Author: Robert Hyatt
Date: 05:15:14 02/25/99
Go up one level in this thread
On February 25, 1999 at 02:20:39, Bruce Moreland wrote: > >On February 24, 1999 at 22:06:34, Robert Hyatt wrote: > >>On February 24, 1999 at 17:48:10, Vincent Diepeveen wrote: >> >>>On February 24, 1999 at 16:04:41, Dann Corbit wrote: >>> >>>>I have an email from Frederic Friedel, and I am fully convinced that Fritz did >>>>not 'cook' anything to get the answers right. >>> >>>I got the same email from Frederic and i'm completely convinced that Fritz >>>did cook. >>> >>>Frederic is playing the innocence himselve, but in the meantime he has >>>ordered to make an auto232 player for fritz that doesn't allow rebel9 >>>to learn, that exchanges colors so that other learners of other chessprograms >>>get confused, and that's just the top of the iceberg. This top has >>>been confirmed by Karlsson, which i honour for being so nice to admit >>>that fritz autoplayer doesn't allow learning. >>> >> >>I disagree with the above. The 'shortcoming' is in Rebel, _not_ in Fritz. >>IE if you can't learn when alternating colors, what good is learning, since >>_most_ real tournaments do alternate? >> >>I've said it before.. what you can fix on _your_ end you should. In this >>case, it is a strange design decision indeed that says you have to play N >>games same color to learn anything... > >I think this is totally wrong. Imagine programming for an OS where the >filesystem uses 8.3 filenames, and you specify on the box that your program is >meant to be used under that OS. > >Suddenly someone uses your thing in a place they aren't supposed to use it, and >they have 256 character filenames, and you can't handle it and somehow a >filename overwrites your piece-square table and your thing plays 1. Nh3 if you >turn the book off. > HOwever the auto232 protocol says _nothing_ about alternating colors, or not alternating colors. You can play one game matches or you can play N game matches. Shouldn't you write code to fit _all_ possible cases from the protocol, or should you write code that just fits one possible scenario because you "know" everyone does it that way? >Fine. This was a short-sighted engineering decision on the part of the >programmer, but the program was bug-free until someone did something unexpected. > >The autoplayer specified by the Fritz guys was incompatible with the Donninger >autoplayer, this is not Ed's fault. I wouldn't say 'incompatible'. This reminds me of my first 'learning' efforts on ICC. People found that I learned 10 moves out of book. So at move 8-9, if they had a good position, they would disconnect and then log back in and resume, but without any 'learning' happening. I could have +noplayed everyone doing that. Or I could have 'fixed it' so that a premature termination still triggers the learning (if possible). I chose the latter. > >It doesn't say on the Rebel box, "don't use some home-brew autoplayer with this >under important circumstances that will impact my sales for years", but Ed had >no reason to expect that this would happen. It isn't something that he should >have had to foresee. Had Donninger done a new autoplayer, which flipped >black/white between games, Ed would have gotten ahold of it, found the problems, >and fixed them. auto232 has a definite 'new game' command. Why wouldn't you 'learn' after getting that no matter _which_ color you just had? Personally I always thought the N games with 1 color was 'broken', rather than alternating colors which is more normal. > >It doesn't sound like he had any opportunity to do this with the Fritz >autoplayer. > >If the autoplayer had been so incompatible with Rebel that it crashed it, the >SSDF wouldn't have scored all of those games as time forfeits for Rebel, nor >would you be arguing that they should, because Ed should have had the foresight >to predict that one of his competitors would write a incompatible version of the >software that he had come to rely upon, and mandate to the SSDF that they use >it, for reasons that are still impossible to understand. > >The only difference here is that the SSDF probably put a lot of time into >playing the games before this bug came to light, and that time would have been >wasted if the games had been replayed. > >This is probably what should have happened though, because Rebel had a bug which >probably reduced its strength, and Ed should have had a chance to fix it. > >Note that *nothing* I have said should be taken to mean that there was any >malice on the part of the Fritz guys, nor do I think that they need to have been >malicious in order to make my arguments valid. > >bruce Had I been in that position, from a software engineering point of view, I would ask the question "how can a game be terminated?" And I would make sure that for each 'answer' I have a methodology to trigger my 'learning code' if I consider learning important. I wouldn't argue that Fritz was right in doing this a new way. But there is blame on 'both ends' since alternating colors ought not be a problem, since that is 'normal'.
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.