Phalanx source code

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

Moderators: hgm, Rebel, chrisw

stevenaaus
Posts: 608
Joined: Wed Oct 13, 2010 9:44 am
Location: Australia

Re: Phalanx source code

Post by stevenaaus »

I'm starting to hit a brick wall here.
One of the problems with phalanx is the mix of versions which differ only in C formating (apart from the CR/LF windows/linux issue.)

The only changes i could gleam from above have been applied to svn.
Interestingly, phalanx (linux) played a proper time-per-game game today for me for the first time after these changes. But the changes seem win32 specific, so perhaps my testing is errroneous.
Haven't tested this last commit with Windows yet, will do so now hopefully.

Anyway - i don't think i can move that polling code to phalanx.h.
cheers, S.A.
Elorejano
Posts: 101
Joined: Sat Mar 20, 2010 3:31 am

Re: Phalanx source code

Post by Elorejano »

Jim, could you give us a little light?
AFAIK, the last source (with ini file) has all the bugfixes, ready for compile in Visual Studio more Danns tweaks
Its right this?
What do you think about 64 bits compiles?
Thanks in advance
stevenaaus
Posts: 608
Joined: Wed Oct 13, 2010 9:44 am
Location: Australia

Re: Phalanx source code

Post by stevenaaus »

It seems to play a well timed game on windows too :)
But my original win32 polling bug still exists.
These four commands don't reset the analysis under windows, though they are fine under Linux

Engine: 8 -1 91 384451 Pf5xe4 Pd2-d4 Ng8-f6
Engine: 9 1 233 919440 Pf5xe4 Pd2-d3 Ng8-f6 Pd3xe4 Pe7-e6 Bc1-g5 Ph7-h6 Bg5xf6 Qd8xf6 Nb1-c3 Bf8-a3
Engine: 9 1 350 1357705 Pf5xe4 Pd2-d3 Ng8-f6 Pd3xe4 Pe7-e6 Bc1-g5 Ph7-h6 Bg5xf6 Qd8xf6 Nb1-c3 Bf8-a3
Scid : exit
Scid : force
Scid : setboard rnbqkbnr/ppppp1pp/8/8/4p3/7P/PPPP1PP1/RNBQKBNR w KQkq - 0 3
Scid : analyze
Engine: 10 10 712 3127860 Pf5xe4 Pd2-d3 Pe4xd3 Bf1xd3 Ng8-f6 Ng1-f3 Nb8-c6 O-O Pd7-d5 Bc1-f4 Pe7-e6
Engine: 10 10 937 3885625 Pf5xe4 Pd2-d3 Pe4xd3 Bf1xd3 Ng8-f6 Ng1-f3 Nb8-c6 O-O Pd7-d5 Bc1-f4 Pe7-e6
stevenaaus
Posts: 608
Joined: Wed Oct 13, 2010 9:44 am
Location: Australia

Re: Phalanx source code

Post by stevenaaus »

Hmmmm - "ponder on" is not an xboard command at all. Crafty acknowledges it, and it is in mainline SCID which made me think it was. :oops: I have replaced SCID's "ponder on/off" with "hard/easy" in Scid vs PC.

Testing phalanx for 1min + 1sec games - it seems it plays fine if pondering (-p+) is enabled, and miserably otherwise This seems to have always been the case.
User avatar
Jim Ablett
Posts: 1384
Joined: Fri Jul 14, 2006 7:56 am
Location: London, England
Full name: Jim Ablett

Re: Phalanx source code

Post by Jim Ablett »

Elorejano wrote:Jim, could you give us a little light?
AFAIK, the last source (with ini file) has all the bugfixes, ready for compile in Visual Studio more Danns tweaks
Its right this?
What do you think about 64 bits compiles?
Thanks in advance

Hi Carlos,

Phalanx version with config file has all the fixes, but is just my own 'fun' version and not something Steven would want to merge into ScidvsPc project.
Download Phalanx src code from ScidvcPc svn repository if you want the latest regular Phalanx version to compile.

Compiling in 64 bit has no speed benefit. Phalanx doesnt use bitboards. In my tests it ran slower.

Jim.
User avatar
Jim Ablett
Posts: 1384
Joined: Fri Jul 14, 2006 7:56 am
Location: London, England
Full name: Jim Ablett

Re: Phalanx source code

Post by Jim Ablett »

Anyway - i don't think i can move that polling code to phalanx.h.
I just moved it there so I could use Bioskey() it in different places while I was tracking down ponder bug.

Jim.
stevenaaus
Posts: 608
Joined: Wed Oct 13, 2010 9:44 am
Location: Australia

Re: Phalanx source code

Post by stevenaaus »

Hmmm - i just found a bug in ScidvsPC's win32 phalanx.
Playing White, It won't allow black to castle in tournament play, but says

Code: Select all

Scid: e8g8
Engine: illegal move: e8g8 
Playing Black, it allows white to castle fine.

It could be an issue with the cygwin libs i used in later version, and i'll package Jim's binary in future i think, unless i can sort it out. Linux seems unaffected.
User avatar
jshriver
Posts: 1342
Joined: Wed Mar 08, 2006 9:41 pm
Location: Morgantown, WV, USA

Re: Phalanx source code

Post by jshriver »

I am having the same problem and have removed it from my testing/tournaments. 50 or so games and realized it always lost on time and never made a move after book :(

-Josh
User avatar
Jim Ablett
Posts: 1384
Joined: Fri Jul 14, 2006 7:56 am
Location: London, England
Full name: Jim Ablett

Re: Phalanx source code

Post by Jim Ablett »

jshriver wrote:I am having the same problem and have removed it from my testing/tournaments. 50 or so games and realized it always lost on time and never made a move after book :(

-Josh
Hi Josh,

I think you're using the old version that had no windows polling code and so no ponder under windows.
Polling had to be switched off on the command line >

Code: Select all

-o <+/->  polling input                       default&#58; on
       Use -o- for running test suites. Phalanx reads the positions from its
       standard input and polling input makes it stop thinking after almost
       zero seconds on each position.
My version fixed polling & ponder under windows >

http://dl.dropbox.com/u/5047625/phalanx-xxiii-ja.zip

Jim.
User avatar
jshriver
Posts: 1342
Joined: Wed Mar 08, 2006 9:41 pm
Location: Morgantown, WV, USA

Re: Phalanx source code

Post by jshriver »

I'll check it out, but I was using it under Linux, and compiling from source.