It is less than 3200 elo but it is also clearly more than 200 elo because improvement from 3000 to 3200 is clearly harder than improvement from 1700 to 2500 and I consider Robert Houdart's achievement as clearly bigger than John boyd's achievement when he started from trace and earned 800 elo(Note that I have nothing against John boyd that was honest about what he is doing)Rebel wrote:Fabio, as a non chess programmer I realize that it must be hard to sense what it takes to write a 3000 elo chess program from the ground up. If I would have the energy to study Ippo and friends I surely can add 100 elo points to my existing one and it still would be far below 3000.Lavir wrote:LOLRebel wrote: Honesty matters because you got the first 3000 elo for free and that in a competitive field of chess programmers who write their own stuff from the ground up.
Is that meant to be comic?
Oh my god... the hypocrisy is seriously high in this place.
I would like you (or anybody else) to mention me a program that is on top 10 that has not taken everything it could from Ippo/Robbo or anywhere else when needed (it is not just because you change the implementation that the thing is suddenly different; everybody can do that).
You seem to be under the impression that taking ideas from open sources magically puts you on top, that's not so. Don recently put it very well.
http://74.220.23.57/forum/viewtopic.php?p=505505#505505
What's far more easy to do is take a 3000 elo open source and add a few of your own (better) ideas to it and gain some 25-50-75 elo with ease. Thereafter things become much more complicated. The credit of Robert is that he added much more than that.
In this competitive field taking the first 3000 elo for free is essential and by denying it an author indirectly is claiming he wrote the first 3000 elo all by himself which is a lie. Robert's contribution to computer chess is about 200 elo, not 3200. That matters.
http://www.members.optusnet.com.au/~john.boyd/
The source of the 2500 and 1700 numbers is
CCRL 40/4 that has
Trace 1.37a 2509
Tscp 1.81 1708