Question for Norman Schmidt ...

[Frank Quisinsky]

Hi Norman,

which version should be play in SWCR?
1.32 or 1.40 xtreme?

On Intel Q9550 for SWCR.
Any important other thing I have to config?

Could you give me a little tip?
Later perhaps checks the configuration?

After the Junior 12.5.0.3 x64 Fire will follow!
I think Tuesday next week in the evening the test can run.

Best
Frank

[Damir]

Hi Frank

I think it is better we wait a little until the bugs are fixed.
The engine is crashing every time with error exception message.

[mwyoung]

Hi Frank

I think it is better we wait a little until the bugs are fixed.
The engine is crashing every time with error exception message.

I am testing Fire 1.4 x64p. I have no problems testing the program so far.

After 9 games Fire 1.4 is on top of Houdini 1.5a playing at TC = 60/5 +3 =4 -2

[Damir]

Maybe offline, but I think you should try and play with Fire on one of the chess servers like Playchess.

[kranium]

Hi Frank-

definitely 1.5 xTreme
it's proving a little stronger, and has fixes/improvements...

http://www.chess2u.com/t2403-fire-15-xtreme-tests#14388

(the two 1.4 releases were both 'beta' test versions, for which i got a lot a great test results and feedback)

i recommend using the 'default' settings...
at the moment, they are the best that I'm aware of.
eventually, of course somebody will find and share an improvement.

if/when that happens i'll let you know..

thanks much for your excellent effort/work!

[Frank Quisinsky]

Hi Norman,

ah, to late, saw your message to late.

Now BugChess2 1.9 x64 is still running in round robin 1.200 games. Fire 1.5 will be follow Monday or Tuesday next week after the BugChess2 1. 9 test.

BugChess2 1.9 is also clearly improved, saw it in 300 test games (here 78 ELO). But 300 games on my test systems only.

Very interesting on your Fire version. Should I gave Fire an own place in SWCR? Predecessor from IvanHoe I think? Do you think that Fire 1.5 and IvanHoe are to different, so I will gave Fire an own place.

In SWCR Fire 1.31 don't have an own place because the best IvanHoe I tested is stronger.

Hope you understand my problem ...
Bright - Spark (one SWCR place)
Glaurung - Stockfish (one SWCR place)
Twisted Logic - Hannibal (one SWCR place)
Fire - IvanHoe (one SWCR place)
Fruit - Cyclone (one SWCR place)

Best
Frank

Thanks for your tip to the configuration and version!
The configuration I will be use - default - next week for the test can be found in my download area (here you can found all configuration from the engines I tested in SWCR).

[kranium]

Hi Norman,

ah, to late, saw your message to late.

Now BugChess2 1.9 x64 is still running in round robin 1.200 games. Fire 1.5 will be follow Monday or Tuesday next week after the BugChess2 1. 9 test.

BugChess2 1.9 is also clearly improved, saw it in 300 test games (here 78 ELO). But 300 games on my test systems only.

Very interesting on your Fire version. Should I gave Fire an own place in SWCR? Predecessor from IvanHoe I think? Do you think that Fire 1.5 and IvanHoe are to different, so I will gave Fire an own place.

In SWCR Fire 1.31 don't have an own place because the best IvanHoe I tested is stronger.

Hope you understand my problem ...
Bright - Spark (one SWCR place)
Glaurung - Stockfish (one SWCR place)
Twisted Logic - Hannibal (one SWCR place)
Fire - IvanHoe (one SWCR place)
Fruit - Cyclone (one SWCR place)

Best
Frank

Thanks for your tip to the configuration and version!
The configuration I will be use - default - next week for the test can be found in my download area (here you can found all configuration from the engines I tested in SWCR).

sorry, i don't know anything about these 'places', or how that works,
or how you testers arrive at such arbitrary decisions,
but i do understand the diffculty with allocating time and resources.

if you want to take only one from Fruit family...i recommend Rybka
if you want to take only one from Ippolit family...i recommend Houdini

if you want to choose one between Fire and IvanHoe, i suggest you run a long match to see which is better,
or simply pick the one with the coolest name!

Norm

[Frank Quisinsky]

Hi Norman,

a little problem in presentation of my list.
Perhaps I should delete the places I set.
In the following example (place 1 for Houdini, 2 for Rybka, 3 for IvanHoe).

Look here ...

