Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: What is good comes fast

Author: Uri Blass

Date: 21:08:18 01/21/02

Go up one level in this thread


On January 21, 2002 at 15:10:49, Bas Hamstra wrote:

>On January 21, 2002 at 04:55:02, Uri Blass wrote:
>
>>On January 21, 2002 at 04:07:11, Tony Werten wrote:
>>
>>>On January 21, 2002 at 03:51:42, Uri Blass wrote:
>>>
>>>>On January 21, 2002 at 02:32:18, Tony Werten wrote:
>>>>
>>>>>On January 20, 2002 at 14:40:08, Uri Blass wrote:
>>>>>
>>>>>>On January 20, 2002 at 12:25:29, Russell Reagan wrote:
>>>>>>
>>>>>>>On January 20, 2002 at 09:56:43, Gian-Carlo Pascutto wrote:
>>>>>>>
>>>>>>>>On January 20, 2002 at 09:28:51, Bas Hamstra wrote:
>>>>>>>>
>>>>>>>>>A dutch saying. I "converted" a colleague that has become a friend at work, I'm
>>>>>>>>>not entirely sure that this is a noble thing. He is pretty smart and has a broad
>>>>>>>>>interest. So we talked a few times about the insides of a chess program. He
>>>>>>>>>became interested and I explained alphabeta. The next day he had alphabeta
>>>>>>>>>completely figured out in a spreadsheet, and understood it. Two weeks later, he
>>>>>>>>>has a working winboard engine... Boy HE goes fast. In May (if there is a Leiden
>>>>>>>>>tournament again) he will show up with "Shark". His main target is to crush Tao
>>>>>>>>>:-) At first I smiled, and said that is was quite possible in a couple of years.
>>>>>>>>>But lately I am beginning to get the feeling that it might be sooner :-) Have
>>>>>>>>>you ever seen someone pick it up SO quickly...?
>>>>>>>>
>>>>>>>>IIRC Bruce started on Ferret in 1994 and had (from his own comments) an engine
>>>>>>>>that could have been commerical by the end of 1995. I think that's certainly
>>>>>>>>a remarkable archievement, especially considering that seven years ago there
>>>>>>>>was much less information and examples available.
>>>>>>>
>>>>>>>Bruce works for Microsoft. He doesn't count.
>>>>>>
>>>>>>
>>>>>>I do not understand what is the importance of the question where someone works
>>>>>>for the discussion.
>>>>>
>>>>>Certain workenvironments give acces to a lot of experienced programmers.
>>>>>Although it's not a sure it helps, it doesn't hurt either.
>>>>>
>>>>>Tony
>>>>
>>>>I believe that there are 2 important things to be good in chess programming.
>>>>
>>>>1)having talent for programming that means the ability to do fast a well defined
>>>>algorithm with no stupid bugs that you need to work hours to discover.
>>>>
>>>>2)having talent for math that can help you to be better in finding
>>>>a better algorithm and to be better in guessing if an idea is wrong or right
>>>>based on previous experience.
>>>>
>>>>The number of people who have special talent for both 1 and 2 is small but I
>>>>believe that someone with special talent for 1 and 2 can do a top program in
>>>>every game in a few weeks.
>>>
>>>I disagree. You forgot 3) Having new ideas. Not everything is published, so you
>>>need a feel of what is happening in your program and then find ways to do that
>>>best. 1 and 2 will only give you a decent program.
>>>
>>>Tony
>>
>>I think that having new ideas is included in 2
>>people who have a special talent for math need to have new ideas in order to
>>discover new things.
>
>I think (and those are the most seldom) people with a special talent for writing
>top programs very fast can do it. But seriously NONE can start from zero an
>write a top program in a few weeks, you must be kidding... As far as I know SMK
>has the absolute world record, doing it within a couple of months (almost a year
>or so I believe). If I am not mistaken he started from zero.


I do not think that SMK is the best(I assume that you mean to the programmer of
shredder).

I know that shredder is not good in tactics relative to Fritz and I expect a
good programmer first to do the program good in tactics(It seems to me easier
than investigating how to get better positional evaluation).

I do not say that SMK is a bad programmer and he is probably better than more
than 99% of the programmers but it is not enough.

You cannot prove that None can start from zero and write a top program in few
weeks when most of the programmers simply did not try.

I believe that the few programmers who can do it can earn more monry from other
things and not from chess programming.

Uri



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.