Author: Dann Corbit
Date: 11:59:21 08/20/02
Go up one level in this thread
On August 20, 2002 at 14:07:27, Dann Corbit wrote: >On August 20, 2002 at 10:55:33, Gian-Carlo Pascutto wrote: > >>On August 20, 2002 at 09:43:17, Sune Fischer wrote: >> >>>I had the same thought, copy-paste from net and you have a thesis, amazing... >>> >>>Looks more like an article for the sunday paper than a thesis. >> >>I would have to disagree 200%. >> >>It's by far one of the most complete accounts of whats needed in the >>actual implementation of a program, it contains several new ideas, >>describes some known ones that weren't formally described before and >>it's written in a very understandable way. >> >>Also note that the main parts of it date back to before 1998. It >>was only published now because there were a few chapters >>that never quite got finished in the years before. This is why some >>parts may be dated now. >> >>But even discounting the latter, your comments are downright >>insulting and injustified, IMHO. >> >>Maybe I should throw a stack of ICCA journals at you, to learn >>to relativate. > >I agree with GCP, and I also think that we should not be expecting radical new >algorithms or some kind of magic from a Master's thesis. Sure, it's a rehash of >known techniques. But the author is (I think) a *very good teacher* as far as >explaining things. > >Let me go further... > >I really, really, wish I had this document at my disposal when I *first* started >learning about computer chess. It would have saved me about TWO YEARS of >effort. > >His writing is clear and understandable. The document has excellent >organization. > >For someone who wants to learn how to write a chess program: >THIS DOCUMENT is better than ANYTHING ELSE for a new beginner. It won't teach >you some kind of magical excellence where suddenly you can pop off an engine >that is clobbering Chess Tiger, Fritz and Junior. But it will show you all the >necessary techniques and ideas to write a good, solid chess engine. What is >more, it explains how they work and why they work in such an easy to understand >manner that *anyone* can grasp it. Personally, I think that is a form of >genius. For example, read Albert Einstein's notebooks. He talks about men in >elevators and guys walking on trains. Simple, clear illustrations that anyone >can understand. If you read Einstein's explanation of relativity, a 12 year old >can understand it. This ability to communicate a difficult subject in a clear >and comprehensible manner is a rare talent. The very fact that reading his >thesis is like reading "Dick and Jane" is actually demonstrating its excellence >rather than showing some sort of lack. For those who have never heard of "Dick and Jane" -- these are the books that used to be used to teach first grade students to read in the US: http://www.valpo.edu/home/faculty/bflak/dickjane/ If you are already an advanced chess programmer, perhaps you won't learn a lot from the Master's thesis. But I suggest that *everyone* read it, [if for nothing else] to see a good example of a clear teaching style. Even if you don't learn any new programming ideas, you will see a very good example of how to communicate technical ideas in a way that absolutely anyone can understand. I think that a chess GM could pick up this paper and immediately understand it without ever having learned anything about programming. In fact, I think it would be an excellent article for the JICCA (or whatever it is called now). >Once again, his thesis is an excellent piece of work and something to be proud >of. I expect that many new chess engines will get a very good start because of >what he has done.
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.