Fire 5 is out!

[velmarin]

The executable compressed with MPRESS causes that some antivirus jump the alarms...

Apart and following fashion , the similarity is completely isolated from Fire 3 and very close to Stockfish, specifically version 6.

Code: Select all

sim version 3
------ Fire 5 x64 (time: 100 ms  scale: 1.0) ------
 52.31  Stockfish 6 64 POPCNT (time: 100 ms  scale: 1.0)
 50.67  Deep Synapse RZ4.8 64bit (time: 100 ms  scale: 1.0)
 49.81  Stockfish 7 64 POPCNT (time: 100 ms  scale: 1.0)
 48.81  Stockfish 8 64 POPCNT (time: 100 ms  scale: 1.0)
 45.96  Fritz 15 (time: 100 ms  scale: 1.0)
 44.89  komodo 9.2 64-bit  (time: 100 ms  scale: 1.0)
 44.88  Rybka 4.1 (time: 100 ms  scale: 1.0)
 44.68  Fire 3.0 x64 (time: 100 ms  scale: 1.0)

Ed Schröder wrote about sim test results in 2014:
"In general one can say that 60% and higher is a sign of non-originality, 65% and higher being a strong sign of cloning, that the grey area is from 55-60% and any engine below 55% has passed the test, meaning just a first indication of originality."

I never said that there is nothing wrong.
But if removed the trees see the forest

What is clear that already know of which foot limps...
There is not so much work, good, copy if work...

Code: Select all

------ Fire 5 x64 (time: 100 ms  scale: 1.0) ------
 52.31  Stockfish 6 64 POPCNT (time: 100 ms  scale: 1.0)

 44.68  Fire 3.0 x64 (time: 100 ms  scale: 1.0)

[kranium]

sim version 3
------ Bouquet 1.8 x64 (time: 100 ms scale: 1.0) ------
65.51 IvanHoe 9.46b x64 (time: 100 ms scale: 1.0)
52.42 Komodo 6 64-bit (time: 100 ms scale: 1.0)
51.77 Stockfish 4 64 SSE4.2 (time: 100 ms scale: 1.0)
51.18 Stockfish 5 64 SSE4.2 (time: 100 ms scale: 1.0)
50.30 Gull 3 x64 (time: 100 ms scale: 1.0)

From 2014, also interesting...it's unfortunate I don't have data anymore that includes open-source Fire 2.2, as I remember,
Bouquet similarity to Fire was around 69%.

[Laskos]

The executable compressed with MPRESS causes that some antivirus jump the alarms...

Apart and following fashion , the similarity is completely isolated from Fire 3 and very close to Stockfish, specifically version 6.

Code: Select all

sim version 3
------ Fire 5 x64 (time: 100 ms  scale: 1.0) ------
 52.31  Stockfish 6 64 POPCNT (time: 100 ms  scale: 1.0)
 50.67  Deep Synapse RZ4.8 64bit (time: 100 ms  scale: 1.0)
 49.81  Stockfish 7 64 POPCNT (time: 100 ms  scale: 1.0)
 48.81  Stockfish 8 64 POPCNT (time: 100 ms  scale: 1.0)
 45.96  Fritz 15 (time: 100 ms  scale: 1.0)
 44.89  komodo 9.2 64-bit  (time: 100 ms  scale: 1.0)
 44.88  Rybka 4.1 (time: 100 ms  scale: 1.0)
 44.68  Fire 3.0 x64 (time: 100 ms  scale: 1.0)

Ed Schröder wrote about sim test results in 2014:
"In general one can say that 60% and higher is a sign of non-originality, 65% and higher being a strong sign of cloning, that the grey area is from 55-60% and any engine below 55% has passed the test, meaning just a first indication of originality."

I never said that there is nothing wrong.
But if removed the trees see the forest

What is clear that already know of which foot limps...
There is not so much work, good, copy if work...

Code: Select all

------ Fire 5 x64 (time: 100 ms  scale: 1.0) ------
 52.31  Stockfish 6 64 POPCNT (time: 100 ms  scale: 1.0)

 44.68  Fire 3.0 x64 (time: 100 ms  scale: 1.0)

52% is low. Basically you show that Fire 5 is original, at least in eval, thanks.

[supersharp77]

Fire 5 is available for download at http://chesslogik.wixsite.com/fire/download. Please test! Many thanks to Norman Schmidt for his hard work!!

w32 compile.....not working.......XP AR....

[Marek Soszynski]

PS - I just updated the download archive to include a 64-bit version that should run in Linux via wine, but it's not well tested.

Appreciated, though a non-popcnt-x64-wine would be handy as the regular x64 doesn't work (on my system) while the w32 does.

[CositasBuenas]

Thanx, but the wine version is not running on linux mint 18.

[drj4759]

Fire 5 does not work on all my Linux/WINE AMD 4-cores 8GB either.

Distro: PCLinuxOS 2016
Kernel: Linux 4.8.4
WINE: 1.8 staging
Arena GUI: 3.5

[Lyudmil Tsvetkov]

Bitdefender says the webpage and the link is trojan.

I can assure you it's not...

PS - I just updated the download archive to include a 64-bit version that should run in Linux via wine, but it's not well tested.

Norman, to give you some good advice for Fire 6: look very carefully at your king safety, Fire 5 loses about 70% of its games (based on what I have watched on TCEC) due to its poor king safety. Fire tends not to castle, frequently staying in the center. Whether it is due to some kind of castling rigth bonus or poor king mg psqt, I do not know, but the problem is quite evident and recurrent.

[Colin-G]

The 32bit executable does not work in linux Mint 17 with wine-1.6.2
I tried with both an old Pentium 4 cpu and on a computer with a modern AMD cpu
It will report the UCI options, but will not make a move in Arena GUI or Xboard (using polyglot)
(I changed the spaces in the file name to underscores)

Code: Select all

colin@PB2009 ~ $ cd .wine/drive_c/Arena/Engines/Fire
colin@PB2009 ~/.wine/drive_c/Arena/Engines/Fire $ wine Fire_5_w32.exe 
Fire 5 w32
uci
id name Fire 5 w32
id author Norman Schmidt
option name Hash type spin default 128 min 1 max 2048
option name Threads type spin default 1 min 1 max 64
option name MultiPV type spin default 1 min 1 max 64
option name Contempt type spin default 0 min -100 max 100
option name Ponder type check default false
option name UCI_Chess960 type check default false
option name Syzygy50MoveRule type check default true
option name SyzygyPath type string default <empty>
option name SyzygyProbeDepth type spin default 8 min 1 max 64
option name SyzygyProbeLimit type spin default 6 min 0 max 6
uciok

[kranium]

Thanks all, for letting me know about the wine issue, I'll spend time on in the next day or two to see if I can provide a functional executable for you Linux users.

Since Windows XP is now ancient, (almost 14 years old), I'm loathe to spend a lot of time figuring out the problem...