Nirvanachess 1.0 Release

[cetormenter]

Hello everybody,

I have decided to release my engine which has been a little hobby of mine for the past few years. I know that a few of you are interested in amateur engines and I hope that this engine is welcomed contribution.

http://nirvanachess.blogspot.com/

Cheers,

Thomas

[Adam Hair]

Thanks for sharing, Thomas. It is always great when a new author shares his work with us.

[Graham Banks]

Hello everybody,

I have decided to release my engine which has been a little hobby of mine for the past few years. I know that a few of you are interested in amateur engines and I hope that this engine is welcomed contribution.

http://nirvanachess.blogspot.com/

Cheers,

Thomas

Thanks Thomas. What would be a rough estimation of its strength?

Graham.

[jshriver]

I look forward to testing it Thanks for the release.

-Josh

[tmokonen]

Hi Thomas. I notice that when I run Nirvana Chess, it sends uciok before it sends option name Hash type spin default 32 min 1 max 256. The uciok should be sent after the option, so that the Hash parameter works properly, at least in Arena.

[jd1]

Hello everybody,

I have decided to release my engine which has been a little hobby of mine for the past few years. I know that a few of you are interested in amateur engines and I hope that this engine is welcomed contribution.

http://nirvanachess.blogspot.com/

Cheers,

Thomas

Thanks Thomas. What would be a rough estimation of its strength?

Graham.

From the website
~ 2350 CCRL

[cetormenter]

Hi Thomas. I notice that when I run Nirvana Chess, it sends uciok before it sends option name Hash type spin default 32 min 1 max 256. The uciok should be sent after the option, so that the Hash parameter works properly, at least in Arena.

Whoops! I never actually used this option. I always just tested at the default of 32 MB.

New Link:
https://mega.co.nz/#!iBZhSLaJ!NvvChDJrN ... CavWmjYPy4

Thanks Thomas. What would be a rough estimation of its strength?

Graham.

From my lastest tournament run at very quick time controls (I don't have enough computing cycles to do long time control tests).

Code: Select all

Cechess - AnMon 5.75         : 10.0/42 7-29-6   24%  -200
Cechess - BigLion            : 37.0/41 35-2-4  90%  +382
Cechess - Dragon 4.6         : 17.0/41 14-21-6  41%   -63
Cechess - Fruit_21           : 3.5/41 0-34-7    9%  -402
Cechess - Hermann 2.6 64 bit : 8.0/41 5-30-6   20%  -241
Cechess - SOS 5.1 for Arena  : 14.5/41 12-24-5  35%  -108

Using the ratings from CCRL 40/4 this gives an estimated rating of 2354 +- 41. However I believe this is an over estimation because it seems that both Dragon and SOS have issues with very fast time controls. Also, Nirvanachess is unable to utilize endgame table bases, which is where it usually ends up losing the game. I am not sure how common it is for engines at this rating to be able to use them.

[tmokonen]

Thank you for your quick response, Thomas, it works well now.

[geots]

Thank you for your quick response, Thomas, it works well now.

Tony, this new link takes me to a page that has no engine in site to download and nothing to click on that would take me to one. ?????


george

[geots]

Hi Thomas. I notice that when I run Nirvana Chess, it sends uciok before it sends option name Hash type spin default 32 min 1 max 256. The uciok should be sent after the option, so that the Hash parameter works properly, at least in Arena.

Whoops! I never actually used this option. I always just tested at the default of 32 MB.

New Link:
https://mega.co.nz/#!iBZhSLaJ!NvvChDJrN ... CavWmjYPy4

Thanks Thomas. What would be a rough estimation of its strength?

Graham.

From my lastest tournament run at very quick time controls (I don't have enough computing cycles to do long time control tests).

Code: Select all

Cechess - AnMon 5.75         : 10.0/42 7-29-6   24%  -200
Cechess - BigLion            : 37.0/41 35-2-4  90%  +382
Cechess - Dragon 4.6         : 17.0/41 14-21-6  41%   -63
Cechess - Fruit_21           : 3.5/41 0-34-7    9%  -402
Cechess - Hermann 2.6 64 bit : 8.0/41 5-30-6   20%  -241
Cechess - SOS 5.1 for Arena  : 14.5/41 12-24-5  35%  -108

Using the ratings from CCRL 40/4 this gives an estimated rating of 2354 +- 41. However I believe this is an over estimation because it seems that both Dragon and SOS have issues with very fast time controls. Also, Nirvanachess is unable to utilize endgame table bases, which is where it usually ends up losing the game. I am not sure how common it is for engines at this rating to be able to use them.

Thomas, I download from the first link you give, I get an 85kb exe. Then I see you talking about a problem- I download from the new link you gave- I get the exact same 85kb executable. Could you please tell me what is going on. I have no idea what or which to use. Basically I am totally confused.


george