Anubis ... a very special chess program!

[Frank Quisinsky]

Hi there,

Anubis 3.06 xb is playing with around 3000 Elo ... if Shredder 12 is playing with 2800 Elo.
But after clearly more as 400 games no of the stronger NN engines can beat this engine in lesser as 45 moves.
Only Hakkapeliitta in one game. A bit sensational if I am looking on the other engine results I have.

Test-run is at the moment still running, but it seems that is a complete other chess engine with a complete other playing style. The Spain programmer do here really a very special work. Very safty chess after the opning books moves. For ETOC-G I am using an equal opening book and can produce a draw quote lesser as 50% for all of the TOP-250 engines = more fun in watching the games!

Interesting ist the next test-run.
The newer Gilipol (now UCI) by the same Spain programmer will be running.

For ETOC-G all of the TOP-250 engines have to play with the same group of 26 opponents.
End of the day ... new knowledge can be generate to all the older and newer engines, today in TOP-250.

Anubis has pleasantly surprised me at the moment!
My congrats, a nice work!

ETOC-G home:
https://www.amateurschach.de/main/_etoc-g.htm
Note: The file "checkfile" are the classical ratings from the results I have (in front of the site).

Best
Frank

[Frank Quisinsky]

In times today ...
Each other playing-style we've got to celebrate.
Definitely with a delicious bratwurst.
Oh, I have to go shopping; I can't just stand here watching any longer. I have to buy a bratwurst.

Anubis, an event ... you know!

[Frank Quisinsky]

Those who see what is against them they usually see their own attack as well.
Not new ...

So, after 650 games ... with 3000 Elo "only" in this very very strong field of 26 opponents (most opponents are good known and very strong attacker).

2x won and "only" 3x lost in lesser as 45 moves to mate.

Strong, Anubis rocked!
Time for my Bratwurst!

Such engines have a lot of potential for further improvments!

[Frank Quisinsky]

Hi there,

I wrote this in TalkChess after Gilipol 1.00 release.
syzygy not working for me!!

I test it normaly with edit an easy position: KNB vs. K and 5-pieces syzygybases.
I don't see how Gilipol would then use the emdgame databases as well.

So, unfortunatley ...
Gilipol 1.0.0 will be start in my test without using the 5-pieces syzygybases.

On the end of the message can be found the Shredder 12 debug file.
You can see, Gilipol give the information all will be OK!

But nothing is OK with syzygybases
Nothing ...
I give up computer chess for ever
For ever
I give up
I'm a complete wreck
It's all over


No, no ... maybe in 20 years.

I think all is very new with uci code and I am sure the programmer offer sooner or later an update.

Shredder 12 GUI
Engine configuration:

Code: Select all

[ENGINE]
Name=Gilipol 1.00
Author=Jose Carlos Martinez Galan
Filename=C:\Chess\Shredder_12\Engines\Gilipol\Gilipol_100.exe
[OPTIONS]
SyzygyPath=c:\ajedrez\TB_syzygy
SyzygyProbeLimit=5

Code: Select all

### 2.: to Gilipol 1.00 (0): quit

### 3.: to Gilipol 1.00 (0): uci

### 4.: to Gilipol 1.00 (0): debug on

### from Gilipol 1.00 (0): Info: Gilipol (Single-thread)

##### Unknown token: Info:
### from Gilipol 1.00 (0): Info: Inicializando estructuras del motor...

##### Unknown token: Info:
### from Gilipol 1.00 (0): Info: Inicializando tablas de ataques...

##### Unknown token: Info:
### from Gilipol 1.00 (0): Info: Tablas de ataques inicializadas.

##### Unknown token: Info:
### from Gilipol 1.00 (0): info string Hash inicializada: 32 MB (524288 clusters)

Hash inicializada: 32 MB (524288 clusters)
### from Gilipol 1.00 (0): Info: NNUE cargada correctamente (net4.bin)

