Author: Robert Hyatt
Date: 08:37:13 10/29/02
Go up one level in this thread
On October 29, 2002 at 05:51:43, Vincent Diepeveen wrote: >On October 28, 2002 at 11:40:00, Robert Hyatt wrote: > >and what if suddenly DIEP forfeits against a program 'F', >because it gives a move 'hahaBf5'. And exactly _how_ would that happen? The FICS server validity-checks each move before relaying it to the opponent. The opponent can _not_ send anything directly to my program... > >Prints itself onto its own screen that it played Bf5, diep >interprets it as bf5 and plays Rxh5. > >Who'se going to lose on clock now? Nobody because it can't happen. Ever had that happen playing on ICC? Didn't think so. It couldn't happen in a tournament either... > >This is daily business with auto232 protocol. Now i know you >find the protocol sucks, but it most definitely is great >for easy connecting machines. I'm not suggesting using that piece of trash. I am suggesting using a chess server in the tournament playing hall, running the FICS server code. Everybody uses any interface they want that can talk to the chess server, which pretty well includes most commercial interfaces plus xboard/winboard/etc... > >Of course i know your solution: make a server solve all this. > >But the average tournament organization can't even get to work >internet properly, not to mention networks. Nobody is saying "use internet". I am saying use an FICS-type server on a local machine. If they don't have the skills to set that up, which is really not very complex, then hire someone to do it once and it is done forever for them. > >If you can do it, why not organize a great new tournament in the >USA at your university? I have _already_ set up FICS years ago. I used it for testing for a long time... It isn't a problem... > >>On October 28, 2002 at 10:31:03, Vincent Diepeveen wrote: >> >>>On October 28, 2002 at 10:05:29, Uri Blass wrote: >>> >>>>On October 28, 2002 at 09:44:40, Vincent Diepeveen wrote: >>>> >>>>>On October 28, 2002 at 09:38:57, Vincent Diepeveen wrote: >>>>> >>>>>I hope you realize that i prefer a time control of 40 in 2. >>>>>However, if you play blitz levels like 90 0, then you have >>>>>consequences like this. If you asked me in advance, then >>>>>i would not have told you to expect that i would get to >>>>>move 187 at all in a computer-computer game which was >>>>>hand operated. >>>>> >>>>>Both operators played like 50 moves extra with very little time >>>>>on the clock. >>>>> >>>>>If you do not find it fair to win on time, then don't play 90 0 >>>>>is my viewpoint. I prefer 40 in 2 myself too. If you go 90 0, >>>>>it is unpreventable that a number of games end in a time win. >>>> >>>>It is possible to prevent it if you do not play games manually. >>>> >>>>I see no reason to play games manually. >>>>I see no reason not to play the games automatically in ICC. >>>> >>>>Uri >>> >>>But this is why your movei will have problems progressing >>>like GCP is progressing fast now with Sjeng. He talks to the >>>programmers live and he is there to operate his program. >>> >>>I will never trust it if we use some kind of protocol where i >>>cannot see my opponent. There is too much money at stake with >>>some people to just *assume* they play honest if you can't see >>>them. >> >>That is all well and good, but it wasn't Uri's point. You can attend a >>tournament, >>but _not_ manually operate. There is absolutely nothing wrong with requiring >>that a programmer attend, if you want. But there is also absolutely nothing >>wrong >>with an automatic interface to actually play the games. And in fact, there is a >>lot >>_right_ with it as it eliminates a significant opportunity for cheating... >> >> >>> >>>May i remind you the FIDE world champs qualifier which was played >>>online? Can you please analyze the games of the first 8 who >>>qualified. How many of them look non-computer helped to you? >>> >>>If you trust all those GMs, how many of the games the GM in person >>>or even FM in person, played an opening he normally plays too? >>> >>>Isn't it weird that some games resemble some Top GMs, who didn't >>>need to qualify, a lot? >>> >>>If you still are under the impression that this wouldn't happen >>>in computerchess, then let me just tell you that for those GMs >>>only up to $10000 was at stake, in order to qualify for moscow >>>2001. In computerchess for the commercials you can add a few zero's >>>to it for some. >>> >>>You still reading this? >>> >>>Best regards, >>>Vincent
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.