Glaurung 2.2

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Harvey Williamson, bob

Tord Romstad
Posts: 1808
Joined: Wed Mar 08, 2006 8:19 pm
Location: Oslo, Norway

Glaurung 2.2

Post by Tord Romstad » Sat Dec 20, 2008 12:50 am

Hi all,

The source code and Mac OS X executables for Glaurung 2.2 are now available from my web site. Executables for GNU/Linux and Windows will follow shortly.

This version is mostly the result of work I did in preparation for the Open Polish Computer Chess Championships in June. When I tested the most recent development version before the tournament, it performed poorly, and I ended up using version 2.1 in the tournament. Evidently, I didn't test thoroughly enough, because recent tests indicate that the version I prepared for the OPCCC is considerably stronger than 2.1. I have therefore fixed a handful of very minor bugs, polished the code a little, and published the result as Glaurung 2.2.

The most important improvement is the addition of space evaluation. This improves the strength (particularly in Chess960, but also in normal chess), and also makes the style of play more attractive.

For the next version, I will have a close look at Marco Costalba's improvements in Stockfish 1.1. Stockfish 1.1 seems extremely strong (possibly stronger than Glaurung 2.2), and it is surprising that it doesn't get more attention. In the current version of Glaurung, there is only one new thing stolen from Stockfish: Razoring has been modified to use only the approximate evaluation of the position, rather than the full eval.

Tord

swami
Posts: 6535
Joined: Thu Mar 09, 2006 3:21 am

Re: Glaurung 2.2

Post by swami » Sat Dec 20, 2008 1:15 am

Thanks, Tord, for the new release.

Glaurung has just entered CCT, I will update it in a while.

Good luck!

Dirt
Posts: 2851
Joined: Wed Mar 08, 2006 9:01 pm
Location: Irvine, CA, USA

Re: Glaurung 2.2

Post by Dirt » Sat Dec 20, 2008 1:38 am

Tord Romstad wrote:Stockfish 1.1 seems extremely strong (possibly stronger than Glaurung 2.2), and it is surprising that it doesn't get more attention.
I think that's partly due to Stockfish 1.0 not performing well.

Dann Corbit
Posts: 8600
Joined: Wed Mar 08, 2006 7:57 pm
Location: Redmond, WA USA
Contact:

Re: Glaurung 2.2

Post by Dann Corbit » Sat Dec 20, 2008 3:58 am

Tord Romstad wrote:Hi all,

The source code and Mac OS X executables for Glaurung 2.2 are now available from my web site. Executables for GNU/Linux and Windows will follow shortly.

This version is mostly the result of work I did in preparation for the Open Polish Computer Chess Championships in June. When I tested the most recent development version before the tournament, it performed poorly, and I ended up using version 2.1 in the tournament. Evidently, I didn't test thoroughly enough, because recent tests indicate that the version I prepared for the OPCCC is considerably stronger than 2.1. I have therefore fixed a handful of very minor bugs, polished the code a little, and published the result as Glaurung 2.2.

The most important improvement is the addition of space evaluation. This improves the strength (particularly in Chess960, but also in normal chess), and also makes the style of play more attractive.

For the next version, I will have a close look at Marco Costalba's improvements in Stockfish 1.1. Stockfish 1.1 seems extremely strong (possibly stronger than Glaurung 2.2), and it is surprising that it doesn't get more attention. In the current version of Glaurung, there is only one new thing stolen from Stockfish: Razoring has been modified to use only the approximate evaluation of the position, rather than the full eval.

Tord
I guess it is due to not enough test results for Stockfish 1.1 yet.
Once a thousand games or so are finished and it is clear how strong it is exactly, it will get plenty of attention if it still looks like a world beater.

Tord Romstad
Posts: 1808
Joined: Wed Mar 08, 2006 8:19 pm
Location: Oslo, Norway

Re: Glaurung 2.2

Post by Tord Romstad » Sat Dec 20, 2008 11:42 am

The 32-bit and 64-bit Windows binaries are now available. Thanks, Jim Ablett! :)

Linux binaries will be added later today.

Tord

User avatar
AdminX
Posts: 4864
Joined: Mon Mar 13, 2006 1:34 pm
Location: Acworth, GA
Contact:

Re: Glaurung 2.2

Post by AdminX » Sat Dec 20, 2008 12:12 pm

"Good decisions come from experience, and experience comes from bad decisions."
__________________________________________________________________
Ted Summers

Tord Romstad
Posts: 1808
Joined: Wed Mar 08, 2006 8:19 pm
Location: Oslo, Norway

Re: Glaurung 2.2

Post by Tord Romstad » Sat Dec 20, 2008 12:19 pm

AdminX wrote:Here is the link:
http://www.mediafire.com/?jnhmhw1wmik
I recommend downloading from my site instead, so you will get the Readme.txt file and the license text in addition to just the binaries.

