Moderators: hgm, Rebel, chrisw
I can't agree more.hgm wrote:This tournament really sucks. Next time we should organize it in a much more professional way: if a higher-rated profgram loses to a lower-rated one, we sould simply replay the game. If need be, severl times. At least as often as needed to get a correct result for the game.
Better yet, we could also move to a server where lower-rated programs are not even allowed to log in. Then we would not have this problem at all. If we only allow Rybka, in enough copies to make it a meaningful tournament. Then we would never be embarrased by having incorrect results, and everyone would take us serious again!
You're missing the point. It's not about being allowed to play on the server but about being able to play there. Unless something has changed recently, Chessbase hasn't published their protocol. That they allow me to connect to their server doesn't help much when they keep all information about how to interact with the server secret.Damir wrote:You can create a nick there for 30 days, after that it will expire, and you would no longer be able to use it unless you renew it by buying a serial from Chessbase...
Just try to make your own CCT. Wish you success.Damir wrote:why not try and make a vote ?
This was not a protocol issue at all. Operator tried to "end" the GUI. GUI asked "do you want to resign this game before logging out?" Operator clicked "yes"...Tord Romstad wrote:The losing side was not awarded a win because the winning side crashed: It was awarded because the winning side resigned. Apparently, there was some sort of bug in the program that caused it to automatically resign when disconnected. That a program loses because a bug causes it to resign is no more unfair than that a program loses because a bug causes it to make a bad move.Damir wrote:It seems that CCT11 event is not stable and should never have been run in the 1st place, if such games can occure where the losing side should be awarded a win because of engine error or crash on the benefit of another engine.....
The TD's decision was therefore entirely correct, but I think this incident does illustrate what I'm been trying to say earlier: The ICC and its protocol are designed for humans, and are unnecessarily difficult for computer players. There are too many technical problems. I would like to see a special server and protocol designed exclusively for running official computer vs computer tournaments, and where an effort is made to make everything as simple as possible for the participants.
Tord
What _are_ you talking about? Every game that is lost is lost because of an "engine error". So I somehow miss the point. Computers crash at the WCCC "in-person" events regularly. This is not new, nor unexpected, nor avoidable.Damir wrote:It seems that CCT11 event is not stable and should never have been run in the 1st place, if such games can occure where the losing side should be awarded a win because of engine error or crash on the benefit of another engine..... It happened in previous event, and it still happens.
Should not the TD make sure such thing no longer occurs, as he is the one responsible, that everything runs smoothly ?
Such results give wrong impressions to the audience.
ICC has been around 3x as long as playchess. ICC works with _open_ gui specifications, playchess works with a commercial interface. ICC had _nothing_ to do with this issue today, it was the participant program that had a problem...Damir wrote:Why not move this event on the Playchess server, since it is much more stable than the one on ICC ??
You don't need to play that many games, of course. however, the number was not exaggerated at all. We actually played that many games last week during final testing and tuning.swami wrote:Damir wrote:because it happens way too often, and not enough has been done to prevent such thing??Harvey Williamson wrote:How does an engine crash make the CCT11 event unstable?Damir wrote:It seems that CCT11 event is not stable and should never have been run in the 1st place, if such games can occure where the losing side should be awarded a win because of engine error or crash on the benefit of another engine..... It happened in previous event, and it still happens.
Should not the TD make sure such thing no longer occurs, as he is the one responsible, that everything runs smoothly ?
Such results give wrong impressions to the audience.
Hyatt(64): I have a solution for those kinds of bugs. Play about 5 million games a week and you wont have 'em.
Ofcourse Bob is exaggerating but he put it quite right.
You show a complete lack of understanding about what is going on. ICC didn't crash today. Rybka was crashing. This was a Rybka problem. Not an ICC problem.Damir wrote:Let's move CCT11 event on Playchess Server, than we will see how many crashes will be there... My guess is none