program, type "quit"
>
> 18. to get list of commands, type "help"
>
Admit it Ed, this is funny, no ? :))))))))))))))))))))))))))))))
Chris Whittington
>
>
>
Chris Whittington
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote in article
<6j4t58$ncs$1...@news01.btx.dtag.de>...
Why do you think I went on strike and stopped posting to CCC back in
November last year ?
You've forgotten who I'ld been submitting to moderation by ?
Chris Whittington
>
>
>
>
>
mclane
This is more complicate than playing chess or doing sex or operating
my computer !
Heavens !
I will try it. But I am sure I am ready with the preparations in july
:-)))
Oh man. How can testing a chess program be so complicate !!
best wishes
mclane
Moritz Berger
>
>
>nob...@nsm.htp.org wrote in article
><1998051019450...@nsm.htp.org>...
>> 4. create "crafty.rc" file and put the following in it"
>>
>> log on
>> resign 3
>> ponder off
don't (!)
Robert Hyatt
nob...@nsm.htp.org wrote:
: On Sun, 10 May 1998 18:16:23 GMT mcl...@prima.ruhr.de (mclane) wrote:
: TheDoDo says
: If under windoze, do this:
: instructions w/o gui - you wont need gui to play cstal, since cstal will
: give you the interface.
: 1. download wcrafty at ftp://ftp.cis.uab.edu/pub/hyatt
: {windoze version has "w" at front of executable.
: 2. download winboard - i.e. only if you want gui
: 3. make crafty directory - "md\crafty" or put it where you like
: 4. create "crafty.rc" file and put the following in it"
: log on
: resign 3
: ponder off
: hash 10M
: hashp 2M
: book random 1
: book width 2
: exit
: <eof>
: 5. download "small" book from hyatt ftp site - and put in crafty directory
: ftp://ftp.cis.uab.edu/pub/hyatt/common/small.txt
Doing ok to here... but "small" is *not* the book to use. It has *no*
useful information other than to produce random openings that have nothing
to do with "soundness" or anything. Far better to go into the KKUP
directory and download the book.bin file there, it is a prebuilt book that
will only work with the windows version of crafty, but it has decent games
with frequency counts that will provide reasonable openings when used.
: 6. download "learn.dat" from same place - and put in same place
: ftp://ftp.cis.uab.edu/pub/hyatt/v14/learn.zip
Not necessary if the above book.bin is downloaded, the learn data is
already included.
:
: 6a. download start.zip and put in crafty directory
: ftp://ftp.cis.uab.edu/pub/hyatt/common/start.zip
: 7. you should have wcrafty-xxx.exe, small.txt, learn.dat, and crafty.rc
: all in same directory now. Make sure it's all unzipped.
: 8. execute crafty - you're now at crafty command prompt
: 9. You need to create opening book now and import learned data:
: at crafty command prompt type:
: "book create small.txt 60 30" {this creates opening book from pgn games}
Skip this step if you download book.bin, as it is already ready to go. You
still need to build books.bin as per the read.me or the instructions given
here..
: "books create start.pgn 60" {you now have a small opening book}
: 10. Import learned data with following command:
: "import learn.dat"
not needed if KKUP/book.bin used...
: 11. For a 40/120 game, type: "time 40/120"
: 12. at the "white" prompt, enter your move w/ standard algebraic notation
: e4 <enter>
: 13. Crafty will reply with simple algebraic response: "e5" etc.
: If out of book, it will start displaying PV.
: 14. the game is automatically stored in last "game.x" file.
: 15. to resign, type "resign"
: 16. to offer draw, type "draw"
: 17. to exit program, type "quit"
: 18. to get list of commands, type "help"
:
Robert Hyatt
Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
: bruce moreland <bru...@seanet.com> wrote in article
: <3556dd7a....@news.seanet.com>...
:> On Sun, 10 May 1998 15:46:34 GMT, "Chris Whittington"
:> <chr...@cpsoft.demon.co.uk> wrote:
:>
:> >Now, Mr Hyatt, if you're going to call on the massed ranks of the
:> >petit-bourgoisie to assemble for the counter-revolution, marching behind
:> >the banner "Crafty has a king-safety function, honest"; then it would be
: as
:> >well if you maintained your promise to give me the great honour and
:> >privilege of being in your kill-file.
:>
:> I would love to see another match between CST and Crafty.
: Binaries at dawn, eh Bruce ?
: I suppose if Crafty wins then it's allowed to dismember drug dealers by
: pulling them apart with buses ?
: Chris Whittington
No, but maybe a "certain poster" will simply stop trying to tell *me* how
*my* program is defectively designed. I see enough defects to go around
the list *twice*.
Rather than telling you your approach is defective, I simply suggest that
you prove mine is?
:>
:> bruce
:>
:>
Keith Ian Price
>
>
> Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
> > 10. Since we had been talking about evaluations and positional
> > understanding, I took the time at this point to bring up my current
> > favorite among the chess programs I have, Chess System Tal. I stated
> > that I was impressed with the amount it accomplished within 3000 nodes
> > per second searched. I said that if its evaluation was able to search
> > at a much higher rate, that I thought it would be much better than the
> > other micro programs. I was surprised by the enthusiasm Hsu showed
> > about the program.
>
> Likewise :)
>
> > He mentioned that how it handled King Safety was
> > much more similar to Deep Blue than the other micro programs, although
> > perhaps a little bit more extreme, and that he, too, was impressed
> > with it.
>
> Hmmm. Not sure I believe this. If DB had king-safety terms like CSTal has
> king safety terms, it would have certainly seen the draw possibilities in
> the famous missed draw game - very early.
Deep Blue was not playing for the draw. True, CSTal decides on Qe3
after 32 minutes if playing Kasparov's side, but it doesn't see a
draw, just the least painful continuation. So, it is not clear that DB
did not see this move, but merely that it didn't see it was a forced
draw. This took a minimum of 30-ply brute force to guarantee not
missing it (according to analysis done by others), so if the selective
extensions had not carried that far, it may have looked like a good
continuation, but not a draw. Besides, he said your king safety
routines were a little excessive. Perhaps if Deep Tal had been playing
with that search depth, it would have resigned, because its King was
so exposed. ;-)
> > He said that many things in CSTal were implemented in Deep
> > Blue, which I found strange, since it wasn't released until after the
> > match. I didn't think of that until later, so I wasn't able to ask him
> > about this. Perhaps he meant implemented similarly, or perhaps ideas
> > from CSTal's style of play exhibited in games Thorsten posted. Or
> > maybe he had access to a beta version, or he was referring to Complete
> > chess system, I don't really know.
>
> He could mean almost anything. But one thing I know for sure, that many
> early, far earlier than pre-beta, versions of CSTal were doing the rounds
> of freaks and programming teams. Despite non-disclosure-agreements,
> personal assurances, and conditions of general morality; the level of
> dishonest behaviour, right at the heart, and right amongst some of the
> loudest copy-protectors of their own software, is really quite high.
Well, I must admit, you've made a cool program. When are you going to
get the Pentium II Xeon processor with 2MB of 450Mhz L2 cache to
prepare for this years WCCC? The difference a little speed made with
your program was apparent to me when CSTal played Re1! in the first
round of the WMCCC last year. It would never have seen that gem on my
150 Mhz Pentium in three minutes.
> Mostly however, and amusingly, these persons didn't recognise what was
> staring them in the face.
And that was? (My face blank, my eyes glazed...)
> I guess it is quite possible that somebody sent a version to Hsu. I
> recollect at the time, that I was part of a vociferous campaign that DB
> didn't stand a chance in hell since it was a materialistic bean-counter in
> the Hyatt-Crafty mould.
Bean there, won that! :-)
> However, Hsu's GMs wouldn't have needed CSTal to tell him that what was
> lacking in the DB program, as in any materialistic-Hyatt style program, was
> dynamic stuff to deal with that which the
> quiescent-seeking-Hyattian-nonsense tried to ignore. Namely positions
> leading to mating attacks. Duh.
And that's supposed to be important when you're a Rook up?
> > I only mention this as it gives a
> > little insight into the approach used in the evaluation. During the
> > presentation Hsu stated that unlike other chess programs, DB's
> > evaluation in not just a matter of adding weights together with
> > bonuses to arrive at a score, but some functions were calculated
> > non-linearly, through multiplication, or other "second level" methods.
> This is repeated again and again, in all kind of situations, and is just
> woffle.
Yes, this was part of the general presentation. I included it because
I was talking about evaluation routines. Woffle?
> I guess if your program has moved beyond the stage of adding up 32 piece
> square table entries, then you might want to start crowing about
> 'non-linear second-level differential equations packed with Greek letters,
> and things to the power of'. It might even do as a ICCAJ article title. But
> meaningless nevertheless.
How many bishops can dance on the head of a pinned Queen? Using the
RMS method, only one, unless underpromotion has taken place earlier in
the game, which would, of course, require the recursive retrograde
Fourier transform regression analysis to determine the answer.
> > When asked about this, he said that some examples of the non-linear
> > evaluations were in the method of calcualating a pawn's value based on
> > it's advance, and its position relative to other pieces and pawns, and
> > King safety, which was an example of what he had referred to as
> > "second level" methods.
> I'd call it 'dynamic'. ie. evaluating with respect to what the pieces can
> do in combination with each other. Very powerful.
A symphony of patterns! I've always thought of a chess game in this
way. Better, of course, when the harmony comes from my side, and the
discordant cacophany issues from the other, but against most chess
machines lately, it's the other way round for me!
> The Hyattian quiescence paradigm can't do this. This can't evaluate
> anything to do with chess, especially not those 'second-level' things. So
> they try and extend to a 'quiet' position (defined as one with no
> captures), so that they can then basically just add up the material, and
> not worry about any dynamic features (because, by their definition, there
> aren't any when quiet).
Does this mean you doubt their definition of quiet, which seems right
to me (no dynamic features possible) or you doubt that they can ever
achieve such a position in an actual game where there are more than 2
Kings left?
> See the big hole in the logic ? They can't evaluate any dynamic king stuff.
> So they don't know whether their 'quiet' position is quiet or not with
> respect to the king. They just ignore it :)
> The result is a materialistic bean-counter. Which, as Thorsten has shown in
> many published test games, has no idea when it is on receiving end of a
> vicious king attack. Or not until too late.
And this is fun, no??
> > This was a question and answer section at the
> > end of the presentation, and since it wasn't my question, I could not
> > ask him to expand on these generalisms.
> Well, that leave me to expand, then ...... :)
> 1. As has been repeatedly said, the bean-counters will get nowhere until
> they abandon their quiesence paradigm.
