TalkChess.com

xr_a_y wrote: ↑

Wed Jan 13, 2021 6:22 pm

No difference in terms of nodes count (or any other statistic Minic is able to print), so no obvious bug.
But 10% lower nps.

10% lower nps - that's at most 10 elo at hyperbullet, realistically about 5.
so I wonder where you get -45 I suspect an issue with testing.

P.S. Martin, comewhere in the Cheng thread you've mentioned something like: "because I've made no progress since 4.40dev and because CEGT is already testing Guenther's compiles" - I just wanted to ask what you do when feel that kind of reaching the limits - when you feel like you can't make an engi...

I see. I would be more than happy with 30+

I see. have you tried to mask it to 32 bits (or 31) first, like hashKey &= 0xffffffff? then use %?
still - if you round number of entries to nearest power of two and then mask with num_entries-1, it should work as is

I don't understand your masking around hashEntries: var hashEntry = hashTable[hashKey & (hashEntries - 4)]; why -4, plus hashEntries seems nowehere close to a power of two (default is some weird value like 16M/80, so I'm puzzled if hashEntries was 2**n, then masking with (hashEntries-1) I would unde...

thanks for the tournament, it was fun to follow except for some hiccups (like streaming stopped after n rounds for some reason etc.)

thanks for testing Cheng (Gabor I guess?)
if this result holds, that would mean that I broke 3k CCRL blitz - a very important milestone to me (even if elo is relative)

Angle wrote: ↑

Fri Jan 08, 2021 3:21 am

Cheng 4.40 is playng very well in my current tourney:
http://talkchess.com/forum3/viewtopic.php?f=6&t=76231
https://www.youtube.com/watch?v=aN9hAzWRwPI

I'm following the tourney. luck is fun, but we'll see how it ends. beating 4.39 is my only concern for the moment, even if it's only 50 games.

Dann, we both know that's not quite true :) The code is unnecessarily verbose and certainly not "beautiful", but at least it doesn't crash when ucinewgame is missing or when illegal fen is passed to the engine, because there's simply no reason to :wink: Wolfgang: that's probably too good to be true,...

Android NDK uses the clang compiler. And with it, you have to change very often parts the source code for proper compiling (but this can be done very often in a minimal invasive way). For Android compiles, i simply use ndk-build from Android NDK. Cheng 4.40: https://app.box.com/s/235inq8ab3rlu0objn...