Author: Uri Blass
Date: 16:50:24 08/25/05
Go up one level in this thread
On August 25, 2005 at 18:34:21, Dan Honeycutt wrote:
>On August 25, 2005 at 16:23:06, José Carlos wrote:
>
>...
>>
>> I understood the idea immediately and I thought it could be a good idea to
>>produce more cutoffs with losing too much information. I decided to try the
>>idea. How could I implement it? Not many ways to choose. I just had to use the
>>same exact line or just discard it.
>> I tried, and it seemed to be a tiny improvement.
>> Months later (note that I programmed Averno about one hour per week, and it's
>>been a year or so since the last time I did it), I figured out possible
>>improvement. I tried (from memory):
>>
>> if (abs(eval) >= 30)
>> {
>> if (abs(eval) >= 150)
>> eval -= (eval % 5);
>> else
>> eval -= (eval %3);
>> }
>>
>> The idea was that when the advantage is bigger, less precision is needed.
>> It seemed to work.
>> Now I ask myself: is this an unfair use of open source? I don't think so, but
>>I'm not sure anymore...
>>
>> José C.
>
>I think you are innocent, José. I also think Bo makes a good point.
>
>I would know if I were taking someone else's idea and implementing it myself,
>even if my code looked exactly like theirs such as in your example.
I agree about it.
>
>I would also know if I were stealing someone else's work even if I rearranged,
>renamed and converted to other structures to the point where my code looked
>nothing like theirs.
>
>Best
>Dan H.
I do not think that the last case is stealing.
If you translate fruit's code to human language and use different structure to
implement the same ideas you steal nothing and I think fabien will agree about
it.
I reject the idea that using good ideas that you understand is stealing
otherwise it is not clear how many ideas you can copy because you think that
they are better and testing showed that they are probably better than what you
have.
Free source programs were generated to help programmers to learn from them.
If programmers are not allowed to use what they learn they worth nothing.
If it is not clear how many ideas they can use then it is also not a good
situation.
Note that in the case of movei I decided to undo most of the ideas that I took
from fruit for the simple fact that testing suggest that they probably make
movei weaker because they do not go together with other evaluation terms that I
already have.
I am even not sure if the part that I left makes it stronger but I think that
the idea that heavy pieces that attack square near the king worth more then
small pieces is a good idea that I plan to use (the code that I have in that
case is also not the same as fruit and I am also not sure if it makes movei
stronger).
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.