> 2. The reason the bean-counting programmers use the quiesence paradigm is
> because they know zilch about chess. Fach-idiots. Experts at programming.
> Stupid at chess.
That's why they chose GM assistants to help them with the evaluation.
> Anyroads, if the GM's chess knowledge + Hsu's programming/hardware skills
> were put together into DB, effectively, then no GM or super-GM would be
> able to withstand.
This was the challenge. I have no idea if GMs would have a chance.
Super-GMs are better than those who were assisting, so they might have
a chance, even given an effective implementation. The difficulty is
that transfer of the GM's chess knowledge into an effective algorithm
that covers all bases. This is where an expert chessplayer-programmer
might have an advantage over a GM and programmer team, since you don't
have to explain to yourself what you want the evaluation to do, and
you wouldn't have to try to figure out the convoluted explanation you
would likely come up with if you had to.
> My only doubt with regard to DB remains with the quiesence paradigm, which
> must have been at the heart of the program. Did they throw the paradigm out
> ? Almost impossible in one year.
He did say they do not do a normal quiescence search. He was answering
someone else's question, and I cannot remember exactly what he did
say.
> Did they implement dynamic king-safety/attack stuff and incorporate into
> the evaluation ?
Yes. He did say this.
> Quite possible. But one of the features of king attack
> situations is their unresolvability within any reasonable time frame. You
> can exhaust captures fairly quickly, and then evaluate; but exhausting
> checks and mate threat attacks ? No way.
Maybe at 200 million nps it is not such a limitation. Too bad Digital
Equipment got bought out by Compaq. You might have convinced them to
let you do a Deep Maroon, and we could have these questions answered.
2000 Alphas at 767Mhz, each with 32 Deep Tal Processors at 8 million
nps, with dynamic king safety extensions, and no Freon or CFCs used to
cool the thing! :-) All kidding aside, I'm sure that the future needs
more hardware chess processors in order to get all the chess knowledge
you want to implement into a program that can exhaust all checks and
mating threats, as well. It would be the next step.
-----------------------------------------------------------------
kp
Keith Ian Price
On Sun, 10 May 1998 13:50:53, mcl...@prima.ruhr.de (mclane) wrote:
> kpr...@spamfree.teleport.com (Keith Ian Price) wrote:
>
> >Sorry for the delay. Here is Part III of my Deep Blue Report:
>
> First let me say thanks for your reports.
> I have attacked this machine so often and projected my prejudices
> towards it, that I am always glad to get information that is not
> arrogant and snobistic.
Thanks for your appreciation. It would be hard to be arrogant since I
had nothing to do with the creation of Deep Blue. I suspect that you
mean, however, that I do not spout the "Deep Blue is a 2900 ELO
monster" line, and that I still think that Kasparov could beat it in a
match where he has time to rest, and is allowed to see its games to
prepare beforehand. While this is what I think, I must admit that I
would not bet a great deal of money on it, since 6 games is not enough
to form a good opinion about such things. Particularly when Kasparov
was playing such strange chess, so that it can't be seen how he would
have fared with a regular closed game instead of his "clam attack".
Even Karpov seemed confused by his play, and after the match (on
Chess.net) he stated that it should have been him playing, he would
have beaten it easily. The real shame is that they will not run it
anymore, since they do not want to risk losing. It would be nice to
know exactly where it stands among the top players. I am not sure that
the future will not bring it back out, after all memory of the match
fades from the general public's mind, and IBM salesman says, "This is
the same sort of scalable processor that was used in Deep Blue", and
the Information Technology manager says, "Deep What?" Then IBM will
say, "Ok, time for the chips to be redone in .08 micron technology,
add 20000 more features, and lets let Kasparov have his rematch, this
time for the World Championship." Who knows...
> When I was a teen I dived deep into computerchess. I read Opfermann's
> strange books about the chess-players all over the world (his style is
> very associative, he build links between chess and other related
> topics), followed each of David Levy's games with my Mephisto II and
> ate any PIONIER article about AI and Botwinnik.
I used to subscribe to a magazine called Personal Computing in the
'70s, which was basically boring to me since it covered all sorts of
computers that I couldn't afford, but it would have a large section
each issue on Computer chess that would fascinate me.
> In germany Helmut-Pfleger and Vlastimil Hort commented TV-world-cups.
> I saw Seiravan, very smart clothes and I am lucky that he has not
> developed negatively since now. I read Frederic's globetrotter
> articles about: Fred visits chess-programmers all over the world, says
> hello to Dave and Kathe and and and.
I sent off for the source code to Sargon when it first came out. They
offered the Z-80 source code, and I purchased it, since I couldn't
afford a Z-80 to run it on. I was hoping to convert it to run on a
PDP-11 I had access to at work, but the graphic were for a Juno
computer (or Jupiter), and since I didn't know how the graphics
routines produced graphics with that system, I would have had to redo
it from scratch in assembler for a Teletype, which seemed to be less
cool than I had hoped. So I left it for the moment until I got a UCSD
pascal pseudocode compiler for the PDP-11, and then I wrote a quick
and cheesy chess program in Pascal. It was so poor compared to my
Chess Challenger 10, even, that I got discouraged and didn't try
anything again, as far as programming. (I did start another one for my
Atari ST, but never finished it.)
Incidentally, Hsu mentioned that your globe-trotting friend visited
them at the Watson Center, acting as a "second" for Kasparov, and
pronounced as he was about to leave that Kasparov could easily beat
Deep Blue if he were to play anti-computer chess, but that he would
never agree to do this. So, if this was not simply a smoke screen,
perhaps Friedel is the advisor behind the strange play of Kasparov.
> In my dreams when I went to sleep in nights dedicated chess computers
> and Super-Constellations and names like HITECH and Berliner circeled
> arround.
> This was like drugs.
I know what you mean.
> We thought: ONE day the AI machine will be developed. ONE day. This
> was my dream. I was very passive, only a consumer. Not able to change
> or help or do anything. I was forced to read the magazines, and that
> was all.
> Now I am 31 and many things have changed.
> When I read about Deep Blue and Kasparov and their "match" i thought
> materialist fight in an advertising-campaign. He gets money, gives his
> name, and they pay and get the games and the fame, if they have a
> change. This was all very materialistic to me.
> After I saw game 2 anything was clear to me. This match is not a real
> match, it is manipulated. This is not the right way to play against a
> machine, like he has done in the first game.
I'm not sure that leaving all pieces on the 3rd rank or less is the
right way to play the first game either, even if he did win it.
Usually White needs to be a little more agressive, or gives up his
initiative.
Well, actually, I believe he just wanted to say he was not giving out
the game scores, but he was being polite as I kept trying different
tacks. If you are still doubting the 10-0 outcome, let me tell you
that after talking to him, I do not. If it was a made-up story, he
would have told me the results and that would have been it. Besides
not having any reason to make up this story, Hsu expressed such
disgust at the way the micros played, that he could have won an Oscar
for the performance if it was faked. I am quite convinced of the 10-0
outcome, but would still like to see the scores for the games to see
how the chess processor plays.
> >10. Since we had been talking about evaluations and positional
> >understanding, I took the time at this point to bring up my current
> >favorite among the chess programs I have, Chess System Tal.
> You did this really ??!!
> Amazing.
Why? I wanted to see what he thought of it, if he had even heard of
it. I got more than I expected.
> > I stated
> >that I was impressed with the amount it accomplished within 3000 nodes
> >per second searched.
> on a 200 Mhz machine, yes.
Actually, I have a 150 non MMX. So it's more like 2500, but I rounded
up.
> >I said that if its evaluation was able to search
> >at a much higher rate, that I thought it would be much better than the
> >other micro programs.
> Could be...
I really think so. I'd love to try it on a 450 Mhz Pentium II Xeon
with 2MB of 450 Mhz L2 cache. It would probably barely fill the 2MB of
cache with its hash tables in 3 minutes.
> > I was surprised by the enthusiasm Hsu showed
> >about the program.
> I am surprised to read this too. Is this materialistic-HSU maybe a
> resigned idealist ?
I think you may have let your prejudice rule over you here. Just
because Hsu uses massive speed for searching doesn't mean that he
sacrifices knowledge. From talking to him, he was so happy to have
Joel Benjamin help him put extra "GM-class" knowledge into the chips,
that I suspect that he is a lot more like you than you think. He
stated what made him start out on his project 12 years ago. He said
that he loved watching Star Trek, and when the Klingons would go into
battle they would say "Today is a glorious day to die". Well, he
thought, if I take on this battle (to beat the World Champion), I am
not likely to die, but if I succeed, it would be glorious. When I
heard him say this, I thought of you, mclane. Maybe you are sure his
is just a fast searcher because he took the job with IBM, and IBM is a
global conglomerate, and so the enemy. I think he took the job with
IBM so he could get better equipment to run his toy on, and to achieve
his goal. Tell me if you had the opportunity to put CSTal on a chip so
you could have the same eval and yet search 2-2.5 million nps, that
you wouldn't do this, because it would be materialistic. I'd have a
tough time believing that.
> If somebody gives up his dreams, or has to give up them, he is maybe
> so depressed that he turnes arround and becomes a materialist.
Hsu is not depressed. He seems quite pleased with what he
accomplished.
> > He mentioned that how it handled King Safety was
> >much more similar to Deep Blue than the other micro programs, although
> >perhaps a little bit more extreme, and that he, too, was impressed
> >with it.
> !!! :-))))
I knew this would make you happy.
> Although we never designed it for Deep-Blue engineers... :-)))
> > He said that many things in CSTal were implemented in Deep
> >Blue, which I found strange, since it wasn't released until after the
> >match.
> You forgot that we worked for arround 6 years on it and that I
> published many games...
> >I didn't think of that until later, so I wasn't able to ask him
> >about this. Perhaps he meant implemented similarly,
> I guess I know what he means.... :-))
> > or perhaps ideas
> >from CSTal's style of play exhibited in games Thorsten posted.
> How shall he know the games I post ? I don't think people like him
> have time to read rgcc ?!
I think he sees most of the things posted here about Deep Blue. He
would not admit to reading rgcc, but he didn't deny it either, when I
asked him.
> >Or
> >maybe he had access to a beta version,
> Clear No !
Chris seems to think it's possible.
> >or he was referring to Complete
> >chess system, I don't really know.
> I don't think you can be enthusiastic by relating to Complete Chess
> system, or ?
> :-)
I didn't think so, either.
<snipped some stuff>
> Thanks for your little report. I laughed much about this. I am sure we
> will kill Deep Blue one day, no matter how many NPS it makes. :-))
> The slowest beats the fastest ? My dreams come through ?!