##### Unknown token: Info:
### from Gilipol 1.00 (0): id name Gilipol 1.00

### from Gilipol 1.00 (0): id author Jose Carlos Martinez Galan

### from Gilipol 1.00 (0): option name Hash type spin default 32 min 1 max 1024

### from Gilipol 1.00 (0): option name Ponder type check default true

### from Gilipol 1.00 (0): option name MultiPV type spin default 1 min 1 max 1

### from Gilipol 1.00 (0): option name SyzygyPath type string default <empty>

### from Gilipol 1.00 (0): option name SyzygyProbeDepth type spin default 1 min 1 max 100

### from Gilipol 1.00 (0): option name Syzygy50MoveRule type check default true

### from Gilipol 1.00 (0): option name SyzygyProbeLimit type spin default 7 min 0 max 7

### from Gilipol 1.00 (0): option name EvalFile type string default net4.bin

### from Gilipol 1.00 (0): uciok

### 5.: to Gilipol 1.00 (0): setoption name Hash value 256

### 6.: to Gilipol 1.00 (0): isready

### from Gilipol 1.00 (0): info string Hash inicializada: 256 MB (4194304 clusters)

Hash inicializada: 256 MB (4194304 clusters)
### from Gilipol 1.00 (0): readyok

setoption name Ponder value false

### 7.: to Gilipol 1.00 (0): setoption name Ponder value false

setoption name SyzygyPath value c:\ajedrez\TB_syzygy

### 8.: to Gilipol 1.00 (0): setoption name SyzygyPath value c:\ajedrez\TB_syzygy

setoption name SyzygyProbeLimit value 5

### 9.: to Gilipol 1.00 (0): setoption name SyzygyProbeLimit value 5

### 10.: to Gilipol 1.00 (0): isready

### from Gilipol 1.00 (0): readyok

Engine: Gilipol 1.00 (256 MB)
by Jose Carlos Martinez Galan

Bug should be fix because such a nice engine should be powered the next 100 years.
I saw in GitHub, programmer is working, have a version 1.03.
But no executables are available.

I give up!
No more computer chess for ever ..
I've just changed my mind!!
My test started around 23:00 German time after ...

"The GREAT" ...
"The small 3000 Elo sensation, the Spain answer against the most powerful attackers in the World of Engines."

Dirty is also such a fine engine.
But Dirty is not the topic!

means Anubis 3.06 xb
Really a very very nice work!

Best
Frank

[Ajedrecista]

Hello Frank:

[...]

Interesting ist the next test-run.
The newer Gilipol (now UCI) by the same Spain programmer will be running.

[...]

Disgusting name to say the least. The Spanish word gilipollas is a hard, common, vulgar insult that could be translated as asshole. The word gilipol does not exist itself, but I can tell you that this is not a coincidence, because the official dictionary of Spanish language lists very few words with this root, all of them related to the same thing. The engine may be good, but the naming is an epic trolling. Just wanted you to know what you have in your hands.

Thank you very much for your long lasting work.

Regards from Spain.

Ajedrecista.

[Frank Quisinsky]

It’s more or less a German idiom:
“I only understand Spanish” (even when a German is speaking German to me and I understand not what he is meaning).

It’s an unusual name for an engine, but why not? I get what you mean. I actually had a Spanish girlfriend once, and eventually I figured out what she was saying when she was swearing. She was always mad at me for having so many chess computers. I just had to look into her eyes and I knew... Frank, put your computers away right away. I learn a bit Spanish at this time in younger years, but only the words never I should learn.

You know... for me, the most beautiful name for an engine ever released is AnMon.

Means...
An = ANgel
Mon = deMON

But today, AnMon is more of an angel than a demon with a 2400 Elo rating.
Also a great engine from Christian BARRETEAU (France).

True engine enthusiasts look more closely.
Unfortunately, many only look at Elo and fail to discover the beauty of the game.

[Ajedrecista]

Hello:

It’s more or less a German idiom:
“I only understand Spanish” (even when a German is speaking German to me and I understand not what he is meaning).

[...] I learn a bit Spanish at this time in younger years, but only the words never I should learn.

You know... for me, the most beautiful name for an engine ever released is AnMon.

Means...
An = ANgel
Mon = deMON

But today, AnMon is more of an angel than a demon with a 2400 Elo rating.
Also a great engine from Christian BARRETEAU (France).

[...]

I did not know about the German idiom. Regarding the first words learnt in any language... this happens always, all the time. Period.

I had not realised about AnMon naming, thanks! I use a version of AnMon embedded at Arena 2.0.1 GUI among other oldies when I want to test some positions with oldies. You know, 'oldie but goldie' that applies to some test positions.

Regards from Spain.

Ajedrecista.

[Frank Quisinsky]

Arena was a great time. But the era of chess computers was even better. Kittinger, Spracklens, Schröder, Lang, Danielsen, Morsch, Kaplan / Barnes, Rathsman and so one.

Twice a year, I wrote to Rudolf Huber (SOS). Back then, I had always wanted to use his engine for my Winboard site (later I gave my work around Winboard to Thomas Meyer, later the downloads and other material to Leo Dijksman). But he always wrote back saying it was strictly a private engine. Then I started working with Martin Blume on Arena. Rudolf noticed this and suddenly sent me SOS. That was the best moment I’ve ever had in computer chess. He wrote, “Frank, for your tireless work, I’m giving you SOS for Arena.” That’s how the engine got its name: SOS for Arena.

Stefan Meyer-Kahlen once told me during a phone call that Rudolf Huber is fundamentally one of the most talented German programmers (not for computer chess only). So Stefan was obviously a fan of his himself. I was as proud as can be. I had SOS and AnMon (both M.T.D.) on my site with more as 250.000 downloads. And Arena have 4 Million downloads only in the best year.

Rudolf Huber is one of the UCI programmers too.

[Frank Quisinsky]

Hi there,

test-run endet for the new "Gilipol 1.0" now with UCI support.
More or less the same results as for "Anubis 3.06 xb".
Anubis is a bit stronger, reason are the not working syzygybases with Shredder 12 GUI.

That was a very interesting test for me.
Thanks for both engines again!

Best
Frank