Critter 1.6 released!

[mclane]

first results from my mobile (LG Optimus Speed p990):

Critter1.6 32bit * 7,5 7 7 21,5 /30 1
Komodo32 3AB 2,5 * 4,5 1 8/30 4
Robbolito085e4 3 5,5 * 4,5 13/30 3
Stockfish2.2.2ja 3 9 5,5 * 17,5/30 2

one core each, permanent brain on.

[Arpad Rusz]

Thanks!
I have tried the 32 bit version in Chessbase 11. First time it immediatelly crashed the program but since then I didn't managed to see a crash again after several tries: it seems that it's working now.

[ernest]

No problem with Nalimov bases either that i can see - Scid or Scid vs PC.

Just curious, why mention Nalimov as Critter only supports Gaviota?

[Marek Soszynski]

The previous Gaviota tablebase bug seems to be fixed on my Linux Mint 32 bit system, cheers. (I don't have the full set of bases installed though.)

Try the new Critter with Tablebase Usage set to "Everywhere" in late middlegame positions (so the tablebases are accessed in the search).

[Tomcass]

Thank you very much for this precious gift, Richard!.

Tom.

[Norm Pollock]

Try the new Critter with Tablebase Usage set to "Everywhere" in late middlegame positions (so the tablebases are accessed in the search).

Critter 1.6 works much better when it is set to the root. TB access wastes time unless it is at the root, or very close to the root.

[Marek Soszynski]

Try the new Critter with Tablebase Usage set to "Everywhere" in late middlegame positions (so the tablebases are accessed in the search).

Critter 1.6 works much better when it is set to the root. TB access wastes time unless it is at the root, or very close to the root.

Two points. 1) Critter crashes on my system when set to "Everywhere," so it's not a question of working less well.
2) I decide what settings to use depending on what sorts of positions I'm studying. I already mentioned "late middlegame positions" in this connexion.

[stevenaaus]

The previous Gaviota tablebase bug seems to be fixed on my Linux Mint 32 bit system, cheers. (I don't have the full set of bases installed though.)

Try the new Critter with Tablebase Usage set to "Everywhere" in late middlegame positions (so the tablebases are accessed in the search).

Trying your test position, no problems with the 32 bit binary (Core2Quad)

Code: Select all

Scid vs. PC <--> Engine communication log file
Engine&#58; Critter 1.6 &#40;32 bit&#41;
Command&#58; critter-16
Directory&#58; /home/steven/.scidvspc
Date&#58; Sun Jun 17 11&#58;41&#58;04 +0000 2012