Or maybe Deep Tal will someday beat Deep Fritz by packing more chess
knowledge into the EEPCPs (Electrically Erasable and Programmable
Chess Processors) of the future.
-----------------------------------------------------------------
kp
Robert Hyatt
Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
: Admit it Ed, this is funny, no ? :))))))))))))))))))))))))))))))
: Chris Whittington
Depends. How many operating systems do *you* support? How many different
computing platforms?
I can run on DOS, Win3.1, win95, winNT, linux, unix, HPUX, digital unix,
Cray unicos, IBM aix, digital VMS, IBM MVS, Amiga, Next, FreeBSD, and others.
I can compile and run on PC's, alphas, crays, PPC's, HP's, Sun's, Fujitsu's,
Hitachi's, parallel machines, uniprocessor machines, etc.
How many computers do you support? How many operating systems do you
support?
That *I* consider funny. I can't do prebuilt books because of endian and
compiler issues, wordsize issues, etc. However, those that follow the read.me
end up with a working program at a reasonable price/performance point. And
they aren't stuck on a uniprocessor PC running either dos or win95. They
might choose a 16 processor Sun, or an 8 processor alpha, or a 4 processor
pentium pro, or you-name-it.
Flexibility has its price. But as far as "funny?" I suppose all the crafty
users can "keep right on laughing"...???
Oh, I almost forgot... what about supporting automatic play on chess servers,
sharing learning among multiple versions? Your refined installation is certainly
one advantage of a commercial package. However, Crafty has a significant
number of advantages that offset the more complicated installation... as the
first time you play a 16 processor ultra sparc machine will show you...
rag...@miracon.net
On 11 May 1998 03:44:53 GMT, Robert Hyatt <hy...@crafty.cis.uab.edu>
wrote:
>That *I* consider funny. I can't do prebuilt books because of endian and
>compiler issues, wordsize issues, etc. However, those that follow the read.me
>end up with a working program at a reasonable price/performance point.
Now _that's_ the understatement of the year!
Joe Ragazzi
bruce moreland
>No.
>
>The hatred is actually from you to me.
>
>Since Paris, and probably before.
>
>Chris Whittington
Yes, I don't like you at all. Since way before Paris.
I can talk with you about computer chess without it coming through
much though.
And I'm not going to insult your program, I think you had some cool
ideas and did a good job building it.
bruce
Howard E
Chris Whittington wrote in message <01bd7bfb$6e3f6ca0$LocalHost@chess>...
>
>
>Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
><ihDRG0dUkHpV-p...@pdx03-pm2-02.teleport.com>...
>> He said that many things in CSTal were implemented in Deep
>> Blue, which I found strange, since it wasn't released until after the
>> match. I didn't think of that until later, so I wasn't able to ask him
>> about this. Perhaps he meant implemented similarly, or perhaps ideas
>> from CSTal's style of play exhibited in games Thorsten posted. Or
>> maybe he had access to a beta version, or he was referring to Complete
>> chess system, I don't really know.
>
>He could mean almost anything. But one thing I know for sure, that many
>early, far earlier than pre-beta, versions of CSTal were doing the rounds
>of freaks and programming teams. Despite non-disclosure-agreements,
>personal assurances, and conditions of general morality; the level of
>dishonest behaviour, right at the heart, and right amongst some of the
>loudest copy-protectors of their own software, is really quite high.
My guess is that someone brought to his attention the game Genius vs CSTal.
My memory is a bit fuzzy on when the game took place but the move played by
CSTal to win that game has held up today as a tough move for other programs
to find. The move was a bishop move to the c file (someone might confirm the
actual position) that was very human-like in that it placed the bishop in
line with the opposing
king with a hoard of pieces inbetween. In my own wacky category of moves I
call this
type of move, "easy for humans but tough for computers". At least it would
be an easy
move to see that deserved serious consideration. I also recall that the
"lowly" P90
was the machine in use for this game.
With Hsu's resources, he probably said, "get me that program!"
So much for my detective Columbo ramblings.
Chris Whittington
Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
<6j5s7l$mn8$1...@juniper.cis.uab.edu>...
> Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
>
>
> : Admit it Ed, this is funny, no ? :))))))))))))))))))))))))))))))
>
>
> : Chris Whittington
>
>
> Depends. How many operating systems do *you* support? How many
different
> computing platforms?
>
> I can run on DOS, Win3.1, win95, winNT, linux, unix, HPUX, digital unix,
> Cray unicos, IBM aix, digital VMS, IBM MVS, Amiga, Next, FreeBSD, and
others.
>
> I can compile and run on PC's, alphas, crays, PPC's, HP's, Sun's,
Fujitsu's,
> Hitachi's, parallel machines, uniprocessor machines, etc.
>
> How many computers do you support? How many operating systems do you
> support?
>
> That *I* consider funny. I can't do prebuilt books because of endian and
> compiler issues, wordsize issues, etc. However, those that follow the
read.me
> end up with a working program at a reasonable price/performance point.
And
> they aren't stuck on a uniprocessor PC running either dos or win95. They
> might choose a 16 processor Sun, or an 8 processor alpha, or a 4
processor
> pentium pro, or you-name-it.
>
> Flexibility has its price. But as far as "funny?" I suppose all the
crafty
> users can "keep right on laughing"...???
>
> Oh, I almost forgot... what about supporting automatic play on chess
servers,
> sharing learning among multiple versions? Your refined installation is
certainly
> one advantage of a commercial package. However, Crafty has a significant
> number of advantages that offset the more complicated installation... as
the
> first time you play a 16 processor ultra sparc machine will show you...
>
Oh, dear :)
Well I just thought it might have reminded Ed of the 100 versions of Crafty
rebuilds required to play one auto232 game last year .....
I think he gave up in end, no ?
BTW what happened to my entry in your kill file ? You promised.
Chris Whittington
Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
<6j5g82$h6a$1...@juniper.cis.uab.edu>...
HELP !
Peter Herttrich
mclane <mcl...@prima.ruhr.de> wrote:
: nob...@nsm.htp.org wrote:
....
: >17. to exit program, type "quit"
: >18. to get list of commands, type "help"
: >
: This is more complicate than playing chess or doing sex or operating
: my computer !
: Heavens !
: I will try it. But I am sure I am ready with the preparations in july
: :-)))
: Oh man. How can testing a chess program be so complicate !!
Du bist doch nicht etwa ein Tastaturleghasteniker ? ;-]
Die wahre Macht ueber ein Betriebssytem hat man nur mit der
Tastatur!
Deswegen: Use UNIX or die! :-)))
Peter
(using LINUX since version 0.9)
--
--
_____________________________________________________________________________
Peter Herttrich email: dh1...@inss1.etec.uni-karlsruhe.de
Universitaet Karlsruhe Tel. +49 721 6083747 FAX +49 721 6086071
Institut fuer Nachrichtentechnik ..life outside caves is complicated ...
_____________________________________________________________________________
Chris Whittington
Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
<6j5gc9$h6a$2...@juniper.cis.uab.edu>...
> Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
>
>
> : bruce moreland <bru...@seanet.com> wrote in article
> : <3556dd7a....@news.seanet.com>...
> :> On Sun, 10 May 1998 15:46:34 GMT, "Chris Whittington"
> :> <chr...@cpsoft.demon.co.uk> wrote:
> :>
> :> >Now, Mr Hyatt, if you're going to call on the massed ranks of the
> :> >petit-bourgoisie to assemble for the counter-revolution, marching
behind
> :> >the banner "Crafty has a king-safety function, honest"; then it would
be
> : as
> :> >well if you maintained your promise to give me the great honour and
> :> >privilege of being in your kill-file.
> :>
> :> I would love to see another match between CST and Crafty.
>
> : Binaries at dawn, eh Bruce ?
>
> : I suppose if Crafty wins then it's allowed to dismember drug dealers by
> : pulling them apart with buses ?
>
> : Chris Whittington
>
> No, but maybe a "certain poster" will simply stop trying to tell *me* how
> *my* program is defectively designed. I see enough defects to go around
> the list *twice*.
>
> Rather than telling you your approach is defective, I simply suggest that
> you prove mine is?
>
That would not be possible. You're paradigm bound.
Chris Whittington
>
>
> :>
> :> bruce
> :>
> :>
Chris Whittington
bruce moreland <bru...@seanet.com> wrote in article
<35569e84....@news.seanet.com>...
> On Sun, 10 May 1998 19:00:33 GMT, "Chris Whittington"
> <chr...@cpsoft.demon.co.uk> wrote:
>
> >No.
> >
> >The hatred is actually from you to me.
> >
> >Since Paris, and probably before.
> >
> >Chris Whittington
>
> Yes, I don't like you at all. Since way before Paris.
>
> I can talk with you about computer chess without it coming through
> much though.
No you can't. It comes over endlessly and repetitively, colouring many of
your moves. Such that the apparent cool rationality of the Bruce is
actually a mask for robotic-like behaviour patterns.
And I don't think it's just me. I think there's a range of
persons/organisations you're reacting on. For one, against the other,
apparently on rational grounds, but, I'm convinced the real reason is
emotional like/dislike.
And you let it get the better of you. I mean your 'it's him or me' in the
CCC was pretty crass, no ?
Chris Whittington
Chris Whittington
Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
<ihDRG0dUkHpV-pn2-Fw9O8bWuCBdf@localhost>...
Well that's the Tal function finding something interesting (less painful if
you like) by evaluation ...
> So, it is not clear that DB
> did not see this move, but merely that it didn't see it was a forced
> draw.
Then it's not evaluating like CSTal.
> This took a minimum of 30-ply brute force to guarantee not
> missing it (according to analysis done by others), so if the selective
> extensions had not carried that far, it may have looked like a good
> continuation, but not a draw.
Again, in this case, it can't be evaluating like CSTal.
> Besides, he said your king safety
> routines were a little excessive.
They are v e r y excessive :)
> Perhaps if Deep Tal had been playing
> with that search depth, it would have resigned, because its King was
> so exposed. ;-)
>
It might have resigned if playing the DB side ? :) yes, very possible.
>
> > > He said that many things in CSTal were implemented in Deep
> > > Blue, which I found strange, since it wasn't released until after the
> > > match. I didn't think of that until later, so I wasn't able to ask
him
> > > about this. Perhaps he meant implemented similarly, or perhaps ideas
> > > from CSTal's style of play exhibited in games Thorsten posted. Or
> > > maybe he had access to a beta version, or he was referring to
Complete
> > > chess system, I don't really know.
