In your original post you said:
right now there is no hashing, no ai, no transposition table no nothing, it's just a simple move generator
"no nothing !!!!!!!"
that's a double negative !!!
How can you program boolean logic when you talk like that ?????
reviving old c# codebase engine, 0x88
Moderators: hgm, Rebel, chrisw
-
- Posts: 27811
- Joined: Fri Mar 10, 2006 10:06 am
- Location: Amsterdam
- Full name: H G Muller
Re: reviving old c# codebase engine, 0x88
I don't even have an engine at the moment where I could test that. I suppose I could test in on Qperft.spirch wrote: ↑Wed Mar 24, 2021 1:33 amcan you tell what kind of % speed increase you got with your engine? if anyhgm wrote: ↑Tue Mar 23, 2021 2:14 pm [Edit] The aligned[] table elements could have the same format as the set of aligned piece-direction combinations. Each piece could then have a mask associated with it (in the piece list) to indicate the piece-direction bits that belong to it. The incremental update would then be
alignedSet &= ~sliderMask[piece];
alignedSet |= sliderMask[piece] & aligned[toSqr - location[opponent+KING]];
But I will surely investigate it in the "mailbox trials" project, when I have time again to work on that.
-
- Posts: 881
- Joined: Sun Dec 27, 2020 2:40 am
- Location: Bremen, Germany
- Full name: Thomas Jahn
Re: reviving old c# codebase engine, 0x88
The copious amount of times you used the NOT operator (!) in your post makes your central point of critizism quite ironic!
-
- Posts: 95
- Joined: Fri Nov 09, 2012 12:36 am
-
- Posts: 27811
- Joined: Fri Mar 10, 2006 10:06 am
- Location: Amsterdam
- Full name: H G Muller
Re: reviving old c# codebase engine, 0x88
No, not yet. At the moment I don't have any pin detection at all. Nor check detection, for that matter; my test search still relies on detecting (and scoring) the King capture in the child node. I am still working on the basic method of move generation.