Search found 1898 matches

by mar
Wed May 22, 2019 3:17 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Possible "real" hash collision
Replies: 27
Views: 569

Re: Possible "real" hash collision

Is everybody else engine seeing the only good move for white (Qf5) ? or after Kh2 the good move for black Rf7? My engine needs 3 seconds and depth 18 to see Rf7 after Kh2, single threaded. It wants to play Qf5 before Kh2 rather quickly, but it's probably simply because it prefers Qf5, not because i...
by mar
Wed May 22, 2019 10:31 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Possible "real" hash collision
Replies: 27
Views: 569

Re: Possible "real" hash collision

If they don't exist, why then verify the tt-move? maybe because of race conditions when multiple threads work on shared TT? (I'm paranoid enough to do it in single-threaded search as well because my makemove would crash on invalid move) the effect of hash collisions along with the impact on root mo...
by mar
Wed May 22, 2019 7:33 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Possible "real" hash collision
Replies: 27
Views: 569

Re: Possible "real" hash collision

xr_a_y wrote:
Wed May 22, 2019 5:32 am
but the position is not lost, it is drawish according to stockfish.
After Kh2, is it immediately a lose.
You're right, I was looking at the second Kh2, sorry.
Could it be that minic simply played too fast? How much time was spent on the move?
by mar
Wed May 22, 2019 5:28 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Possible "real" hash collision
Replies: 27
Views: 569

Re: Possible "real" hash collision

9 of 9 "real" hash collisions are either a fantasies or bugs
why do you think kh2 a blunder? the position is lost anyway
by mar
Sat May 18, 2019 8:01 pm
Forum: Computer Chess Club: General Topics
Topic: Luciole (UCI engine written in Lua)
Replies: 3
Views: 340

Re: Luciole (UCI engine written in Lua)

Very nice. Luciole seems to play fixed depth + moves seem a bit random to me though (maybe I'm doing something wrong) I've noticed you use math.mininteger/math.maxinteger from Lua 5.3, if you replace this with constants, Luciole can be run in LuaJIT http://luajit.org/luajit.html which is an awesome ...
by mar
Fri May 17, 2019 12:24 pm
Forum: Computer Chess Club: General Topics
Topic: Sideloading on iOS now Possible
Replies: 12
Views: 413

Re: Sideloading on iOS now Possible

I dont know if you're a millennial or gen-z but i feel i get static with the younger generation on talkchess. Yea it's possible to develop for ios without great difficulty beyond 99 a year. Though you can side load on your own device for free. Android is no walk in the park. Any development is a so...
by mar
Fri May 17, 2019 11:37 am
Forum: Computer Chess Club: General Topics
Topic: Sideloading on iOS now Possible
Replies: 12
Views: 413

Re: Sideloading on iOS now Possible

Simply sharing some frustrations I had when developing for iOS.
I also forgot the pain to generate zillions of icon sizes for various devices, I guess automatic image resampling would be too much to ask.
(my points are perfectly valid though - even if you bury your fanboy head deep in the sand :)
by mar
Fri May 17, 2019 10:38 am
Forum: Computer Chess Club: General Topics
Topic: Sideloading on iOS now Possible
Replies: 12
Views: 413

Re: Sideloading on iOS now Possible

I have a few complaints: iOS: - you can't spawn processes on iOS => loading external engines is a no-go (this is not about safety because the app already runs sandboxed) - you can't allocate executable pages so no JITting for you (again this is not about safety because the app already runs sandboxed...
by mar
Thu May 16, 2019 9:01 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Method to interpret the third field of a FEN string in a program playing FRC
Replies: 2
Views: 190

Re: Method to interpret the third field of a FEN string in a program playing FRC

Well, in X-FEN you don't specify the rook file (AHah - that's Shredder notation if I'm not mistaken), so you simply have to scan for the rooks to determine rook file. I think this is probably what you do. It's a bit trickier if you have both rooks on either side, for more information: https://en.wik...
by mar
Wed May 15, 2019 10:16 pm
Forum: Computer Chess Club: General Topics
Topic: Avalanche of Elo
Replies: 2
Views: 315

Re: Avalanche of Elo

...I bet new Houdini will require a strong GPU :)