Note&#58; First line from engine seen; sending it initial commands now.
Scid  &#58; uci
Engine&#58; Critter v1.6 32-bit, by Richard Vida
Engine&#58; hardware POPCNT disabled
Engine&#58; id name Critter 1.6 32-bit
Engine&#58; id author Richard Vida
Engine&#58; option name Hash type spin default 64 min 8 max 1024
Engine&#58; option name Ponder type check default false
Engine&#58; option name OwnBook type check default true
Engine&#58; option name Book File type string default book.cbk
Engine&#58; option name UCI_Chess960 type check default false
Engine&#58; option name MultiPV type spin default 1 min 1 max 100
Engine&#58; option name Threads type spin default 4 min 1 max 32
Engine&#58; option name Split Depth type spin default 5 min 4 max 15
Engine&#58; option name GaviotaTbPath type string default /gtb
Engine&#58; option name GaviotaTbCache type spin default 32 min 4 max 256
Engine&#58; option name GaviotaTbCompression type combo default cp4 var uncompressed var cp1 var cp2 var cp3 var cp4
Engine&#58; option name Tablebase Usage type combo default Only At Root var Disable var Only At Root var Everywhere
Engine&#58; option name Clear Hash type button
Engine&#58; option name Use Session File type check default false
Engine&#58; option name Session File type string default session.csf
Engine&#58; option name SF Size type spin default 32 min 2 max 1024
Engine&#58; option name SF Write Depth type spin default 8 min 2 max 32
Engine&#58; option name SF Material Limit type spin default 0 min 0 max 32
Engine&#58; option name SF Move Limit type spin default 0 min 0 max 999
Engine&#58; option name Clear SF type button
Engine&#58; option name Resolve Score Drops type check default false
Engine&#58; uciok
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name Hash value 64
Scid  &#58; isready
Engine&#58; info string Opening book "book.cbk" not found
Engine&#58; info string GTB Init FAILED
Engine&#58; failed to allocate large pages, errno = 22
Engine&#58; info string Using 4 thread&#40;s&#41;, 64 MB hash LP=off
Engine&#58; readyok
Scid  &#58; setoption name OwnBook value false
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name MultiPV value 1
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name Book File value book.cbk
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name Threads value 4
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name Split Depth value 5
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name GaviotaTbPath value /usr/share/scid/gtb
Scid  &#58; isready
Engine&#58; info string GTB Init OK &#40;3 piece set&#41;
Engine&#58; info string Using 4 thread&#40;s&#41;, 64 MB hash LP=off
Engine&#58; readyok
Scid  &#58; setoption name GaviotaTbCache value 32
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name GaviotaTbCompression value cp4
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name Tablebase Usage value Everywhere
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name Use Session File value false
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name Session File value session.csf
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name SF Size value 32
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name SF Write Depth value 8
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name SF Material Limit value 0
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name SF Move Limit value 0
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; setoption name Resolve Score Drops value false
Scid  &#58; isready
Engine&#58; readyok
Scid  &#58; position fen 4r1k1/3p4/8/8/8/8/4P1PP/4N1K1 w - - 0 1
Scid  &#58; go infinite
Engine&#58; info multipv 1 depth 2 seldepth 2 nodes 162 time 0 nps 0 hashfull 0 tbhits 0 score cp -37 pv g1f2 g8f7 e1d3 
Engine&#58; info multipv 1 depth 3 seldepth 3 nodes 254 time 1 nps 254000 hashfull 0 tbhits 0 score cp -37 pv g1f2 g8f7 e1d3 
Engine&#58; info multipv 1 depth 4 seldepth 5 nodes 461 time 1 nps 461000 hashfull 0 tbhits 0 score cp -33 pv g1f2 g8f7 e1f3 f7f6 
Engine&#58; info multipv 1 depth 5 seldepth 6 nodes 953 time 1 nps 953000 hashfull 0 tbhits 0 score cp -33 pv g1f2 g8f7 e1d3 d7d6 d3f4 
Engine&#58; info multipv 1 depth 6 seldepth 7 nodes 1641 time 2 nps 820500 hashfull 0 tbhits 0 score cp -29 pv g1f2 g8f7 e1d3 d7d6 d3f4 e8a8 
Engine&#58; info multipv 1 depth 7 seldepth 9 nodes 6608 time 5 nps 1321600 hashfull 0 tbhits 0 score cp -33 pv g1f2 e8a8 e1d3 g8g7 d3c5 a8h8 h2h3 h8f8 f2e3 
