Search found 274 matches

by zenpawn
Sun Jun 03, 2018 2:15 pm
Forum: Computer Chess Club: General Topics
Topic: Komodo 12 and MCTS
Replies: 178
Views: 33620

Re: Exploiting versus exploring

mjlef wrote:
Sun Jun 03, 2018 2:32 am
I just checked the latest Komodo development version and it find Qd4 almost instantly. We have improved it a lot. It will be a free update for Komodo 12 owners.

Mark
Do you mean 23.Qc4 ? That's the move AlphaZero played vs Stockfish 8.
by zenpawn
Sat May 12, 2018 10:43 am
Forum: Computer Chess Club: General Topics
Topic: Proposal: name confusion
Replies: 6
Views: 802

Re: Proposal: name confusion

Sven wrote:
Sat May 12, 2018 10:28 am
zenpawn wrote:
Sat May 12, 2018 12:53 am
Good idea. Unfortunately, a new signature doesn't apply to old posts (unless I'm missing a setting).
It does. The signature is a property of the user, not of a post.
By old posts, I guess I mean the ones made before the upgrade. I've checked some of mine, and they don't have my signature.
by zenpawn
Sat May 12, 2018 12:53 am
Forum: Computer Chess Club: General Topics
Topic: Proposal: name confusion
Replies: 6
Views: 802

Re: Proposal: name confusion

Good idea. Unfortunately, a new signature doesn't apply to old posts (unless I'm missing a setting).
by zenpawn
Fri May 11, 2018 7:59 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Arasan test suite with minimal depth
Replies: 20
Views: 2492

Re: Arasan test suite with minimal depth

Dann Corbit wrote:
Thu May 10, 2018 6:53 pm
Grouped by 4, each position (of the 4) should have identical evaluation. If not, then there is an evaluation bug.
It looks like the castling rights aren't being adjusted on the color flips.
by zenpawn
Tue May 08, 2018 10:43 pm
Forum: Computer Chess Club: General Topics
Topic: Regarding Talkchess update and features
Replies: 33
Views: 4586

Re: Regarding Talkchess update and features

Seems faster, is easy to use, and the code blocks having scrollbars instead of taking up the whole page is nice. Just need to get the diagram and pgn tags working again, and we're back in business. :) Oh, and I agree the real names display in the old forum was nice, though it still showed usernames ...
by zenpawn
Tue May 08, 2018 11:56 am
Forum: Computer Chess Club: Tournaments and Matches
Topic: TCEC Division 2/Arasan
Replies: 5
Views: 625

Re: TCEC Division 2/Arasan

TCEC Division 2 is still ongoing. Arasan is participating but was disqualified after disconnecting three times (other engines have also had disconnects but so far no one else has three). I have been following the chat but I am not participating because it apparently requires login with Facebook, an...
by zenpawn
Sun Apr 29, 2018 11:37 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: About a rebel evaluation study
Replies: 16
Views: 2911

Re: About a rebel evaluation study

Uri Blass wrote: Edit 2:I can add that I believe that only piece square table with good search should have already 2600 elo
I really hope that's not true considering all the work, including eval tuning, I've put into creating an almost 2400-rated engine. :)
by zenpawn
Thu Apr 26, 2018 6:42 pm
Forum: Computer Chess Club: General Topics
Topic: You can run your engine on lichess.org now
Replies: 18
Views: 3255

Re: You can run your engine on lichess.org now

xr_a_y wrote:Nope forget what I said, I was wrong, time and otime seem to be the "new" way of manaing time (protocol version 2), Weini use level for this ...
Well, even with time, level is important, since it tells you the time control, whereas time tells you how much is left.
by zenpawn
Thu Apr 26, 2018 9:30 am
Forum: Computer Chess Club: General Topics
Topic: You can run your engine on lichess.org now
Replies: 18
Views: 3255

Re: You can run your engine on lichess.org now

I didn't have any problem with Weini using both uci (newly implemented) or xboard when I use the new bot functionality of lichess. Just that for xboard, it seems you have to implement the old "time" feature ... RookieMonster uses "time" -- didn't realize there was a newer way for the protocol to re...