> >
> > He could mean almost anything. But one thing I know for sure, that many
> > early, far earlier than pre-beta, versions of CSTal were doing the
rounds
> > of freaks and programming teams. Despite non-disclosure-agreements,
> > personal assurances, and conditions of general morality; the level of
> > dishonest behaviour, right at the heart, and right amongst some of the
> > loudest copy-protectors of their own software, is really quite high.
>
> Well, I must admit, you've made a cool program. When are you going to
> get the Pentium II Xeon processor with 2MB of 450Mhz L2 cache to
> prepare for this years WCCC?
I'm recruiting bodyguards right now :)
> The difference a little speed made with
> your program was apparent to me when CSTal played Re1! in the first
> round of the WMCCC last year. It would never have seen that gem on my
> 150 Mhz Pentium in three minutes.
>
> > Mostly however, and amusingly, these persons didn't recognise what was
> > staring them in the face.
>
> And that was? (My face blank, my eyes glazed...)
That the king has value = infinity
>
> > I guess it is quite possible that somebody sent a version to Hsu. I
> > recollect at the time, that I was part of a vociferous campaign that DB
> > didn't stand a chance in hell since it was a materialistic bean-counter
in
> > the Hyatt-Crafty mould.
>
> Bean there, won that! :-)
But is it a bean-counter a la Hyatt, or not ? Nobody seems to know.
>
> > However, Hsu's GMs wouldn't have needed CSTal to tell him that what was
> > lacking in the DB program, as in any materialistic-Hyatt style program,
was
> > dynamic stuff to deal with that which the
> > quiescent-seeking-Hyattian-nonsense tried to ignore. Namely positions
> > leading to mating attacks. Duh.
>
> And that's supposed to be important when you're a Rook up?
In the rather large game tree space that arises from a relatively balanced
position, most of the lines of consequence are going to be material
equal-ish.
Most of them are also going to be king-attack equal-ish also.
And it's the handling of these situations that leads to win/lose, no ?
And rook-up doesn't come into it.
The Hyatt-paradigm handles the former case, but not the latter. That's my
point.
>
> > > I only mention this as it gives a
> > > little insight into the approach used in the evaluation. During the
> > > presentation Hsu stated that unlike other chess programs, DB's
> > > evaluation in not just a matter of adding weights together with
> > > bonuses to arrive at a score, but some functions were calculated
> > > non-linearly, through multiplication, or other "second level"
methods.
>
> > This is repeated again and again, in all kind of situations, and is
just
> > woffle.
>
> Yes, this was part of the general presentation. I included it because
> I was talking about evaluation routines. Woffle?
It's ok, it wasn't your woffle :)
1. Without building comprehensive attack tables, and performing substantial
processing, they can't make any meaningful assessment of king-attack. They
say they do, but they are lying/mistaken/bluffing.
2. I can go for the quiet concept based on material only. You can exhaust
the captures relative fast. Fair enough.
3. But I don't go for it based on king attack. You can and do and see
highly tence king attack situations maintained over many moves. These lines
do not necessarily exhaust. Therefore, IMO, you have to try and evaluate
them.
4. The Hyattian paradigm doesn't either evaluate, nor try to exhaust, it
just ignores.
>
> > See the big hole in the logic ? They can't evaluate any dynamic king
stuff.
> > So they don't know whether their 'quiet' position is quiet or not with
> > respect to the king. They just ignore it :)
>
> > The result is a materialistic bean-counter. Which, as Thorsten has
shown in
> > many published test games, has no idea when it is on receiving end of a
> > vicious king attack. Or not until too late.
>
> And this is fun, no??
Of course.
>
> > > This was a question and answer section at the
> > > end of the presentation, and since it wasn't my question, I could not
> > > ask him to expand on these generalisms.
>
>
> > Well, that leave me to expand, then ...... :)
>
> > 1. As has been repeatedly said, the bean-counters will get nowhere
until
> > they abandon their quiesence paradigm.
>
> > 2. The reason the bean-counting programmers use the quiesence paradigm
is
> > because they know zilch about chess. Fach-idiots. Experts at
programming.
> > Stupid at chess.
>
> That's why they chose GM assistants to help them with the evaluation.
Interestingly, only after Match 1. And presumably after being bollocked by
IBM officials. I don't think the natural tendency in the DB team was to
evaluate chess. It was to be fast.
>
> > Anyroads, if the GM's chess knowledge + Hsu's programming/hardware
skills
> > were put together into DB, effectively, then no GM or super-GM would be
> > able to withstand.
>
> This was the challenge. I have no idea if GMs would have a chance.
> Super-GMs are better than those who were assisting, so they might have
> a chance, even given an effective implementation. The difficulty is
> that transfer of the GM's chess knowledge into an effective algorithm
> that covers all bases. This is where an expert chessplayer-programmer
> might have an advantage over a GM and programmer team, since you don't
> have to explain to yourself what you want the evaluation to do, and
> you wouldn't have to try to figure out the convoluted explanation you
> would likely come up with if you had to.
True.
>
> > My only doubt with regard to DB remains with the quiesence paradigm,
which
> > must have been at the heart of the program. Did they throw the paradigm
out
> > ? Almost impossible in one year.
>
> He did say they do not do a normal quiescence search. He was answering
> someone else's question, and I cannot remember exactly what he did
> say.
>
> > Did they implement dynamic king-safety/attack stuff and incorporate
into
> > the evaluation ?
>
> Yes. He did say this.
Yes, but Hyatt says he does this too. It's not so.
>
> > Quite possible. But one of the features of king attack
> > situations is their unresolvability within any reasonable time frame.
You
> > can exhaust captures fairly quickly, and then evaluate; but exhausting
> > checks and mate threat attacks ? No way.
>
> Maybe at 200 million nps it is not such a limitation. Too bad Digital
> Equipment got bought out by Compaq. You might have convinced them to
> let you do a Deep Maroon, and we could have these questions answered.
> 2000 Alphas at 767Mhz, each with 32 Deep Tal Processors at 8 million
> nps, with dynamic king safety extensions, and no Freon or CFCs used to
> cool the thing! :-) All kidding aside, I'm sure that the future needs
> more hardware chess processors in order to get all the chess knowledge
> you want to implement into a program that can exhaust all checks and
> mating threats, as well. It would be the next step.
>
After finishing off the Windows version here, the next step is to redesign
CSTal more cleanly; speed it up, change a few ideas .....
One major change will be, instead of inefficient programmer, me, doing the
lower level driving routines (data structure/move/unmove/movegen etc),
we've recruited (and are still recruiting) programmers to do this for me.
I'll then just concentrate on the top level stuff and testing.
So you can look forward to Crafty II v CVSTal II and the result 20-0 to
Crafty II :)
Chris Whittington
>
> -----------------------------------------------------------------
> kp
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
Chris Whittington
Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
<ihDRG0dUkHpV-pn2-bVSTNpUPoFqc@localhost>...
I said I thought it was possible, but that it wouldn't make much difference
if he had. The GM's would have told him what to do. Maybe the first step is
just knowing that making a maniac is possible. But you can see that from a
game score, you don't need the program.
However, remember when Vincent came to us at AEGON begging for a test
version since he was an impecunious student ..... ? And then wrote
disparaging comments ? And we told him it was a test version, not for
publicity ? And he said so what, I'll do what I want ...... ?
There were quite a few people we (mistakenly?) gave test versions to. Any
of them could have leaked them around, no ?
Chris Whittington
Chris Whittington
nob...@nsm.htp.org wrote in article
<1998051109334...@nsm.htp.org>...
> >No you can't. It comes over endlessly and repetitively, colouring many
of
> >your moves. Such that the apparent cool rationality of the Bruce is
> >actually a mask for robotic-like behaviour patterns.
> >
> >And I don't think it's just me. I think there's a range of
> >persons/organisations you're reacting on. For one, against the other,
> >apparently on rational grounds, but, I'm convinced the real reason is
> >emotional like/dislike.
> >
> >And you let it get the better of you. I mean your 'it's him or me' in
the
> >CCC was pretty crass, no ?
> >
> >Chris Whittington
> >
>
>
> TheDoDo says:
>
> Were you drunk when you wrote this?
>
>
At 09:49 on a Monday morning ? Do me a favour.
Chris Whittington
>
>
>
>
>
>
>
>
Chris Whittington
nob...@nsm.htp.org wrote in article
<1998051109234...@nsm.htp.org>...
> On Mon, 11 May 1998 09:10:24 GMT "Chris Whittington"
<chr...@cpsoft.demon.co.uk> wrote:
>
> >One major change will be, instead of inefficient programmer, me, doing
the
> >lower level driving routines (data structure/move/unmove/movegen etc),
> >we've recruited (and are still recruiting) programmers to do this for
me.
> >I'll then just concentrate on the top level stuff and testing.
>
> TheDoDo says:
>
> Always in the driver's seat, are ya?
>
Seems so. Jealous ?
Chris Whittington
Rolf Tueschen
"Chris Whittington" <chr...@cpsoft.demon.co.uk> wrote:
>Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
><6j5s7l$mn8$1...@juniper.cis.uab.edu>...
>Oh, dear :)
BUT Chris, it was made quite clear by our bus-driver (you remember,
who wanted to tear apart human bodies, you know, the driver with the
gasolining anomalia), that he was still willing to answer
*reasonable*, on topic stuff .........
Let's analyse what you had written:
[quote]
>> Admit it Ed, this is funny, no ? :))))))))))))))))))))))))))))))
>>
>>
>> : Chris Whittington
[quote end]
Isn't it fuuunnny??
And here is what the Alabama Killer Bob, you know, the one with
DrDeath Kevorkian in the cooperation pack, had written ----------
[quote]
>Why don't you stop acting like a jerk, and *carefully* read (b) above.
>Let me give you a hint:
>
> "except for posts that are "on topic" and reasonably "polite" and
> also "to the point"...
>Get it now? Maybe? Your post was reasonable, was about computer chess,
>and I responded.
[quote end from the gasoliner]
But I just have read another cute piece of impostordom. Read what the
gasoliner Bob confessed in confidential talks:
[quote on]
>> I can run on DOS, Win3.1, win95, winNT, linux, unix, HPUX, digital unix,
>> Cray unicos, IBM aix, digital VMS, IBM MVS, Amiga, Next, FreeBSD, and
>>others.
[quote end from the bus driving gasoliner]
ROTFL. Chris, I had expected all sorts of fun, but this guy is still
confusing me. Now we know that he can drive a bus. We know that he
wanted to gasoline many black people over there in the States. He can
run on WIN95, I do concentrate on the one I'm able to live on too.
