Is Belka a Rybka?

[GenoM]

I think that his reason was to prove that rybka is based on fruit's ideas
and to prove that Rybka gives wrong information about things like nodes per second.

Uri

Now I am really curious. Does this mean that Rybka is illegal (either by using GPL source or by disassembling code? Does this mean that Rajlich tried to conceal illegal content?

I don't know what was V.Rajlich trying to hide but iT's interresting that CCRL doesn't have any data about ponder-hit matches corelation between Rybka 1.0 and Fruit 2.1.
Or I can't find it? May be Graham or Kirr would like to say something about it?

Regards,
Geno

Fruit 2.1 did not play against Rybka 1.0.

Nice concise answer.
As I understand it Fruit 2.1 just denied to play vs such a rookie as Rybka 1.0?

[Tony Thomas]

I think that his reason was to prove that rybka is based on fruit's ideas
and to prove that Rybka gives wrong information about things like nodes per second.

Uri

Now I am really curious. Does this mean that Rybka is illegal (either by using GPL source or by disassembling code? Does this mean that Rajlich tried to conceal illegal content?

I don't know what was V.Rajlich trying to hide but iT's interresting that CCRL doesn't have any data about ponder-hit matches corelation between Rybka 1.0 and Fruit 2.1.
Or I can't find it? May be Graham or Kirr would like to say something about it?

Regards,
Geno

Fruit 2.1 did not play against Rybka 1.0.

Nice concise answer.
As I understand it Fruit 2.1 just denied to play vs such a rookie as Rybka 1.0?

At the time of the release wasnt there Fruit 2.2.1? Why would anyone in their right mind waste time on Fruit 2.1 vs Rybka??

[GenoM]

At the time of the release wasnt there Fruit 2.2.1? Why would anyone in their right mind waste time on Fruit 2.1 vs Rybka??

May be just to see which is the strongest free engine? What about such idea, eh?

[mjlef]

NOW from the early 1990's certainly had a bonus for a bishop on the long diagonal in it's piece-square tables. I am sure others did as well.

[Uri Blass]

NOW from the early 1990's certainly had a bonus for a bishop on the long diagonal in it's piece-square tables. I am sure others did as well.

I do not understand the logic behind that bonus and why to think about a bonus for this specific diagnol.

As a chess player I never learned that it is better for a bishop to be in the long diagnol.

I know that you may need a bishop at g2 in case that there are pawns h2 g3 f2 but this is not a reason to have a general bonus for the long diagnol
that can also encourage the bishop to be there when there is no reason to be there in order to protect the king.

Uri

[Tony Thomas]

At the time of the release wasnt there Fruit 2.2.1? Why would anyone in their right mind waste time on Fruit 2.1 vs Rybka??

May be just to see which is the strongest free engine? What about such idea, eh?

Rybka was so strong when it was released, everyone was trying to play it against the strongest engines they had. Fruit 2.2.1, and later released Zappa Paderborn on 4 CPU etc. They probably didnt see a need for a Fruit 2.1 vs Rybka game. If you are so interested why dont you ask CCRL if you can send them 30 games adapted to their time control? Rybka was definitley not a clone, it might have used ideas, but not a clone.

[John Conway]

Thanks.

Is there a good set of the parameters that does pretty well. All zeros must be pretty bad!

Mark

I found the following settings on a well known site, the name of which I am not permitted to mention.

SideToMoveBonus 3
PawnFileOpening 35
PawnFileEndgame 5
KnightCentrOpening 150
KnightCentrEndgame 40
KnightRankOpening 50
KnightTrapped 1000
BishopCentrOpening 40
BishopCentrEndgame 50
BishopBackRankOpening 120
RookFileOpening 130
QueenCentrOpening 20
QueenCentrEndgame 120
QueenBackRankOpening 40
KingFileOpening 120
KingCentrEndgame 90
PawnFile -5 0 5 20 20 5 0 -5
KnightLine -8 -2 0 2 2 0 -2 -8
KnightRank -3 -1 0 1 4 6 4 1
BishopLine -2 2 5 7 7 5 2 -2
RookFile -1 0 0 1 1 0 0 -1
QueenLine -3 -1 1 3 3 1 -1 -3
KingLine -5 -2 0 5 5 0 -2 -5
KingFile 5 10 1 -1 -1 1 10 5
CenterPawns 750
KnightMobilityOpening 125
KnightMobilityEndgame 100
BishopMobilityOpening 95
BishopMobilityEndgame 85
RookMobilityOpening 70
RookMobilityEndgame 60
QueenMobilityOpening 8
QueenMobilityEndgame 45
RookHalfOpenOpening 80
RookHalfOpenEndgame 160
RookOpenOpening 1200
RookOpenEndgame 200
RookHalfOpenKingAttack 600
RookChecksOpening 25
Rook7Opening 1000
Rook7Endgame 450
Queen7Endgame 50
KnightKingAttack 1500
BishopKingAttack 450
RookKingAttack 500
QueenKingAttack 650
KingAttackWeight 15 35 50 65 75 85 95 120 320 420 520 720 1300 1700 1920 2820
PassedPawnValue1 0 10 110 900 1100 1500 2100 0
PassedPawnValue2 0 30 90 400 1500 2800 3600 0
PassedPawnValue3 0 10 20 45 260 330 420 0
PassedPawnValue4 0 10 20 45 265 340 450 0
PassedPawnValue5 0 10 20 40 345 440 590 0
PassedPawnValue6 0 10 20 30 390 850 1100 0
PassedPawnValue7 0 10 20 35 550 1150 1950 0
PawnDoubledEndgame 310
PawnIsolatedOpening 130
PawnIsolatedEndgame 280
PawnOpenIsolatedOpening 90
PawnOpenIsolatedEndgame 240
PawnWeakEndgame 300
PawnOpenWeakOpening 200
PawnWeakOpening 250
PawnPassedScoreOpening 0 100 500 1000 1300 1900 0 0
PawnPassedScoreEndgame 0 300 500 700 1300 1600 0 0

[Tony Thomas]

Why are you not allowed to mention the name of the website, unless it has illegal contents? If it does have pirated software, it probably isnt well known.

[John Conway]

Why are you not allowed to mention the name of the website, unless it has illegal contents? If it does have pirated software, it probably isnt well known.

I would guess most members here know the name, even if they haven't visited the site.

[matejst]

Now I understand why the moderators deleted my last post: I wrote this adress. I was persuaded that it was the site of a sympathic member of this forum.

Kind regards!

BS