Fridolin 2.0

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

Moderators: hgm, Rebel, chrisw

Engin
Posts: 918
Joined: Mon Jan 05, 2009 7:40 pm
Location: Germany
Full name: Engin Üstün

Re: Fridolin 1.0 and 2.0 available!

Post by Engin »

fridokar wrote:I would say not more than 2300 CEGT Elos. But there are running tests in the CEGT tournament. So please be patient... :wink:
hello Christian,
hope i was not to fast to spot light your engine here that will be downloadable at your site.
fridokar
Posts: 9
Joined: Thu Dec 04, 2014 2:14 pm

Re: Fridolin 1.0 and 2.0 available!

Post by fridokar »

Hi Engin,

not at all. I'am happy about your posting. Thank you! I hope to see you again somewhere!
supersharp77
Posts: 1242
Joined: Sat Jul 05, 2014 7:54 am
Location: Southwest USA

Re: Fridolin 1.0 and 2.0 Test Games

Post by supersharp77 »

Fridolin 1.0 and Fridolin 2.0 test games..v1.0 solid play but not spectacular
good results overall (see pgn file) Fridolin 2.0 solid but not spectacular..a bit deeper search than v1.0.....major flaw in Fritz gui..Lots of Time Losses
(almost every game)...takes its sweet time making the moves..needs some adjustment..Thx AR :) :wink:


[url]: http://s000.tinyupload.com/?file_id=088 ... 7185146230

Fridolin Games pgn file
User avatar
Graham Banks
Posts: 41423
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: Fridolin 1.0 and 2.0 Test Games

Post by Graham Banks »

supersharp77 wrote:Fridolin 1.0 and Fridolin 2.0 test games..v1.0 solid play but not spectacular
good results overall (see pgn file) Fridolin 2.0 solid but not spectacular..a bit deeper search than v1.0.....major flaw in Fritz gui..Lots of Time Losses
(almost every game)...takes its sweet time making the moves..needs some adjustment..Thx AR :) :wink:


[url]: http://s000.tinyupload.com/?file_id=088 ... 7185146230

Fridolin Games pgn file
What time control are you using?
No losses on time here so far?
gbanksnz at gmail.com
gerold
Posts: 10121
Joined: Thu Mar 09, 2006 12:57 am
Location: van buren,missouri

Re: Fridolin 1.0 and 2.0 Test Games

Post by gerold »

supersharp77 wrote:Fridolin 1.0 and Fridolin 2.0 test games..v1.0 solid play but not spectacular
good results overall (see pgn file) Fridolin 2.0 solid but not spectacular..a bit deeper search than v1.0.....major flaw in Fritz gui..Lots of Time Losses
(almost every game)...takes its sweet time making the moves..needs some adjustment..Thx AR :) :wink:


[url]: http://s000.tinyupload.com/?file_id=088 ... 7185146230

Fridolin Games pgn file
Running 5 minute games in Arena with no time lost games by Fridolin.

Best,
Gerold.
Modern Times
Posts: 3546
Joined: Thu Jun 07, 2012 11:02 pm

Re: Fridolin 1.0 and 2.0 Test Games

Post by Modern Times »

800 chess960 games played perfectly here, no problems at all.
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Fridolin 2.0

Post by MikeB »

Engin wrote:Fridolin 2.0 Leiden that played in WCCC 2015 is for download here:

https://sites.google.com/site/fridolinchess/

would be interesting to include this engine in tournaments besides Maverick engine
Awesome - just added a Mac exe to your package - you may download it here.

https://www.dropbox.com/s/nd8qxbhsq8rqf ... 0.zip?dl=0

Also added the compiler options I used. Did not spend a lot of time - the standard icc defauts worked pretty well.