But now the hype. This guy claims in public, thast he also can run on
"others". Chris, that made my day. This bastard is running on all of
us too ... That's the reason why I felt so weak these days. My
girl-friends already told me, "Rolf, you have become _so_ *weak*, stop
playing chess and those talks on usenet ...".
Master, what shall I do? Stopping to read the dangerous killer from
Alabama?? Playing more CSTal? Studying theology to become an
evangelical pastor like Frick.? Or digging my hole -- like Ed
Schroder? Founding a private computerchess group? Stealing Hyatt
source code and making a new "LoboR" (in the way of the little
<knee-shot>Kerrigan [born in 1978]?)
Master, tell me ...
In case of danger, I will also be glad to take part in the monster
movie Rolf vs. Two Greyhound Buses.
I will keep two buses near-by and prevent them from driving in
opposite directions. I will cause serious damages in the tires. Until
-- after 25 tire changes -- the show will be stopped by Ralph Nader.
I'm *game*.
Chris Whittington
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote in article
<6j6k78$5u9$1...@news02.btx.dtag.de>...
Finally I understand the address.
UA Bus Station.
Hey Rolf, how long are we going to keep this going for ...... ? Forever ?
Or do you have a goal in mind ?
I mean I'm *game*.
Kris Witty
>
>
Robert Hyatt
Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
: Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
: Oh, dear :)
: Well I just thought it might have reminded Ed of the 100 versions of Crafty
: rebuilds required to play one auto232 game last year .....
: I think he gave up in end, no ?
: BTW what happened to my entry in your kill file ? You promised.
: Chris Whittington
No I didn't so long as your posts were about chess. If you'd prefer to
have me ignore anything you say, simply say so. I can handle that...
The "100 versions" were the *first* attempt to get auto232 working. It
was not easy... Crafty "out of the box" works easily and well when used
to play against the person installing it. To get it up on ICC is another
level of difficulty since winboard has to be installed and set up as well.
Auto232 is the most painful of all, particularly since I can not do any
testing at all here, with only unix machines and winNT machines at my
disposal.
Robert Hyatt
Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
: Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
:>
:> No, but maybe a "certain poster" will simply stop trying to tell *me* how
:> *my* program is defectively designed. I see enough defects to go around
:> the list *twice*.
:>
:> Rather than telling you your approach is defective, I simply suggest that
:> you prove mine is?
:>
: That would not be possible. You're paradigm bound.
: Chris Whittington
Here's a "paradigm" for you: log on to ICC, type "match crafty r 60 120"
and play some games. There's no "paradigm" at stake there. When finished,
count the number of 0's and 1's. If you get more 1's, you made your point,
if not...
Chris Whittington
Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
<6j6t0e$5tb$3...@juniper.cis.uab.edu>...
> Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
>
>
> : Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
>
> : Oh, dear :)
>
> : Well I just thought it might have reminded Ed of the 100 versions of
Crafty
> : rebuilds required to play one auto232 game last year .....
>
> : I think he gave up in end, no ?
>
> : BTW what happened to my entry in your kill file ? You promised.
>
> : Chris Whittington
>
>
> No I didn't so long as your posts were about chess. If you'd prefer to
> have me ignore anything you say, simply say so. I can handle that...
>
This is what you said:
(2) There is another option. For this to work, *every* person here will
have
to subscribe to this plan. If you are interested, read on:
(a) immediately add rolf, chris and anonymous to your kill files.
(b) after doing this, *absolutely* do not respond to anything they write,
for *any* reason, on *any* topic. Simply behave as if they don't
exist,
except for posts that are "on topic" and reasonably "polite" and also
"to the point." This is the critical step. They must get *no*
responses
no matter *what* they say to provoke. Even one reply will keep the
problem alive.
You can't have a kill-file "except for posts that are "on topic" and
reasonably "polite" and also "to the point."."
That would be known as not having a kill-file, reading posts, and then
deciding.
So, I'm waiting for you to do me the honour of being in your kill-file. You
promised.
Chris Whittington
> The "100 versions" were the *first* attempt to get auto232 working. It
> was not easy... Crafty "out of the box" works easily and well when used
> to play against the person installing it. To get it up on ICC is another
> level of difficulty since winboard has to be installed and set up as
well.
> Auto232 is the most painful of all, particularly since I can not do any
> testing at all here, with only unix machines and winNT machines at my
> disposal.
>
Chris Whittington
Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
<6j6t63$5tb$4...@juniper.cis.uab.edu>...
> Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
>
>
> : Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
> :>
> :> No, but maybe a "certain poster" will simply stop trying to tell *me*
how
> :> *my* program is defectively designed. I see enough defects to go
around
> :> the list *twice*.
> :>
> :> Rather than telling you your approach is defective, I simply suggest
that
> :> you prove mine is?
> :>
>
> : That would not be possible. You're paradigm bound.
>
> : Chris Whittington
>
>
> Here's a "paradigm" for you: log on to ICC, type "match crafty r 60 120"
> and play some games. There's no "paradigm" at stake there. When
finished,
> count the number of 0's and 1's. If you get more 1's, you made your
point,
> if not...
Dear Mr Bean-Counter,
You promised to do me the honour of being in your kill-file.
Promised, yes ?
Chris Whittington
Robert Hyatt
Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
: Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
: Interestingly, only after Match 1. And presumably after being bollocked by
: IBM officials. I don't think the natural tendency in the DB team was to
: evaluate chess. It was to be fast.
Actually it was *both*... as has been mentioned by them many times. That
was the main reason for the redesign between match 1 and match 2.. to get
more "evaluation" into the hardware... according to Hsu...
: Yes, but Hyatt says he does this too. It's not so.
I said I have *done* this. And I did... first order pawn structure,
second order piece coordination, third order attacks around the king. And
I ran it for several months... and decided that it played better without it,
even after much tuning. I now do a primary and secondary analysis that is
much faster, and, so far, just as effective...
Fortunately, there are still many roads to the same destination...
Robert Hyatt
Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
: So, I'm waiting for you to do me the honour of being in your kill-file. You
: promised.
: Chris Whittington
Done...
caio...
Chris Whittington
Robert Hyatt <hy...@crafty.cis.uab.edu> wrote in article
<6j6tj5$5tb$5...@juniper.cis.uab.edu>...
> Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
>
>
> : Keith Ian Price <kpr...@spamfree.teleport.com> wrote in article
>
> : Interestingly, only after Match 1. And presumably after being bollocked
by
> : IBM officials. I don't think the natural tendency in the DB team was to
> : evaluate chess. It was to be fast.
>
> Actually it was *both*... as has been mentioned by them many times. That
> was the main reason for the redesign between match 1 and match 2.. to get
> more "evaluation" into the hardware... according to Hsu...
>
>
> : Yes, but Hyatt says he does this too. It's not so.
>
> I said I have *done* this. And I did... first order pawn structure,
> second order piece coordination, third order attacks around the king.
And
> I ran it for several months... and decided that it played better without
it,
> even after much tuning. I now do a primary and secondary analysis that
is
> much faster, and, so far, just as effective...
>
> Fortunately, there are still many roads to the same destination...
Hah, you won't read this :)))))))
No, it's not the same destination. You're climbing a different hill.
Four hills so far, I think, might be more, and might be a little random:
1. Test-suite hill. Almost all programs tried climbing this one, and then
got off, and went elsewhere.
2. SSDF hill. Almost all the programs are on this climb. But it's boring,
incestuous. You know the rest.
3. ICC hill. Crafty, Ferret, others. Better than the SSDF hill, but not by
much.
4. Let's have fun hill. That's my one.
See ? The whole process is a genetic algorithm, and there are lots of local
peaks. And nobody knows which is the highest one.
(1) works by trying to fit to some german third order differential equation
(2) and (3) work on 0's and 1's
(4) works on how orgasmic Thorsten gets
And who's right ?
Now don't answer this one Bob. You just kill-filed me, remember ?
Chris Whittington
Chris Whittington
Anonymous <nob...@REPLAY.COM> wrote in article
<1998051113...@basement.replay.com>...
> >So, I'm waiting for you to do me the honour of being in your kill-file.
You
> >promised.
> >
> >Chris Whittington
>
>
> TheDoDo says:
>
> Hey Chris, put yourself in your kill-file.
>
Sorry, I don't have the technology. Not for me, nor for anyone else. Don't
need it.
Chris Whittington
Peter Coleman
A mispelling of the word WAFFLE I presume?!
Keith Ian Price wrote in message ...
Rolf Tueschen
"Chris Whittington" <chr...@cpsoft.demon.co.uk> wrote:
Chris, he meant a hangover, so to speak. I also asked myself if you
were crying about a past love affair somehow...
>Chris Whittington
>>
>>
>>
>>
>>
>>
>>
>>
Komputer Korner
I thought Deeper Blue was a full width program that did singular
extension. If it really is a selective searcher to 30 plies with a
minimum full width of (who knows what number but less than 30
obviously) and along with that, singular extension of 70 plies
as well, then the knowledge guys have won the war.
--
--
Komputer Korner
The inkompetent komputer
To send email take the 1 out of my address. My email address is
kor...@netcom.ca but take the 1 out before sending the email.
Robert Hyatt wrote in message <6j4e19$2cf$2...@juniper.cis.uab.edu>...
>"30 plies" means 30 plies *selectively*... *not* full-width. They do
>this by extending what looks interesting, not extending what doesn't.
>There's no guarantee that what they think is uninteresting really is,
>so "solved" is way too big a stretch..
Chris Whittington
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote in article
<6j7u26$a15$1...@news01.btx.dtag.de>...
I don't think it was ever that ...... :)
Chris Whittington
>
> >Chris Whittington
>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
> >>
>
>
>
Robert Hyatt
Komputer Korner <kor...@netcom.ca> wrote:
: I thought Deeper Blue was a full width program that did singular
: extension. If it really is a selective searcher to 30 plies with a
: minimum full width of (who knows what number but less than 30
: obviously) and along with that, singular extension of 70 plies
: as well, then the knowledge guys have won the war.
Don't mix concepts. Singular extensions is a form of selective
search. In "classic selective search" some branches are tossed out
early so that others may be searched more deeply. In the other approach,
we simply follow "interesting" branches deeper, which means that we don't
search *every* move as deep as we did before. IE DB should easily be
able to hit 20 plies with a program approach like crafty. They have
elected to search only 10-12 plies along *all* branches so that they
can search much deeper along "interesting" lines... which is why Hsu
says they "average" 30 plies... many branches are stopped at 10-12,
but many go well beyond 30...
: --
Rolf Tueschen
"Chris Whittington" <chr...@cpsoft.demon.co.uk> wrote:
What was it?
