Gaviota display suggestion

[mhalstern]

Miguel A. Ballicora,

I saw you reply to my gauntlet post. Not a problem - considering the strong engines, Gaviota performed very well. I became interested in your engine, because you were posting about the new 5 men tablebases that you created. This was an incredible undertaking and you should be commended.

I know that the compressed tablebases are working. Tbtest reports:

tbtest 5

CHECKING LONGEST MATE POSITIONS FROM TABLEBASES
221 positions checked, wrong: 0, correct: 221
tablebases test ----------------------------> PASSED

I also manually checked a number of fen positions which were reported correctly.

When gaviota is playing, the user does not see any evidence that tablebases are being accessed.

When other engines are configured to query nalimov tablebases, they will report things such as "tb=700," indicating that the tablebases were accessed 700 times.

When watching gaviota play the gauntlet, I was a little worried that the tablebases were not being accessed, as I did not see any similar output.

[michiguel]

Miguel A. Ballicora,

I saw you reply to my gauntlet post. Not a problem - considering the strong engines, Gaviota performed very well. I became interested in your engine, because you were posting about the new 5 men tablebases that you created. This was an incredible undertaking and you should be commended.

I know that the compressed tablebases are working. Tbtest reports:

tbtest 5

CHECKING LONGEST MATE POSITIONS FROM TABLEBASES
221 positions checked, wrong: 0, correct: 221
tablebases test ----------------------------> PASSED

I also manually checked a number of fen positions which were reported correctly.

When gaviota is playing, the user does not see any evidence that tablebases are being accessed.

When other engines are configured to query nalimov tablebases, they will report things such as "tb=700," indicating that the tablebases were accessed 700 times.

When watching gaviota play the gauntlet, I was a little worried that the tablebases were not being accessed, as I did not see any similar output.

Is that shown in the principal variation? I guess I could do that.

With that strong opposition, one way to know Gaviota is using TBs is when it starts to announce is being mated in 53

Another thing you could do is (in winboard) to go to the menu and try to see the "book move" when you are in a position with 5 or fewer pieces. It will tell you all the information about that position

You can seethe picture at the bottom:
http://sites.google.com/site/gaviotache ... blebases-1

Miguel

[mhalstern]

The Tablebase hits are displayed in the field where the Number of nodes (either per second or total, depending on the setting) is displayed.