Danasah beta 2.89

[Tony Thomas]

I cant seem to download it, can someone send it to me with the source Code? All I am getting is an empty file.

Tony

My email address is apocalypse112at aol.com

[Olivier Deville]

I cant seem to download it, can someone send it to me with the source Code? All I am getting is an empty file.

Tony

My email address is apocalypse112at aol.com

Sent

Olivier

[Pedro Castro]

From version 2.85 (although it does well in CRCL) I have some problem when the program accedes to bitbases 5 men, I want to remove a version with some small modifications.

Version 2.89 win in a test to the 2.85, but I believe that she makes against others engines worse, in addition each 30 games the program causes an error (crash). For that reason for games in tournaments I prefer that it continues playing the 2.85.

[Tony Thomas]

From version 2.85 (although it does well in CRCL) I have some problem when the program accedes to bitbases 5 men, I want to remove a version with some small modifications.

Version 2.89 win in a test to the 2.85, but I believe that she makes against others engines worse, in addition each 30 games the program causes an error (crash). For that reason for games in tournaments I prefer that it continues playing the 2.85.

I thought the performance of Danasah 2.85 was so phenomenal that I didnt expect any improvements in such a short period of time. I am still playing with 2.85, it has close to 250 games in a database I am creating mainly to test beta versions of Romichess.

[Pedro Castro]

In the last month I have not been able to improve.

The 2.89beta version does very badly in short times where the depths search are small, Michael (Romichess) already has demonstrated it with 2 test against versions 2.85 and 2.89 with times of 1+1, in longer times the thing is perhaps more even.

I am discovered some problems in the 2.85, that perhaps do not affect in your test, if the program proves 5 men of bitbases in short times can lose per time, Olivier has already sent me some game with this problem and also I have corrected a problem in the endgame with great depths the program can do crash, commented me in an occasion George Lyapko but I did not know to see it.

Perhaps in the next days I put a version bugfix but without improving the game.

[Tony Thomas]

In the last month I have not been able to improve.

The 2.89beta version does very badly in short times where the depths search are small, Michael (Romichess) already has demonstrated it with 2 test against versions 2.85 and 2.89 with times of 1+1, in longer times the thing is perhaps more even.

I am discovered some problems in the 2.85, that perhaps do not affect in your test, if the program proves 5 men of bitbases in short times can lose per time, Olivier has already sent me some game with this problem and also I have corrected a problem in the endgame with great depths the program can do crash, commented me in an occasion George Lyapko but I did not know to see it.

Perhaps in the next days I put a version bugfix but without improving the game.

I do not use bitbases or Tablebases...just my own rule, I do have my reasons, but they arent solid, they change everyday. At first it was my slow internet connection, then it was the fact that most of the time my blitz games end way before it get to 4 man or 5 man, then there are other reasons as well.