Code: Select all

   1    1 Houdini 1.5 x64                3003   17   16  1640   80%  2765   27% 
   2    2 Rybka 4 x64 Exp. 42            2970   20   19  1240   80%  2720   25% 
        3 Rybka 4.1 x64 Exp. 79TD v.1    2965   20   19  1200   79%  2738   26% 
        4 Rybka 4 x64 Exp. 61            2962   21   21  1000   78%  2744   27% 
        5 Rybka 4.1 x64                  2956   20   19  1040   77%  2754   32% 
        6 Houdini 1.03a x64              2946   21   21  1000   80%  2713   30% 
   3    7 IvanHoe B47cB x64              2945   19   19  1080   75%  2762   36% 
        8 Rybka 4 x64                    2940   17   17  1520   80%  2701   29% 
        9 IvanHoe B49jA x64              2934   19   18  1160   76%  2740   34% 
       10 IvanHoe B52aC x64              2920   20   20  1000   76%  2728   34% 
   4   11 Stockfish 1.8.0 JA x64         2907   18   18  1200   75%  2716   33% 

Perhaps better could be ...

Code: Select all

    1 Houdini 1.5 x64                3003   17   16  1640   80%  2765   27% 
    2 Rybka 4 x64 Exp. 42            2970   20   19  1240   80%  2720   25% 
    3 Rybka 4.1 x64 Exp. 79TD v.1    2965   20   19  1200   79%  2738   26% 
    4 Rybka 4 x64 Exp. 61            2962   21   21  1000   78%  2744   27% 
    5 Rybka 4.1 x64                  2956   20   19  1040   77%  2754   32% 
    6 Houdini 1.03a x64              2946   21   21  1000   80%  2713   30% 
    7 IvanHoe B47cB x64              2945   19   19  1080   75%  2762   36% 
    8 Rybka 4 x64                    2940   17   17  1520   80%  2701   29% 
    9 IvanHoe B49jA x64              2934   19   18  1160   76%  2740   34% 
   10 IvanHoe B52aC x64              2920   20   20  1000   76%  2728   34% 
   11 Stockfish 1.8.0 JA x64         2907   18   18  1200   75%  2716   33% 

More easy!

Best
Frank

[kranium]

Hi Norman,

a little problem in presentation of my list.
Perhaps I should delete the places.

Look here ...

Code: Select all

   1    1 Houdini 1.5 x64                3003   17   16  1640   80%  2765   27% 
   2    2 Rybka 4 x64 Exp. 42            2970   20   19  1240   80%  2720   25% 
        3 Rybka 4.1 x64 Exp. 79TD v.1    2965   20   19  1200   79%  2738   26% 
        4 Rybka 4 x64 Exp. 61            2962   21   21  1000   78%  2744   27% 
        5 Rybka 4.1 x64                  2956   20   19  1040   77%  2754   32% 
        6 Houdini 1.03a x64              2946   21   21  1000   80%  2713   30% 
   3    7 IvanHoe B47cB x64              2945   19   19  1080   75%  2762   36% 
        8 Rybka 4 x64                    2940   17   17  1520   80%  2701   29% 
        9 IvanHoe B49jA x64              2934   19   18  1160   76%  2740   34% 
       10 IvanHoe B52aC x64              2920   20   20  1000   76%  2728   34% 
   4   11 Stockfish 1.8.0 JA x64         2907   18   18  1200   75%  2716   33% 

Perhaps better could be ...

Code: Select all

    1 Houdini 1.5 x64                3003   17   16  1640   80%  2765   27% 
    2 Rybka 4 x64 Exp. 42            2970   20   19  1240   80%  2720   25% 
    3 Rybka 4.1 x64 Exp. 79TD v.1    2965   20   19  1200   79%  2738   26% 
    4 Rybka 4 x64 Exp. 61            2962   21   21  1000   78%  2744   27% 
    5 Rybka 4.1 x64                  2956   20   19  1040   77%  2754   32% 
    6 Houdini 1.03a x64              2946   21   21  1000   80%  2713   30% 
    7 IvanHoe B47cB x64              2945   19   19  1080   75%  2762   36% 
    8 Rybka 4 x64                    2940   17   17  1520   80%  2701   29% 
    9 IvanHoe B49jA x64              2934   19   18  1160   76%  2740   34% 
   10 IvanHoe B52aC x64              2920   20   20  1000   76%  2728   34% 
   11 Stockfish 1.8.0 JA x64         2907   18   18  1200   75%  2716   33% 

More easy!

Best
Frank

Honestly, the latter makes a lot more sense to me...
of course i'm not a genealogist,
so, IMO the exact DNA 'gene pool' of each and every participant is pretty irrelevant.

Since Rybka has been tested for years, (as Houdini is now also), it seems the major testing groups and most individuals agree.

of course, self-pronounced 'original' engine programmers will cry 'foul'...their engine will now be much further down,
i.e. and the list will have been polluted with engines of dubious 'pedigree'...

[Frank Quisinsky]

Hi,

yes, and each one see it with an other look.
I think I will delete my "own" places in the list with the next update I have to make.

Have a nice day!

Best
Frank