King of Bullet -Turnier mit 100 Engines- Hauptturnier gestar

[Uwe Jakoby]

Geschrieben von: / Posted by: Uwe Jakoby at 11 April 2002 19:25:57:
Fritz 7 Gui, UCI-Protokoll (Wb2Uci-Adapter), eigenes Buch (falls vorhanden), Bedenkzeit 1 Minute

Welche Engine wird King of Bullet ?

Die Engines:
Fritz 6
Hiarcs 7.32
(außer Konkurenz - zum Leistungsvergleich)
Amy 0.7
Amyan 1.52
AnMon5.19
Averno 0.40
Baron 0.29
Beowulf 2.0
Bestia 0.8
Bionic 4.01
Comet B42
Crafty 18-14
Delfi 2.4
Dragon 4.2
ESC 1.09
EXchess 4.03a
Fortress 1.62
Gaviota 0.30
Genesis 1.03
Gerbil 0.2
GnuChess 5.04
GreenLightChess 2.18
Gromit 3.8.2
Hagrid 0.6.34
Horizon 3.0
King of Kings 2.02
LambChop 7.1
Leila 0.52F
Little Goliath 3.5
LordKing 3.2
Monarch 2002-04
MrChess 2.1
Nejmet 3.05
Nimzo2000
Pepito v1.42
Phalanx 22
Pharaon 2.62
Quark 1.5
Queen 2.25
Requiem 0.52
Resp 0.14
Sjeng 12.11
SmarThink 0.05a
SOS.2 for Arena
Tao 5.0
Terra 2.71
The Crazy Bishop
Ufim 1.43
WildCat 2.61
Yace 0.99.56


Uwe's Schachseite

[Jose Carlos]

Geschrieben von: / Posted by: Jose Carlos at 12 April 2002 08:58:06:
Als Antwort auf: / As an answer to: King of Bullet -Turnier mit 100 Engines- Hauptturnier gestartet geschrieben von: / posted by: Uwe Jakoby at 11 April 2002 19:25:57:

Fritz 7 Gui, UCI-Protokoll (Wb2Uci-Adapter), eigenes Buch (falls vorhanden), Bedenkzeit 1 Minute

Welche Engine wird King of Bullet ?
Die Engines:
Fritz 6
Hiarcs 7.32
(außer Konkurenz - zum Leistungsvergleich)
Amy 0.7
Amyan 1.52
AnMon5.19
Averno 0.40
Baron 0.29
Beowulf 2.0
Bestia 0.8
Bionic 4.01
Comet B42
Crafty 18-14
Delfi 2.4
Dragon 4.2
ESC 1.09
EXchess 4.03a
Fortress 1.62
Gaviota 0.30
Genesis 1.03
Gerbil 0.2
GnuChess 5.04
GreenLightChess 2.18
Gromit 3.8.2
Hagrid 0.6.34
Horizon 3.0
King of Kings 2.02
LambChop 7.1
Leila 0.52F
Little Goliath 3.5
LordKing 3.2
Monarch 2002-04
MrChess 2.1
Nejmet 3.05
Nimzo2000
Pepito v1.42
Phalanx 22
Pharaon 2.62
Quark 1.5
Queen 2.25
Requiem 0.52
Resp 0.14
Sjeng 12.11
SmarThink 0.05a
SOS.2 for Arena
Tao 5.0
Terra 2.71
The Crazy Bishop
Ufim 1.43
WildCat 2.61
Yace 0.99.56

Hi, does Averno run under Fritz 7? If so, they must have changed the protocol (not sending the command "fritz" to the engine) and that would be IMO a very big mistake . Any programmer out there can confirm? (I don't have Fritz 7).
Please, don't use Averno under any Chessbase GUI even if it "seems to work".
José C.

http://www.geocities.com/josec_averno

[Alex]

Geschrieben von: / Posted by: Alex at 12 April 2002 18:17:55:
Als Antwort auf: / As an answer to: Fritz protocol geschrieben von: / posted by: Jose Carlos at 12 April 2002 08:58:06:

Hi, does Averno run under Fritz 7? If so, they must have changed the protocol (not sending the command "fritz" to the engine) and that would be IMO a very big mistake . Any programmer out there can confirm? (I don't have Fritz 7).
Please, don't use Averno under any Chessbase GUI even if it "seems to work".
José C.

