tiger wrote:
Thank you for the trust you put in my ability to improve Chess Tiger...
// Christophe
Don't take any notice of those comments Christophe.
Tiger is a fine engine and I believe that it will continue to get stronger..
Regards, Graham.
Thank you Graham.
I think Eriq did not mean to be harsh, but I think he just doesn't know me well enough.
BTW... you make it more and more difficult for me to express my criticism toward the CCRL list. But I will have to criticize it anyway.
// Christophe
Hi Christophe
I think your new update is quite improved It would have been
nice though, if you had mentioned that you also fixed CTs timing
bug, which caused time losses under WB, because CT allocated
sometimes more time than allowed in difficult positions.
(This did not happen under CB or Arena IIRC, because those
GUIs measure thinking time much less precise and in case of Arena
even allow negative time, which was proven by debugs long ago.
...The (in)famous Monarch case
Anyhow a silent bugfix is still a bugfix Thanks!
I don't regret to have bought it.
Regards,
Guenther
P.S. Any chance for a newer and up to date book at some time?
Yes I have changed the time management. The primary goal was to have a better time allocation, for example by taking more time at the begining of the game.
It seems that this change has also fixed the time management bug (overstepping the time control in X moves in Y minutes mode), but I am not sure, because I have not found this bug in the unmodified code.
So I did not want to claim a victory too fast and did not mention this fix.
Unfortunately the book will stay as it is for some time, because its author, Jeroen Noomen, now works for the Rybka team.
chessfurby wrote:Despite the friendly words and hopefully your honest intention ... to put it in a nutshell what you are saying is:
"Your engine is weaker then Rybka, so give up."
So forgive me my bluntness but ...
Maybe you could exclude us from any of your future opinions
Do you not agree with Erik? Well, me too, I think he was unhaply in his post, but I do not agree with
your words:" Maybe you could exclude us from any of your future opinions". Are you "banning" Erik from CCC?
Yes I have changed the time management. The primary goal was to have a better time allocation, for example by taking more time at the begining of the game.
It seems that this change has also fixed the time management bug (overstepping the time control in X moves in Y minutes mode), but I am not sure, because I have not found this bug in the unmodified code.
So I did not want to claim a victory too fast and did not mention this fix.
...
// Christophe
CT 2007.1 is still losing on time here under winboard. If I'm the only one then of course the problem is not in your code! The fischer time control works well though.
Good luck with the further development of Chess Tiger,
chessfurby wrote:Despite the friendly words and hopefully your honest intention ... to put it in a nutshell what you are saying is:
"Your engine is weaker then Rybka, so give up."
So forgive me my bluntness but ...
Maybe you could exclude us from any of your future opinions
Do you not agree with Erik? Well, me too, I think he was unhaply in his post, but I do not agree with
your words:" Maybe you could exclude us from any of your future opinions". Are you "banning" Erik from CCC?
Paulo Soares
Jens is a tester of Deep Sjeng. As we all know Sjeng is currently weaker than Tiger.
Yes I have changed the time management. The primary goal was to have a better time allocation, for example by taking more time at the begining of the game.
It seems that this change has also fixed the time management bug (overstepping the time control in X moves in Y minutes mode), but I am not sure, because I have not found this bug in the unmodified code.
So I did not want to claim a victory too fast and did not mention this fix.
...
// Christophe
CT 2007.1 is still losing on time here under winboard. If I'm the only one then of course the problem is not in your code! The fischer time control works well though.
Good luck with the further development of Chess Tiger,
I had none so far in 50 games, when I had around 1 per 10-15 games
before. It still could have been lucky though so far?
I play only ponder off games here. It would be easy to prove
if it is still Tigers fault, if you send me Tiger logs of those lost on time
games, because they clearly state that Tiger would have used
a few tenths of a second more than allowed in its last move.
Yes I have changed the time management. The primary goal was to have a better time allocation, for example by taking more time at the begining of the game.
It seems that this change has also fixed the time management bug (overstepping the time control in X moves in Y minutes mode), but I am not sure, because I have not found this bug in the unmodified code.
So I did not want to claim a victory too fast and did not mention this fix.
...
// Christophe
CT 2007.1 is still losing on time here under winboard. If I'm the only one then of course the problem is not in your code! The fischer time control works well though.
Good luck with the further development of Chess Tiger,
I had none so far in 50 games, when I had around 1 per 10-15 games
before. It still could have been lucky though so far?
I play only ponder off games here. It would be easy to prove
if it is still Tigers fault, if you send me Tiger logs of those lost on time
games, because they clearly state that Tiger would have used
a few tenths of a second more than allowed in its last move.
Regards,
Guenther
Hi Guenther,
I don't have logs right now (I'm too lazy to manage this usually) and I unthicked "Auto Flag" to make sure the games don't end prematurately. I still get "White Flag fell" or "Black Flag Fell" at least once in 10 games - usually 40/4- with Tiger 2007.1 .
Next time Tiger plays games, I'll try to think about saving .debug files and having Auto Flag = on.
mclane wrote:i have not one time out in all my tiger games. i do play under arena.
CB and Arena don't measure time usage that precisely. Arena
even allows explicetely negative time before the next cut.
The bug report about this is years old...
Tiger overstepped the time between 0.25 and 0.07 seconds
sometimes in difficult positions. Tiger itself logged this for
its last move and it was confirmed by Lokasoft via email.
The new version so far didn't overstep the time so far in 50
games. (moves/minutes)
It would have been
Thanks!