Engine&#58; info multipv 1 depth 8 seldepth 10 nodes 11410 time 12 nps 950833 hashfull 0 tbhits 0 score cp -30 pv g1f2 e8a8 e1d3 g8g7 d3c5 d7d5 h2h4 a8f8 f2e3 
Engine&#58; info multipv 1 depth 9 seldepth 11 nodes 25628 time 161 nps 159180 hashfull 0 tbhits 0 score cp -33 pv g1f2 g8g7 e1d3 g7f6 d3f4 f6e5 f2e3 e8g8 f4d3 e5f5 
Engine&#58; info multipv 1 depth 10 seldepth 13 nodes 48802 time 271 nps 180081 hashfull 0 tbhits 0 score cp -27 pv g1f2 g8g7 e1d3 d7d5 g2g3 d5d4 d3f4 e8e3 f4d5 e3a3 
Engine&#58; info multipv 1 depth 11 seldepth 15 nodes 165867 time 309 nps 536786 hashfull 18 tbhits 0 score cp -23 pv g1f2 e8f8 e1f3 f8f4 h2h4 g8g7 g2g3 f4a4 f3e5 d7d6 e5c6 a4a2 h4h5 
Engine&#58; info multipv 1 depth 12 seldepth 18 nodes 425535 time 385 nps 1105285 hashfull 39 tbhits 0 score cp -17 pv g1f2 e8f8 f2e3 g8g7 g2g4 f8e8 e3f2 e8f8 f2g3 d7d5 h2h4 f8e8 g3f3 d5d4 e1c2 e8f8 f3e4 
Engine&#58; info multipv 1 depth 13 seldepth 18 nodes 516768 time 414 nps 1248231 hashfull 41 tbhits 0 score cp -17 pv g1f2 e8f8 f2e3 g8g7 g2g4 f8e8 e3f2 e8a8 h2h4 a8a2 f2f3 g7g6 e1d3 d7d6 d3f4 g6g7 h4h5 
Engine&#58; info multipv 1 depth 14 seldepth 21 nodes 888944 time 518 nps 1716108 hashfull 67 tbhits 0 score cp -17 pv g1f2 e8f8 f2e3 g8g7 g2g4 f8e8 e3f2 e8a8 h2h4 a8a2 f2f3 d7d6 h4h5 g7h6 e2e4 h6g5 e1d3 a2a4 
Engine&#58; info multipv 1 depth 15 seldepth 23 nodes 2264479 time 873 nps 2593904 hashfull 190 tbhits 0 score cp -16 pv g1f2 e8f8 e1f3 g8g7 g2g4 g7h6 h2h4 f8f4 f2g3 f4e4 f3g5 e4e7 g3f3 d7d5 e2e3 h6g6 f3f4 
Engine&#58; info multipv 1 depth 16 seldepth 25 nodes 4543007 time 1420 nps 3199300 hashfull 373 tbhits 4 score cp -13 pv g1f2 e8e4 f2f3 e4h4 h2h3 g8g7 e1d3 h4h8 d3f4 h8b8 e2e3 b8b1 h3h4 b1h1 f3g4 h1c1 h4h5 
Engine&#58; info currmove g1f1 currmovenumber 2 nodes 4543007 nps 3199300 hashfull 373 tbhits 4
Engine&#58; info currmove e1d3 currmovenumber 3 nodes 4543725 nps 3193060 hashfull 373 tbhits 4
Engine&#58; info currmove g2g4 currmovenumber 4 nodes 4545665 nps 3192180 hashfull 373 tbhits 4
Engine&#58; info currmove h2h4 currmovenumber 5 nodes 4551262 nps 3191628 hashfull 373 tbhits 4
Engine&#58; info currmove e2e3 currmovenumber 6 nodes 4552981 nps 3188362 hashfull 373 tbhits 4
Engine&#58; info currmove e2e4 currmovenumber 7 nodes 4553442 nps 3186453 hashfull 373 tbhits 4
Engine&#58; info currmove e1f3 currmovenumber 8 nodes 4553780 nps 3186689 hashfull 373 tbhits 4
Engine&#58; info currmove e1c2 currmovenumber 9 nodes 4554790 nps 3185167 hashfull 373 tbhits 4
Engine&#58; info currmove h2h3 currmovenumber 10 nodes 4556491 nps 3184130 hashfull 374 tbhits 4
Engine&#58; info currmove g2g3 currmovenumber 11 nodes 4556760 nps 3175442 hashfull 374 tbhits 4
Engine&#58; info currmove g1h1 currmovenumber 12 nodes 4557040 nps 3173426 hashfull 374 tbhits 4
Engine&#58; info depth 17
Engine&#58; info currmove g1f2 currmovenumber 1 nodes 4557239 nps 3171356 hashfull 374 tbhits 4
Engine&#58; info multipv 1 depth 17 seldepth 26 nodes 7051679 time 2069 nps 3408254 hashfull 491 tbhits 26 score cp -7 pv g1f2 e8f8 e1f3 g8g7 g2g4 g7h6 h2h4 f8f4 f2g3 f4e4 f3g1 h6g7 g3f3 d7d5 h4h5 g7h6 g1h3 e4c4 h3f4 c4c3 f4d3 h6g5 
Engine&#58; info currmove g1f1 currmovenumber 2 nodes 7051679 nps 3408254 hashfull 491 tbhits 26
Engine&#58; info currmove g2g4 currmovenumber 3 nodes 7052151 nps 3403547 hashfull 491 tbhits 26
Engine&#58; info currmove e1d3 currmovenumber 4 nodes 7052998 nps 3403956 hashfull 491 tbhits 26
Engine&#58; info currmove h2h4 currmovenumber 5 nodes 7054061 nps 3401186 hashfull 491 tbhits 26
Engine&#58; info currmove e2e3 currmovenumber 6 nodes 7055534 nps 3400257 hashfull 491 tbhits 26
Engine&#58; info currmove e2e4 currmovenumber 7 nodes 7056237 nps 3400596 hashfull 491 tbhits 26
Engine&#58; info currmove e1c2 currmovenumber 8 nodes 7057266 nps 3399453 hashfull 491 tbhits 26
Engine&#58; info currmove e1f3 currmovenumber 9 nodes 7058213 nps 3398272 hashfull 491 tbhits 26
Engine&#58; info currmove h2h3 currmovenumber 10 nodes 7058461 nps 3396756 hashfull 491 tbhits 26
Engine&#58; info currmove g2g3 currmovenumber 11 nodes 7059435 nps 3390698 hashfull 491 tbhits 26
Engine&#58; info currmove g1h1 currmovenumber 12 nodes 7060519 nps 3389591 hashfull 491 tbhits 26
Engine&#58; info depth 18
Engine&#58; info currmove g1f2 currmovenumber 1 nodes 7061156 nps 3388270 hashfull 491 tbhits 26
Engine&#58; info nodes 11228831 nps 3629227 hashfull 669 tbhits 38
Scid  &#58; stop
Scid  &#58; quit

