Going SMP is a card you can play anytime. It is not really chess programming, it is just solving a technical problem that gives you about 70 ELO. In my opinion it is a good idea to delay the SMP stuff as long as possible (if you run out of any other idea then maybe it's time for it)
Going SMP is a card you can play anytime. It is not really chess programming, it is just solving a technical problem that gives you about 70 ELO. In my opinion it is a good idea to delay the SMP stuff as long as possible (if you run out of any other idea then maybe it's time for it)
Thomas...
I completely agree. Also, as a single threaded program, I have a lot of global variables and functions that must be relocated. An example is the move_do function. Is global and makes de move in the global board... I have to do a big effort for putting the board methods in a board class... In these days, for me is a lot of work to be done.
At this point, endgame knowledge is more important. Another improvement that can be done is to change the pawn structure evaluation ( for example, rhetoric doesn't know anything about candidate passers), add a pawn hash...
Maybe in a future 2.0 version I will add SMP, but for the moment I will be (very) happy if Rhetoric can reach iCE level in a single core.
We don't need more power to be crushed in every game we play. What we need is chess knowledge and rich positions to enjoy the game!
Still learning how to play chess...
knigths move in "L" shape ¿right?
Setting the material at 66% produces amazing chess!
The standard Rhetoric sets the pawn value to 71 for the opening and 153 for the endgame.
With 66%, the value of a pawn in the opening is 46 cp!! it is a value easily surpased with the evaluation of mobility and king safety. Here Rhetoric gave 2 pawns!!
At endgames, 156*0.66 = 102 cp. That parameter makes Rhetoric to consider the endgames using the average "standard" values for the pieces. So it does not make big errors when is transforming an advantage into a superior endgame.
Joe, you found the magic number!! material = 66!!
This is real chess!
Still learning how to play chess...
knigths move in "L" shape ¿right?
I have no objection, but can you put the link to my web page in the download description? I have a donation button, and would be nice if one of your visitors is so kind to give a present by the way.
PS: You have a nice web site. I like it.
Still learning how to play chess...
knigths move in "L" shape ¿right?
Going SMP is a card you can play anytime. It is not really chess programming, it is just solving a technical problem that gives you about 70 ELO. In my opinion it is a good idea to delay the SMP stuff as long as possible (if you run out of any other idea then maybe it's time for it)
Of course, Alberto. I will do so! Your work ist absolutly worth to donate it.
Can I ask You for some features that I would like to have in Rhetoric:
1) make Rhetoric look a few plys deeper into the game when possible
2) add a personality-option wiht adjustable parameters
3 spend an limith-strenght option to this engine
i will make a report about Rhetoric on my website in the next days to make Rhetoric more public visible
please can You send me Your mailadress for direct contacting when needed, i don't like to use webforms
Don't let Your tricky mind kick You out of paradise!
Henryval wrote:Of course, Alberto. I will do so! Your work ist absolutly worth to donate it.
Can I ask You for some features that I would like to have in Rhetoric:
1) make Rhetoric look a few plys deeper into the game when possible
2) add a personality-option wiht adjustable parameters
3 spend an limith-strenght option to this engine
i will make a report about Rhetoric on my website in the next days to make Rhetoric more public visible
please can You send me Your mailadress for direct contacting when needed, i don't like to use webforms
1) I ask for this feature each chrismas
2) This is what I intended with the material option. Material is stored in a precalculated table, so it was easy to change the piece values. The rest of the evaluation parameters are another story. Not so easy as it must be tested very well, as the evaluation is the monster's heart.
3) The limit-strength option is implemented via the standard UCI parameters UCI_LimitStrength and UCI_Elo. It must work reasonably well running on a standard PC. It is implemented using a delay in eval.
Playing at 1600 elo can be tricky but accesible for an average player.
I sent you a PM with my mail address.
Still learning how to play chess...
knigths move in "L" shape ¿right?