Hi,
he uses the new wb2uci adapter by Odd Gunnar Malin. The adapter converts the UCI protocol to 'real' WB protocol.
The Fritz7 adapter is the same as under Fritz6, and sends the 'fritz'.
Alex


Odd Gunnar Malin

[Jose Carlos]

Geschrieben von: / Posted by: Jose Carlos at 12 April 2002 22:55:17:
Als Antwort auf: / As an answer to: Re: Fritz protocol geschrieben von: / posted by: Alex at 12 April 2002 18:17:55:

Hi, does Averno run under Fritz 7? If so, they must have changed the protocol (not sending the command "fritz" to the engine) and that would be IMO a very big mistake . Any programmer out there can confirm? (I don't have Fritz 7).
Please, don't use Averno under any Chessbase GUI even if it "seems to work".
José C.

Hi,
he uses the new wb2uci adapter by Odd Gunnar Malin. The adapter converts the UCI protocol to 'real' WB protocol.
The Fritz7 adapter is the same as under Fritz6, and sends the 'fritz'.
Alex

Hello. Thanks for your answer. After reading what Odd's adapter does, and given the nature of UCI protocol, I don't believe the conversion can be good enough not to hurt winboard engines. It's not Odd's fault, but UCI lacks some information WB protocol II needs.
So, after all, my opinion remains the same. I strongly recommend not tu use Averno under Fritz. If anyone wants to do it, it's fine for me, but I'd like people running Averno under Fritz in tournaments understand Averno is not playing full strength there.
José C.


http://www.geocities.com/josec_averno

[Odd Gunnar Malin]

Geschrieben von: / Posted by: Odd Gunnar Malin at 14 April 2002 14:53:21:
Als Antwort auf: / As an answer to: Re: Fritz protocol geschrieben von: / posted by: Jose Carlos at 12 April 2002 22:55:17:

Hi, does Averno run under Fritz 7? If so, they must have changed the protocol (not sending the command "fritz" to the engine) and that would be IMO a very big mistake . Any programmer out there can confirm? (I don't have Fritz 7).
Please, don't use Averno under any Chessbase GUI even if it "seems to work".
José C.

Hi,
he uses the new wb2uci adapter by Odd Gunnar Malin. The adapter converts the UCI protocol to 'real' WB protocol.
The Fritz7 adapter is the same as under Fritz6, and sends the 'fritz'.
Alex

Hello. Thanks for your answer. After reading what Odd's adapter does, and given the nature of UCI protocol, I don't believe the conversion can be good enough not to hurt winboard engines. It's not Odd's fault, but UCI lacks some information WB protocol II needs.
So, after all, my opinion remains the same. I strongly recommend not tu use Averno under Fritz.
If anyone wants to do it, it's fine for me, but I'd like people running Averno under Fritz in tournaments understand Averno is not playing full strength there.
José C.

I must say that I don't understand why you not want people to use your engine under Fritz, but you could add this to your web-page and I'm sure most people would respect your decision.
The only thing I have seen that could harm an engine is if it rely the learning on the result of it's game. This would not affect so much a tournament game like this but more if it is a match over many games like the SSDF do.
Another thing is if it is wise to use the result at all for the learning, when I have watched games with some top engines, mostly Shredder and Tiger, the result is very much depended on luck in the endgame and not so much because of bad opening. I think I read in CCC some month (years?) ago that Dr. Hyatt don't recommanded this but rather look at the eval for some moves after the engine is out of book to see if the book need to be adjusted. I must say that this method looks a little bit more trusty.
I have not implemented any book learning in my engine so I'm not an expert on this but have instead consentrated my work on the book itself to let it variate it's play when it face a human with small book so he could use the engine for learning openings etc.
Beside from this, is it so bad if the engine is dumbed down a little? Isn't the top of the elo curvee (standard deviation) between 1400-1500, so that your engine would suit most people if your engine plays with a strength a few elo points higher than this.
And as a last thing, if a user prefer a GUI, why would you recommand him to not use it. When I play against Averno or another engine I prefer to use either Fritz or Chessmaster because these gui's have focus on playing human-computer and thereby give more pleasure to play with. I think Chessmaster neither send the result so I guess your recommandation goes to this gui too.
Odd Gunnar