Tord

User avatar
Denis P. Mendoza
Posts: 410
Joined: Fri Dec 15, 2006 8:46 pm
Location: Philippines
Contact:

Re: Glaurung 2.2

Post by Denis P. Mendoza » Sat Dec 20, 2008 1:24 pm

Tord Romstad wrote:Hi all,

The source code and Mac OS X executables for Glaurung 2.2 are now available from my web site. Executables for GNU/Linux and Windows will follow shortly.

This version is mostly the result of work I did in preparation for the Open Polish Computer Chess Championships in June. When I tested the most recent development version before the tournament, it performed poorly, and I ended up using version 2.1 in the tournament. Evidently, I didn't test thoroughly enough, because recent tests indicate that the version I prepared for the OPCCC is considerably stronger than 2.1. I have therefore fixed a handful of very minor bugs, polished the code a little, and published the result as Glaurung 2.2.

The most important improvement is the addition of space evaluation. This improves the strength (particularly in Chess960, but also in normal chess), and also makes the style of play more attractive.

For the next version, I will have a close look at Marco Costalba's improvements in Stockfish 1.1. Stockfish 1.1 seems extremely strong (possibly stronger than Glaurung 2.2), and it is surprising that it doesn't get more attention. In the current version of Glaurung, there is only one new thing stolen from Stockfish: Razoring has been modified to use only the approximate evaluation of the position, rather than the full eval.

Tord
Thanks Tord for the new update... I'm a Glaurung fan too!

You're right about Stockfish1.1. It's quite strong indeed. I have been interested in testing/tweaking Glaurung and Stockfish lately due to this curiosity. I noticed that Stockfish 1.1 performs well on blitz on these settings compared to default (I could be wrong):
Futility Margin 2 = 200
Maximum Razoring Depth = 2
Razoring Margin = 200
It's just preliminary, and results may vary. The one thing I like is, it is beating the latest Togas, go toe-to-toe with some old Rybkas, and giving the new sensation, 'Inert Thinker' a taste of Glaurung's medicine!

Here're my quick 1"/1' results. I just called it Glaurung-64 as I've been adding more UCI options on the code to tweak, yet everything was still original except for the stated parameters:
-----------------GlaurungII-64-----------------
GlaurungII-64 - GrapeFruit1beta : 7.5/10 7-2-1 (111101=011) 75% +191
GlaurungII-64 - Rybka v1.1.x64 : 2.5/10 1-6-3 (==0100=000) 25% -191
GlaurungII-64 - Rybka v1.2f.x64 : 6.0/10 5-3-2 (01101=10=1) 60% +70
GlaurungII-64 - SP-Inert-Thinker_32-bit : 10.0/10 10-0-0 (1111111111) 100% +1200
GlaurungII-64 - SP-Inert-Thinker_64-bit : 4.5/10 3-4-3 (0=11=0010=) 45% -35
GlaurungII-64 - TogaI143JD15 : 6.0/10 5-3-2 (10111=010=) 60% +70
Congrats and good luck at CCT 2009!


Denis

royb
Posts: 507
Joined: Wed Mar 08, 2006 11:53 pm

Re: Glaurung 2.2

Post by royb » Sat Dec 20, 2008 3:17 pm

I've attempted to compile Glaurung-2.2 for on my Ubuntu-8.10 (64-bit) system and I get this:

roy@roy-laptop:~/chess/glaurung22/src$ make
g++ -DNDEBUG -Wall -g -O3 -fno-exceptions -fomit-frame-pointer -fno-rtti -fstrict-aliasing -c -o bitboard.o bitboard.cpp
In file included from direction.h:27,
from bitboard.h:42,
from bitboard.cpp:26:
square.h: In function ‘int file_distance(File, File)’:
square.h:141: error: ‘abs’ was not declared in this scope
square.h: In function ‘int rank_distance(Rank, Rank)’:
square.h:149: error: ‘abs’ was not declared in this scope
bitboard.cpp: In function ‘void<unnamed>::init_ray_bitboards()’:
bitboard.cpp:401: warning: suggest parentheses around + or - in operand of &
bitboard.cpp: In function ‘void<unnamed>::init_attacks()’:
bitboard.cpp:426: error: ‘abs’ was not declared in this scope
make: *** [bitboard.o] Error 1

What customizations should I make to the Makefile to correct this?

Also, I'd like to produce a 64-bit binary so tweaks to the Makefile to allow that are appreciated also.

Thanks,

Roy

Spock

Re: Glaurung 2.2

Post by Spock » Sat Dec 20, 2008 3:19 pm

Tord Romstad wrote: The most important improvement is the addition of space evaluation. This improves the strength (particularly in Chess960, but also in normal chess), and also makes the style of play more attractive.

Tord
Thanks Tord, another chess960 engine for me to run :)

Post Reply