Benchmark: - Start Fridolin with no settings applied - type "go depth 17" (must type "uci" first)

Last login: Thu Aug 6 18:35:05 on ttys000
iMac:~ michaelbyrne$ /Users/michaelbyrne/Downloads/Fridolin-2.00/src/Fridolin ; exit;

Fridolin 2.00 x64 (Aug 6 2015)
created by C.Sommerfeld

uci
id name Fridolin 2.00
id author C.Sommerfeld
option name Hash type spin min 0 max 16384 default 64
option name Threads type spin min 1 max 8 default 1
option name Ponder type check default true
option name UCI_Chess960 type check default false
option name UCI_AnalyseMode type check default false
option name Bitbases Path type string default <empty>
option name Bitbases Cache type spin min 1 max 256 default 32
option name Bitbases Load type combo var None var 4Men Full var 5Men Smart var 5Men Full var 6Men Smart default None
option name Bitbases MinDepth type spin min 1 max 100 default 3
option name NalimovPath type string default <empty>
option name NalimovCache type spin min 1 max 256 default 32
option name Tablebases Usage type combo var Never var OnRoot var InSearch default OnRoot
option name Tablebases MinDepth type spin min 1 max 100 default 6
option name MultiPV type spin min 1 max 20 default 1
option name UCI_ShowCurrLine type check default false
option name UCI_ShowRefutations type check default false
option name UCI_EngineAbout type string default Fridolin by C.Sommerfeld
option name Logging type check default false
option name Contempt Factor type spin min -500 max 500 default 0
uciok
go depth 17

---

info time 9202 nodes 37693644 nps 4096244 cpuload 998 hashfull 634 tbhits 0
bestmove e2e4 ponder e7e5
Alexander Schmidt
Posts: 1203
Joined: Thu May 10, 2007 2:49 pm

Re: Fridolin 2.0

Post by Alexander Schmidt »

Engin wrote:Fridolin 2.0 Leiden that played in WCCC 2015 is for download here:

https://sites.google.com/site/fridolinchess/

would be interesting to include this engine in tournaments besides Maverick engine
Did someone get it to work under Linux? It compiles without problems, and I can get it to work in the terminal. Just a GUI don't recognize it:

PyChess: "There is something wrong with the executeable"
Scid hangs when loading Fridolin.
abulmo
Posts: 151
Joined: Thu Nov 12, 2009 6:31 pm

Re: Fridolin 2.0

Post by abulmo »

Under Linux, input & output are buffered.
I was able to make it work by modifying the function CConsole::init() in the file system.cpp. I just moved an #endif directive so that the setvbuf() calls are also executed under Linux. The corrected function code is:

Code: Select all

void CConsole&#58;&#58;init&#40;void&#41;
&#123;
    m_waiting = true;
    m_logging = false;

#if defined&#40;IS_WINDOWS&#41;

    DWORD dw;
    m_stdin = GetStdHandle&#40;STD_INPUT_HANDLE&#41;;
    m_pipe = !GetConsoleMode&#40;m_stdin, &dw&#41;;

    if (!m_pipe&#41;
    &#123;
        SetConsoleMode&#40;m_stdin, dw &~ &#40;ENABLE_MOUSE_INPUT | ENABLE_WINDOW_INPUT&#41;);
        FlushConsoleInputBuffer&#40;m_stdin&#41;;
    &#125;
    else
#endif
    &#123;
        setvbuf&#40;stdin, NULL, _IONBF, 0&#41;;
        setvbuf&#40;stdout, NULL, _IONBF, 0&#41;;
    &#125;

&#125;
Richard
Alexander Schmidt
Posts: 1203
Joined: Thu May 10, 2007 2:49 pm

Re: Fridolin 2.0

Post by Alexander Schmidt »

abulmo wrote:Under Linux, input & output are buffered.
I was able to make it work by modifying the function CConsole::init() in the file system.cpp. I just moved an #endif directive so that the setvbuf() calls are also executed under Linux. The corrected function code is:

Code: Select all

void CConsole&#58;&#58;init&#40;void&#41;
&#123;
    m_waiting = true;
    m_logging = false;

#if defined&#40;IS_WINDOWS&#41;

    DWORD dw;
    m_stdin = GetStdHandle&#40;STD_INPUT_HANDLE&#41;;
    m_pipe = !GetConsoleMode&#40;m_stdin, &dw&#41;;

    if (!m_pipe&#41;
    &#123;
        SetConsoleMode&#40;m_stdin, dw &~ &#40;ENABLE_MOUSE_INPUT | ENABLE_WINDOW_INPUT&#41;);
        FlushConsoleInputBuffer&#40;m_stdin&#41;;
    &#125;
    else
#endif
    &#123;
        setvbuf&#40;stdin, NULL, _IONBF, 0&#41;;
        setvbuf&#40;stdout, NULL, _IONBF, 0&#41;;
    &#125;

&#125;
That works, thank you very much!!