Because I could only see someone who sold his most worthyful argument
under price ...
Keep in mind what a third rate milestone software stealing bean
counter without any own genuine contributions to computer chess you've
talked to ...
You, if I could judge that as a newbie, as *the* expert, the only one?
who favoured the so called knowledge path.
Moreland is a little proxy with no own standpoints, with no respect
for your dignity. More. With no respect for human dignity at all. With
the exception of his own probably. Hyatt non plus, but at least he's
micro bean counter number one ... Former macro bean counter of course.
He defines the paroles for the many proxies ... He's the wild / over
the edge one. Let's not fritter away! :)
>Chris Whittington
>>
>> >Chris Whittington
>>
bruce moreland
>No, it's not the same destination. You're climbing a different hill.
>
>Four hills so far, I think, might be more, and might be a little random:
>
>1. Test-suite hill. Almost all programs tried climbing this one, and then
>got off, and went elsewhere.
>
>2. SSDF hill. Almost all the programs are on this climb. But it's boring,
>incestuous. You know the rest.
>
>3. ICC hill. Crafty, Ferret, others. Better than the SSDF hill, but not by
>much.
>
>4. Let's have fun hill. That's my one.
The first games my program played, against any entity other than me,
were against people on ICC. I found out about ICC in September 1994,
made an account, started playing my program there, and became hooked.
For the following three months I made no changes to my program, I just
watched it play.
I certainly didn't start doing this as some sort of substitution for
the SSDF list.
I don't make changes to my program based upon its ICC rating. The
same version can roam around in a 300 point rating range, it is
useless to try to make any decisions based upon ICC rating.
It is interesting to listen to people ask me if I've made changes to
my program when it bottoms out at 2850 blitz, and then ask me again
when it hits 3120. Same version since April 2nd.
There are plenty of beans there to count, but it doesn't make sense to
count any of them, everything is pretty random.
You can get an idea of how you stack up, though, against the general
pool of people and programs who play there. You can do some
experiments fairly easy, too, for instance I wanted to know if my
program could still compete on a P6/200, so I made a new account and
ran a few hundred games through it.
You can also meet interesting people. It's where I met Bob, Martin
Zentner, Stefan Meyer-Kahlen, Tom Kerrigan, Dave Slate, John Stanback,
Mark Lefler, Peter McKenzie, and dozens of others who don't write
programs but who are very interested in this field. When I got off
the plane in the '95 Paderborn WMCCC, I had friends already.
You can do chess, too. Sometimes I get to do analysis with people
such as GM Tisdall (DB-Kasparov, 2nd match, game 2), I get to meet
interesting titled players such as IM Schroer and IM Borris, and today
I had a nice conversation with GM Speelman.
I don't get as many GM games as Bob does, but I get a lot.
Here is one of the strangest ways I have seen to draw a won ending.
This game was played today against Crafty as part of a 14-game 5 0
match. I think simply 42. gxf8=Q+ is +-, but I underpromoted because
I thought that black had to take the piece anyway.
[Event "ICC 5 0 05/12/1998"]
[Site "Internet Chess Club"]
[Date "1998.05.12"]
[Round "-"]
[White "Mink"]
[Black "crafty"]
[Result "1/2-1/2"]
[WhiteElo "2913"]
[BlackElo "2915"]
[ECO "C92"]
[NIC "RL.26"]
[LongECO "Ruy Lopez: closed, Flohr-Zaitsev system (Lenzerheide
variation)"]
1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6
5. O-O Be7 6. Re1 b5 7. Bb3 d6 8. c3 O-O
9. h3 Bb7 10. d4 Re8 11. Nbd2 Bf8 12. a4 h6
13. Bc2 exd4 14. cxd4 Nb4 15. Bb1 c5 16. d5 Nd7
17. Ra3 f5 18. exf5 Nf6 19. axb5 axb5 20. Ne4 Bxd5
21. Rxa8 Qxa8 22. Nxf6+ gxf6 23. Rxe8 Qxe8 24. Nh4 Bc4
25. Qg4+ Kh7 26. Bd2 Bd3 27. Bxd3 Nxd3 28. Bc3 Bg7
29. Qg3 Ne5 30. f4 Nc6 31. Qg6+ Qxg6 32. fxg6+ Kh8
33. Nf5 d5 34. Kf2 b4 35. Bd2 d4 36. Kf3 c4
37. Ke2 b3 38. Kd1 Bf8 39. g4 Kg8 40. h4 Ne7
41. g7 Nxf5 42. gxf8=B Nxh4 43. Bxh6 Nf3 44. Bb4 Kf7
45. g5 fxg5 46. Bxg5 d3 47. Bc3 Ke6 48. Bg7 Kf5
49. B7f6 Ke4 50. Bc3 Kd5 51. Bd8 Ke4 52. Bc7 Nh4
53. Kd2 Nf3+ 54. Kd1 Ng1 55. B3e5 Nf3 56. Bb8 Kf5
57. Bh8 Ke4 58. Bhe5 Nh4 59. Bbd6 Nf3 60. Bc3 Kd5
61. Bc7 Ke4 62. Bb8 Ng1 63. Bce5 Nf3 64. Bbd6 Kf5
65. Bc7 Ke4 66. Bf6 Ke3 67. Bd6 Ke4 68. Bfe5 Nh4
69. Bb8 Ke3 70. Bf6 Nf3 71. Bc3 Ke4 72. Bbe5 Kf5
73. Bd6 Ke4 74. Bb8 Kf5 75. Bce5 Ke4 76. Bbc7 Nh4
77. Bcd6 Nf3 78. Bc7 Nh4 79. Ba5 Ke3 80. Bd8 Nf3
81. Bg5 Ke4 82. Bgf6 Ke3 83. Bd6 Ke4 84. Bc7 Kf5
85. Bc3 Ke4 86. B7e5 Kd5 87. Bf6 Ke4 88. Bce5 Ke3
89. Bc7 Ke4 90. Bb8 Ke3 91. Bc3 Ke4 92. Bg7 Nh4
93. f5 Nf3 94. Bc7 Kxf5 95. Bb6 Ke4 96. Bc5 Ng5
97. Be7 Ne6 98. Bgf6 Nf4 99. Bd6 Nd5 100. Bde5 Ne3+
101. Ke1 Ng4 102. Bd4 Nh2 103. Kd1 Nf3 104. Bf2 Nh2
105. Bc5 Kd5 106. Bcd4 Nf3 107. Be3 Kd6 108. Bc3 Ke6
109. Bh8 Kf5 110. Bc3 Ke4 111. Bf2 Nh2 112. Bc5 Kd5
113. Be3 Nf3 114. Bb4 Ke4 115. Bf2 Ne5 116. Bc3 Nf3
117. Bg7 Ng5 118. Bf6 Nh3 119. Bg3 Ng1 120. Kd2 Nf3+
121. Kc1 Ke3 122. Kd1 Ng1 123. Bd6 Ne2 124. Bde5 Ke4
125. Kd2 Ng1 126. Bc7 Nf3+ 127. Kd1 Kf5 128. Bg7 Kg5
129. Bb6 Kg6 130. Bc3 Kf5 131. Bc5 Ng5 132. Be3 Ne6
133. Kd2 Ke4 134. Bh8 Nc7 135. Bh6 Nd5 136. Bc3 Kf3
137. Be5 Ke4 138. Bc3 Nb6 139. Bg5 Nd5 140. Bg7 Kf5
141. Be3 Ke4 142. Beh6 Nb6 143. Bc3 Na4 144. Bhg7 Nc5
{Game drawn by the 50 move rule} 1/2-1/2
If that isn't fun, I don't know what is.
bruce
Komputer Korner
I didn't mix concepts. From the way you describe, Deep Blue's
selectivity is a far different program than Deep Thought's brute force
+ singular extension. It looks like the knowledge guys have won and
selective search is the way to go especially if you are fast enough or
the time controls are long enough.
--
--
Komputer Korner
The inkompetent komputer
To send email take the 1 out of my address. My email address is
kor...@netcom.ca but take the 1 out before sending the email.
Robert Hyatt wrote in message <6j9dov$8o7$1...@juniper.cis.uab.edu>...
Robert Hyatt
Komputer Korner <kor...@netcom.ca> wrote:
: I didn't mix concepts. From the way you describe, Deep Blue's
: selectivity is a far different program than Deep Thought's brute force
: + singular extension. It looks like the knowledge guys have won and
: selective search is the way to go especially if you are fast enough or
: the time controls are long enough.
Nope... DB uses the same sort of extensions as DT... they have some
other chess-specific extensions as well... but still brute-force to
some depth, plus numerous extensions along selected lines... just like
always...
Komputer Korner
I remember a quote from Hsu that said that "brute force was the
program never having to say you are sorry"
This must have mislead us all into thinking that Deep Thought was
basically a brute forcer with singular extensions , when in fact what
he meant was brute force to a certain % of the search. nowadays there
are no real strictly selective searches left. they all brute force to
some % with the rest of the % being selectivity. Either the above is
true or else Deep Blue was more selective than Deep Thought . The
knowledge guys have won!!
--
--
Komputer Korner
The inkompetent komputer
To send email take the 1 out of my address. My email address is
kor...@netcom.ca but take the 1 out before sending the email.
Robert Hyatt wrote in message <6jb5kf$vp$1...@juniper.cis.uab.edu>...
Chris Whittington
Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote in article
<6j98p7$fom$3...@news01.btx.dtag.de>...
Perhaps a little unfair.
However, Bruce's output, or more especially, Bruce's non-output interests
me.
What he doesn't say is usually much more interesting. Or what he says and
then leaves unsaid.
'dislike a lot'
'either him or me'
He says, and then he stops, short. Very strange.
I first got this taste of the strong reaction in Paris. Couple of things
never revealed before; in the said and not said mode:
1. There were seven of us, self-selected to be the
icca-programmers-advisory group or whatever it's called. Myself and Bruce
included. We were standing around mulling over who was to be
spokesman/chairman. Various people disqualified themselves quickly. I said
nothing. And then Bruce said: "I won't do it, and neither should Chris, he
has too much power already".
2. Also, I took my 14 year old son to the tourney in Paris. He met various
people, had a good time. After the event he told me that Bruce had come up
to him, and said: "hi, I'm Bruce, I expect your father has said a lot of
bad things about me, but I want you to know they're not true".