[gaard]

Thanks for the release!

I have an issue when I am using MultiPV where the PVs are sometimes identical. It seems to be unrelated the Threads/Session file/Resolve score drops. I am using 64-bit version with POPCNT enabled and large page support.

info multipv 1 depth 15 seldepth 32 nodes 7607378 time 4275 nps 1779503 hashfull 36 tbhits 0 score cp -3 pv b6b4 b1b4 b8b4 a4a7 h7h6 g5f6 e7f6 g2e4 b4c4 f1e1 f6g5 a2a4 g5d2 e4d3 c4e6 e1d1 d2c3 a7b7 f7f5 d1b1 e5e4 b1b6

info multipv 2 depth 15 seldepth 32 nodes 7607378 time 4275 nps 1779503 hashfull 36 tbhits 0 score cp -3 pv b6b4 b1b4 b8b4 a4a7 h7h6 g5f6 e7f6 g2e4 b4c4 f1e1 f6g5 a2a4 g5d2 e4d3 c4e6 e1d1 d2c3 a7b7 f7f5 d1b1 e5e4 b1b6

info multipv 3 depth 15 seldepth 32 nodes 7607378 time 4275 nps 1779503 hashfull 36 tbhits 0 score cp -25 pv f6d5 g5e7 d5e7 f1c1 b6b5 a4b3 a7a5 a2a3 f8d8 g2e4 e7d5 e4d5 d8d5 c1c2 a5b4 b1c1 d5d8 c2c7 d8f8

info multipv 4 depth 15 seldepth 32 nodes 7607378 time 4275 nps 1779503 hashfull 36 tbhits 0 score cp -30 pv e7b4 g5f6 g7f6 a4d7 b8c8 d7a7 b6a6 a7b7 c8b7 g2b7 a6b6 b7e4 f8b8 f1c1 b4c3 a2a4 b6b2 g1g2 g8g7 b1b2 b8b2 c1c2 b2c2 e4c2 h7h5 c2e4

[stevenaaus]

Actually , using this fen and gtb, 8/8/8/8/4K3/2k5/4B3/8 w - - 0 34
there is a minor bug in ScidvsPC and Scid4.3 which often stops the pv from being shown at all.
Because critter gives the info so quickly , it happens before Scid has finished it's PV initialisation, and gets confused.
I have / am testing a fix for ScidvsPC. Scid git has some serious lag on its engines nowadays, and perhaps it is unaffected.