Peter Skinner wrote:With the Ktulu situation, Edwin was having a horrible time getting Ktulu to work with the latest Winboard version and the latest Polyglot. Over 5 rounds Edwin must have tried 20+ different combinations fo Winboard + Polyglot and configuration files that I was sending him. (We he downloaded them from my site as GMail will not accept an exe file even if zipped.) I must apologize to some of Edwin's opponents, as I was specifically asking him to disconnect during games to try one or two different settings to get kibitzing to work. In round 5, we finally hit the jackpot, and Ktulu was kibitzing normally, and operating well.
I don't understand how this can happen when people have one week to test, and the rules tht prescribe kibitz rather than wisper are known 6 months in advance... Besides, there is an installer contaning a working version of WinBoard and Polyglot available from the standard distribution site, which should run UCI engines and kibitz without any problems. This story really sounds like massive incompetence... Didn't we have rules that operators should actually be acquainted with the program, and know how to use it?
Maybe it would be a good idea holding a preliminary 1-round-tournament one week before the real tournament starts just to check that every participant is following/understanding the rules. There would be enough time to sort out any remaining problems.
Solving problems while the tournament is going on is a bit late ...
Peter Skinner wrote:With the Ktulu situation, Edwin was having a horrible time getting Ktulu to work with the latest Winboard version and the latest Polyglot. Over 5 rounds Edwin must have tried 20+ different combinations fo Winboard + Polyglot and configuration files that I was sending him. (We he downloaded them from my site as GMail will not accept an exe file even if zipped.) I must apologize to some of Edwin's opponents, as I was specifically asking him to disconnect during games to try one or two different settings to get kibitzing to work. In round 5, we finally hit the jackpot, and Ktulu was kibitzing normally, and operating well.
I don't understand how this can happen when people have one week to test, and the rules tht prescribe kibitz rather than wisper are known 6 months in advance... Besides, there is an installer contaning a working version of WinBoard and Polyglot available from the standard distribution site, which should run UCI engines and kibitz without any problems. This story really sounds like massive incompetence... Didn't we have rules that operators should actually be acquainted with the program, and know how to use it?
Actually the problem was the 4.4.2cct version.
When we finally went back to Winboard 4.4.2b and a previous version of polyglot, everything went fine.
Even though we had set the number of engines in the options panel, Ktulu would not use the book correctly, and wouldn't kibitz.
It was an add situation. Someone else stated they had to do the same, being going back to version 4.4.2b.
Peter
I was kicked out of Chapters because I moved all the Bibles to the fiction section.
You are completely right MR Muller. However this is not the first time Ktulu operator has pulled something like this off. On the Playchess server where I play TD had to wait for him several days, because of his '' unstable'' connection postponing the game, making all the other participants wait. Nice to know that even on FICS, he can not stay away from trouble...
Peter Skinner wrote:Actually the problem was the 4.4.2cct version.
When we finally went back to Winboard 4.4.2b and a previous version of polyglot, everything went fine.
Even though we had set the number of engines in the options panel, Ktulu would not use the book correctly, and wouldn't kibitz.
It was an add situation. Someone else stated they had to do the same, being going back to version 4.4.2b.
Well, there is virtually no difference between 4.4.2b and 4.4.CCT12. And the fixes that were made are mostly cosmetic. Like making sure exactly the correct stuff was suppressed in the console window, when using -autoKibitz and chat boxes. (In 4.4.2 partial lines that should have been suppressed could be printed, and continution lines were not captured together with the line they belonged to in kibitz or chat.) But this should have no effect on operation, just on the display.
So what you say is a bit hard to believe. Are you sure the 'someone else' you are taking about did not switch back from 4.20100118 or some other development version? That indeed invokes Polyglot in a completely different way, which would not work with old Polyglots.
So it is more likely that whatever changed, was due to the Polyglot substitution.
Didn't we have rules that operators should actually be acquainted with the program, and know how to use it?