(btw a little aside for Bruce: hadn't mentioned you :)
Interesting. Well I think so :) There seems to be this very cool, logical,
rational person, of few words, but with very strong negative emotional
reaction, occasionally bursting into reality through some thin surface. Or
better described, as external reality burting *in* through the thin
surface.
dislike
either him or me
too much power
says 'bad things about me'
I guess if it wasn't on target lock-on chrisw, I'ld probably be less
inclined to ponder on it.
Or if I didn't see Bruce as archetypical of the programmer community. Even
as part of the testing community.
Bruce has his answer already; he's given it before: "don't try and
psychoanalyse me". I remember that well.
Now we operate here in a kind of two-dimensional world. By definition. It's
the medium. What 'we' do is more one-dimensional. Computer chess. But the
one-dimensionalty of it is inherently unsatisfying to some. viz. Fernando
on ccc. There are others.
Some find the two-dimensionality unsatisfactory. It leads too easily to
box-ing people. Hyatt=fascist. Tueschen=madman. Bruce=emotional keg of
Dynamite. Chrisw=scum (did I get Hyatt's box-ing correct ?). So we can all
happily hate/censor/whatever each other in the easy 2-D world.
And some threaten to burst through the veneer, into three dimensions. How
badly you want to stay in 1-D or 2-D will colour, entirely, your response
to the 3-D threat.
So be a man, take your responsibilities, run and hide, Jakarta, Tueschen,
private 2D worlds. Is what it is about. 3-D.
Chris Whittington
>
> >Chris Whittington
>
> >>
> >> >Chris Whittington
> >>
>
>
>
Robert Hyatt
Komputer Korner <kor...@netcom.ca> wrote:
: I remember a quote from Hsu that said that "brute force was the
: program never having to say you are sorry"
: This must have mislead us all into thinking that Deep Thought was
: basically a brute forcer with singular extensions , when in fact what
: he meant was brute force to a certain % of the search. nowadays there
: are no real strictly selective searches left. they all brute force to
: some % with the rest of the % being selectivity. Either the above is
: true or else Deep Blue was more selective than Deep Thought . The
: knowledge guys have won!!
: --
: --
This has *nothing* to do with "knowledge". Otherwise, programs have
been "knowledge-based" since the late 70's... and don't forget that
singular extensions were publicly explained in the late 80's although
they were tried earlier...
But it definitely (this part of the search) has nothing to do with a
"knowledge issue."
Komputer Korner
Bob, you are taking me too literally. If the selective searchers
didn't have more knowledge than the brute forcers they would get
killed. They wouldn't be able to tell which lines weren't worth
looking at. Sure there are middle of the road programs like Crafty
with a fair amount of knowledge and little selectivity , but in
general selective search
programs have huge amounts of knowledge built in. You are right that
knowledge doesn't exactly equal selective search as a straight line
relationship , but the whole point of discarding lines is to save
time for searching the most important lines and for that you need lots
of knowledge.
--
--
Komputer Korner
The inkompetent komputer
To send email take the 1 out of my address. My email address is
kor...@netcom.ca but take the 1 out before sending the email.
Robert Hyatt wrote in message <6jc7o3$fo5$1...@juniper.cis.uab.edu>...
Rolf Tueschen
"Chris Whittington" <chr...@cpsoft.demon.co.uk> wrote:
>Perhaps a little unfair.
This is sort of always righteous phrase. Perhaps on behalf of Moreland
I should add that all I've said was on the base of my experience with
his online existence. But also on the told story of Paris about his
protest against the conditions of pitters' participation. Pitters was
handicapped at the time.
However, if you had my experience, you'd understand why my verdict is
on to the point. The very point of Bruce. And not knowing my special
reasoning you gave a little proof for exactly my point. See below.
>However, Bruce's output, or more especially, Bruce's non-output interests
>me.
I'm interested in all sorts of output. Always keeping in mind that
usenet is NOT the whole truth or reality.
>What he doesn't say is usually much more interesting. Or what he says and
>then leaves unsaid.
My point. I had the like ideas in mind right from the beginning. But
that's basically a trivial point. It's always interesting to notice
the different key-words or -topics or -persons that provoke someone to
interfere (or intervene?). Excuse my weak English.
Then it's also interesting to see the way someone jumps in -- here
also depending on topic and person.
Well, it provides you with a mass of important data.
Finally you also have to keep an eye on all sorts of fake and jokes
and anonymity. Of course I would say that the ones who only play games
(fakes and jokes) in a chess group are the ones to pity. But usenet is
usenet, also for chess groups.
Anonymity is allowed in case someone has to oppose a non-democratic
majority. More so if he/she's a famous person. I for one did never use
such hidance. With the apparent result as a scapegoat. But as
psychologist it's exactly what is the most interesting. And this way
others are safe who could be shattered more easily.
>'dislike a lot'
>'either him or me'
>He says, and then he stops, short. Very strange.
Not at all. Here's my simple explanation/guess. :)
He simply has misunderstood what he'd seen on the screen when these
proxy press officers have to tell something without saying too much.
And believe me. *If* Bruce is right on to the point, then his style is
very successful. But, alas, this style is also a great restriction. It
leads to a certain one-dimensional picture he's painting about
himself. But we all have to live with pictures. Therefore I stated
that I for one kept in mind the difference between usenet and real
life. (The boundary could also be taken away. That is what is
happening right now with the funny and disgustful tries of the
gasoliner Fach-Idiot. This will lead to his own defeat by definition.
Because how could a gasoliner pp. claim to be disturbed somehow? When
he's the disturbance as such?)
>I first got this taste of the strong reaction in Paris. Couple of things
>never revealed before; in the said and not said mode:
>1. There were seven of us, self-selected to be the
>icca-programmers-advisory group or whatever it's called. Myself and Bruce
>included. We were standing around mulling over who was to be
>spokesman/chairman. Various people disqualified themselves quickly. I said
>nothing. And then Bruce said: "I won't do it, and neither should Chris, he
>has too much power already".
Keeping in mind the big difference in your age, I wouldn't see too
much of importance here. I would always prefere a student who tried to
to challenge me. Instead of a brown-noser.
>2. Also, I took my 14 year old son to the tourney in Paris. He met various
>people, had a good time. After the event he told me that Bruce had come up
>to him, and said: "hi, I'm Bruce, I expect your father has said a lot of
>bad things about me, but I want you to know they're not true".
>(btw a little aside for Bruce: hadn't mentioned you :)
The latter gives some hope. Lets see a very important quality of BM.
Certain class. On the other hand it's somewhat arrogant too.
But the other story. All depends on the style he had said tthat. You
see. It's not possible to talk about all aspects of a human and his
behavior at the same time on usenet level. I would NEVER comment too
much on the "told" superficial story alone. Because the story teller
might have overseen a certain aspect that would - for me - change the
analysis completely into the opposite. Because the problem you have
here, on one side a 14 y. old youngster and on the other a still very
young technical expert. What to say at all in such a situation.
Basically Bruce had told his statement to the false adressee. Probably
he took your son as sort of under-state-secretary of some sort or
nation. :))
It's true, Bruce is unexperienced. What do you think did Kohl say to
Gorbatchov -- after he had denounced the Russian No One as sort of
second Goebbels ... ;-)
To be relaxed requires perhaps more experience than B. could ever
present at his age.
And he's an American. Horrido type of guy, if I'm not totally false ..
I remember the time when he was home back from Jakarta. I hadn't read
a word from him before. And surely he couldn't know anything about me.
But he still jumped in ands explained me how we should take the German
KZ past and how we should not work it into debates about gasolining
deathrow people's bodies ... :) <cynical mode on>
And that was already the crucial moment between us. I could easily
correct him. He took it like a man ... I'd thought. But then it was
clear that he prepared his come back "against" me and I gave him a lot
of possibilities. Just to see if my verdict right from the beginning
was justified. That he was a very uneducated but typically American
cowboy -- completrely innocent about the difficulties of a lot of
things he probably never heard of before ...
Well, that's a base for life-long hate and competition. And time alone
is also not the assurance for approvement as the case Hyatt proves.
>Interesting. Well I think so :) There seems to be this very cool, logical,
>rational person, of few words, but with very strong negative emotional
>reaction, occasionally bursting into reality through some thin surface. Or
>better described, as external reality burting *in* through the thin
>surface.
Ok, we all have something that will be bursting through the surface
after some time. With me it's the always ready commenting about
atrocities or stupid reasonings in off-topics of our experts.
Especially in fields where I was clearly the victim. As in the
Schroder character assassination. He insulted me as a pig, then sued
me then slipped away into an asylum like private censored group/club.
But still his stuff is debated here, he makes announcements for his
company and I assure you, he reads still this group ... So, I also
want to comment on Ed Schroder a little bit longer...
>dislike
>either him or me
>too much power
>says 'bad things about me'
>I guess if it wasn't on target lock-on chrisw, I'ld probably be less
>inclined to ponder on it.
Yes, that's probably the most important difference between the two of
us.
I give my statement, also if I`m not at all involved personally.
With the clear stress on the topics and not my personal dislikes.
And I think I should have a right to do so, if I don't push such
opinions for my own advantages. On the contrary, I get 'hell' more
than enough.
The very illogical and stupid variation of Hyatt is to believe that he
as a real expert of computerchess should have a right to set the
conditions for others like me.
His stupid stuff of "it's in your hands", "it's up to you" alone is a
farce. It sounds as if he, the big expert hat to give something of a
granting or such. No. This is a normal user like you and me, and Hyatt
has not to write such a nonsense. Because what lies in my hands, that
is my thing. And I know quite well, what is my thing here on usenet.
Or it's up to me. Fine. I fell real relaxation to know that at least.
That I have something to decide. I really was completely innocent
about it -- before.
No, this is all very stupid. And also with my 23 y. I can judge that
Hyatt has no rights whatsoever to mute or censor me. And he will NOT
succeed to reach that goeal with my server. Even if he engaged an
armada of some thousand engaged helpers who sent copies of my postings
to Germany ...
It will end in the same way, as Ed Schroder had to bite the dust. I
tell you, I'm not proud of it. On the contrary I feel ashamed that I
had to show Bob in such a disatvantaged light. I didn't want it. It
was Bob who forced the development. And I see a reason why it could
help all of us in the end. Because this sort of mambo hoolup can't
succeed. Must not succeed. A man, a two time Wch, propagates here
atrocities like gasoling and tearing apart bodies with two buses ...
and then is also the sort of judge here for all of us who decides what
people like me should be allowed to post and what not. This is against
democracy, against the American constitution, against the German Basic
Law, against human decency, against logic an --- well, excuse me if I
forgot something important.
>Or if I didn't see Bruce as archetypical of the programmer community. Even
>as part of the testing community.
>Bruce has his answer already; he's given it before: "don't try and
>psychoanalyse me". I remember that well.
>Now we operate here in a kind of two-dimensional world. By definition. It's
>the medium. What 'we' do is more one-dimensional. Computer chess. But the
>one-dimensionalty of it is inherently unsatisfying to some. viz. Fernando
>on ccc. There are others.
Let me go farther. I'm sure that all of us here are smart enough to be
able to understand all this to a certain extent. Not everybody can be
a philosoph or sociologist or whatever. But that all this is here only
parts of a whole entity -- that is quite clear, no?
Although in war-like quarrels, its difficult to keep that in mind.
It's easier to reduce someone on certain few aspects. Thus you simply
have a simpler job to do. I agree I often observed the like.
And especially the scandal that verdicts like "you're mentally ill",
"you should seek help" could happen is very bad for our group. But if
it hits a psychologist it's not so bad at all. At least I could
understand the mechanism of the like stupidities ...
>Some find the two-dimensionality unsatisfactory. It leads too easily to
>box-ing people. Hyatt=fascist.
NB!! This is NOT true. I did never reduce Hyatt on this simple point.
The scandal is that Bob has this aspect. At least in his postings here
on rgcc. But Bob was NEVER reduced on the fascist part alone.
>Tueschen=madman.
Don't worry. I never took it as a personal offense. It clearly told me
more about the authors themselves.
It came out of being angry or upset. And I would aggree that I have
posted stuff that had to make people angry. Especially uneducated
people. But I wouls always repeat it. Because some are still there who
understand it better. Dejas contains all posts. And let's later
generations (! :)) decide who was wrong and who was right.
NB!! Hyatt defined me also as the most prominent Nazi worshipper in
the whole world ... ;-) or :))
>Bruce=emotional keg of
>Dynamite. Chrisw=scum (did I get Hyatt's box-ing correct ?). So we can all
>happily hate/censor/whatever each other in the easy 2-D world.
>And some threaten to burst through the veneer, into three dimensions. How
>badly you want to stay in 1-D or 2-D will colour, entirely, your response
>to the 3-D threat.
I can't understand this due to my English. But if you meant, what Phil
also mentioned, that I had also better things to do than bring the
sword for decisions ... and that I would stay very restrictedly on a
certain level if I would do this or that ...
Well. This is not true. I could reach another performance if I would
also use different personalities ... But I dont do this. So. Of
course, it's a triviality. If I defend me in a war against fascist
censorship, this is predefining my position. Yes, but I never said
that I were Jesus II. I'm really more on the Old Testament. This has
to be decided. An eye for an eye. And what Ed Schroder and helpers did
against me will cost them all some of their reputation.
But read my proposals. From the beginning on I wrote about peace. It
was also you who declined this becuase I were sort of Hitler ... You
have written this. And you had possible influenced Ed to the worse.
Therefore I still hope that you migfht help to find a peace. Also
Hyatt could have a peaceful Rolf. But NOT on the base of atrocities
and the verdict that I were the greastest sympathisant of the Nazis ..
>So be a man, take your responsibilities, run and hide, Jakarta, Tueschen,
>private 2D worlds. Is what it is about. 3-D.
I see. I agree. And I had not to answer for myself. Ok, but hopefully
I could give some explanations. The point is, that some here are
unable to understand what we mean, Chris. I receive emails with just
two or three words. "If I were completely mad or what??" Funny if the
poster didn't have a clue of what I was talking about ...
BTW I want to clarify that the mere word Fach-Idiot is in my
unsderstanding not a simple insult. It's the short key word for sort
of deeper analysis. And Bob Hyatt is probably the greatest example for
rgcc. I'm still sure that Bob could understand the meaning. And the
moment he starts to grin for himself -- to start with -- then we could
seek direct solutions ... But if I see how Bob gave his verdict about
Kasparov, a certain probability is there that he might NOT be able to
realize the foolish things he had done in our group.
Your private Pope, Chris, -- Rudolpherus XXIII.
Robert Hyatt
Komputer Korner <kor...@netcom.ca> wrote:
: Bob, you are taking me too literally. If the selective searchers
: didn't have more knowledge than the brute forcers they would get
: killed. They wouldn't be able to tell which lines weren't worth
: looking at. Sure there are middle of the road programs like Crafty
: with a fair amount of knowledge and little selectivity , but in
: general selective search
: programs have huge amounts of knowledge built in. You are right that
: knowledge doesn't exactly equal selective search as a straight line
: relationship , but the whole point of discarding lines is to save
: time for searching the most important lines and for that you need lots
: of knowledge.
Maybe I didn't understand your comment. I took it as "Deep Blue" is
showing that a "knowledge based selective searcher" can do well. DB
has a lot of knowledge... but most of the extensions are extensions
in the forward direction, rather than pruning as the selective
programs are doing... If you prune beyond 7 so you can reach 12, you
miss lots of things beyond 7. If you extend beyond 12 so you can
reach 30 (assuming enough horsepower, of course) then you miss nothing
within 12, and you see *lots* beyond 12... like DB
bruce moreland
On Wed, 13 May 1998 09:51:04 GMT, "Chris Whittington"
<chr...@cpsoft.demon.co.uk> wrote:
>
>
>Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote in article
><6j98p7$fom$3...@news01.btx.dtag.de>...
>> What was it?
>>
>> Because I could only see someone who sold his most worthyful argument
>> under price ...
>>
>> Keep in mind what a third rate milestone software stealing bean
>> counter without any own genuine contributions to computer chess you've
>> talked to ...
>>
>> You, if I could judge that as a newbie, as *the* expert, the only one?
>> who favoured the so called knowledge path.
>>
>> Moreland is a little proxy with no own standpoints, with no respect
>> for your dignity. More. With no respect for human dignity at all. With
>> the exception of his own probably. Hyatt non plus, but at least he's
>> micro bean counter number one ... Former macro bean counter of course.
>> He defines the paroles for the many proxies ... He's the wild / over
>> the edge one. Let's not fritter away! :)
>
>Perhaps a little unfair.
Oh, just a tad.
>However, Bruce's output, or more especially, Bruce's non-output interests
>me.
>
>What he doesn't say is usually much more interesting. Or what he says and
>then leaves unsaid.
>
>'dislike a lot'
>
>'either him or me'
>
>He says, and then he stops, short. Very strange.
I don't want to waste time, bandwidth, and energy, helping you advance
your agenda, which is in no way constructive. Your threads benefit
nobody, and any significant contributions on my part just add to the
general decay and give you more material you can use to generate
rhetoric.
I figure that you have enough history here that people can understand
your deal without my having to point it out.
bruce
mclane
"Chris Whittington" <chr...@cpsoft.demon.co.uk> wrote:
>I first got this taste of the strong reaction in Paris. Couple of things
>never revealed before; in the said and not said mode:
>1. There were seven of us, self-selected to be the
>icca-programmers-advisory group or whatever it's called. Myself and Bruce
>included. We were standing around mulling over who was to be
>spokesman/chairman. Various people disqualified themselves quickly. I said
>nothing. And then Bruce said: "I won't do it, and neither should Chris, he
>has too much power already".
>2. Also, I took my 14 year old son to the tourney in Paris. He met various
>people, had a good time. After the event he told me that Bruce had come up
>to him, and said: "hi, I'm Bruce, I expect your father has said a lot of
>bad things about me, but I want you to know they're not true".
>(btw a little aside for Bruce: hadn't mentioned you :)
>Interesting. Well I think so :) There seems to be this very cool, logical,
>rational person, of few words, but with very strong negative emotional
>reaction, occasionally bursting into reality through some thin surface. Or
>better described, as external reality burting *in* through the thin
>surface.
>Chris Whittington
I don't think bruce is rational. He is like a ninja-turtle. Outside
rational. But inside very emotional. Those people always try to react
very cool and rational. But in the first moment they have to decide
something with normal-rationality they show that their inner parts are
out of balance.
I guess your experiences with him shows this.
I have made similar experiences with him.
So - no matter if he is vulcan or romulan or klingon, if he is like
spock or like data, whatever - he is bruce. And i like BRUCE.
best wishes
mclane
Chris Whittington
bruce moreland <bru...@seanet.com> wrote in article
<355c0862....@news.seanet.com>...
> On Wed, 13 May 1998 09:51:04 GMT, "Chris Whittington"
> <chr...@cpsoft.demon.co.uk> wrote:
>
> >
> >
> >Rolf Tueschen <TUESCHEN.MEDIZ...@t-online.de> wrote in
article
> ><6j98p7$fom$3...@news01.btx.dtag.de>...
>
> >> What was it?
> >>
> >> Because I could only see someone who sold his most worthyful argument
> >> under price ...
> >>
> >> Keep in mind what a third rate milestone software stealing bean
> >> counter without any own genuine contributions to computer chess you've
> >> talked to ...
> >>
> >> You, if I could judge that as a newbie, as *the* expert, the only one?
> >> who favoured the so called knowledge path.
> >>
> >> Moreland is a little proxy with no own standpoints, with no respect
> >> for your dignity. More. With no respect for human dignity at all. With
> >> the exception of his own probably. Hyatt non plus, but at least he's
> >> micro bean counter number one ... Former macro bean counter of course.
> >> He defines the paroles for the many proxies ... He's the wild / over
> >> the edge one. Let's not fritter away! :)
> >
> >Perhaps a little unfair.
>
> Oh, just a tad.
>
> >However, Bruce's output, or more especially, Bruce's non-output
interests
> >me.
> >
> >What he doesn't say is usually much more interesting. Or what he says
and
> >then leaves unsaid.
> >
> >'dislike a lot'
> >
> >'either him or me'
> >
> >He says, and then he stops, short. Very strange.
>
> I don't want to waste time, bandwidth, and energy, helping you advance
> your agenda, which is in no way constructive.
No, you just like to do that from the safety of the censored group.
Quite why someone so cool and rational as yourself wishes to keep bringing
my name up there, informing the world (well, those allowed to read) of your
personal likes and dislikes; or your power struggle mentality of 'him or
me' escapes me.
Or, it doesn't escape me really, you do it because it's important for you.
> Your threads benefit
> nobody, and any significant contributions on my part just add to the
> general decay and give you more material you can use to generate
> rhetoric.
Now then Bruce. You imagine that making attacks on a censored person from
within the safety of your censored news group; and such the the censored
person has no right of reply on the censored news group, is an isolateable,
one-dimensional act over which you retain total control ? Do you imagine
that ?
>
> I figure that you have enough history here that people can understand
> your deal without my having to point it out.
And you Bruce :)
Chris Whittington
>
> bruce
>
>