ShashChess
Moderator: Ras
-
- Posts: 1439
- Joined: Sat Oct 27, 2018 12:58 am
- Location: Germany
- Full name: N.N.
Re: ShashChess
I played on PlayChess 62 games with Shashchess 20.2 AVX2 (MCTS + 1 thread) without problems.
-
- Posts: 142
- Joined: Wed Feb 28, 2018 2:50 pm
-
- Posts: 352
- Joined: Tue Dec 05, 2017 2:42 pm
Re: ShashChess
I warn all users of the forum that from now on those who find problems can open an issue on the repository.
-
- Posts: 142
- Joined: Wed Feb 28, 2018 2:50 pm
-
- Posts: 73
- Joined: Wed Sep 01, 2021 2:05 am
- Location: London, England
- Full name: Vernon Crawford
Re: ShashChess
Posting as much info and detail about the error on github can be quite helpful to the developer...
refusing to do that, and repeatedly posting criticisms here is counter-productive.
-
- Posts: 142
- Joined: Wed Feb 28, 2018 2:50 pm
Re: ShashChess
and now he deleted his buggy version from github.Vernon Crawford wrote: ↑Mon Jan 17, 2022 1:35 pmPosting as much info and detail about the error on github can be quite helpful to the developer...
refusing to do that, and repeatedly posting criticisms here is counter-productive.
-
- Posts: 73
- Joined: Wed Sep 01, 2021 2:05 am
- Location: London, England
- Full name: Vernon Crawford
Re: ShashChess
Shashchess & Brainlearn are interesting derivatives which clearly involves significant work to maintain.criko wrote: ↑Mon Jan 17, 2022 1:37 pmand now he deleted his buggy version from github.Vernon Crawford wrote: ↑Mon Jan 17, 2022 1:35 pmPosting as much info and detail about the error on github can be quite helpful to the developer...
refusing to do that, and repeatedly posting criticisms here is counter-productive.
You could have stopped after losing 1 game and contacted him concerning the error.
Instead you continued to lose several more games, and now it seems you're angry and lashing out.
I reported your other topic which was intended to humiliate...thankfully the Mods removed it.
Shame on you!
-
- Posts: 142
- Joined: Wed Feb 28, 2018 2:50 pm
Re: ShashChess
Now he changed the sources and executables.Vernon Crawford wrote: ↑Mon Jan 17, 2022 1:53 pmShashchess & Brainlearn are interesting derivatives which clearly involves significant work to maintain.criko wrote: ↑Mon Jan 17, 2022 1:37 pmand now he deleted his buggy version from github.Vernon Crawford wrote: ↑Mon Jan 17, 2022 1:35 pmPosting as much info and detail about the error on github can be quite helpful to the developer...
refusing to do that, and repeatedly posting criticisms here is counter-productive.
You could have stopped after losing 1 game and contacted him concerning the error.
Instead you continued to lose several more games, and now it seems you're angry and lashing out.
I reported your other topic which was intended to humiliate...thankfully the Mods removed it.
Shame on you!
Still buggy.
BTW: I don't care for elo.

-
- Posts: 3311
- Joined: Sat Feb 16, 2008 7:38 am
- Full name: Peter Martan
Re: ShashChess
You're fully right.
Just one question to the new MCTS- params, please.
Choosing MCTS checked with big number of MTCS- threads means same as "Multi" MCTS in version before, doesn't it? Thus only main line isn't MCTS-searched, right?
Quoted from Readme
I just ask, because Shredder- GUI, giving much information with start- stop of analysis, lists all possible moves after stop with wdl- chances in percentage, even if number of MCTS- threads is set to 1 only. Does that mean all threads (possible first moves) are MCTS- checked yet too?MCTSThreads
Integer, Default: 0, Min: 0, Max: 512 The number of settled threads to use for MCTS search except the first (main) one always for alpha-beta search.
Choosing 2 as number of MCTS- threads doesn't change output in Shredder- GUI, shouldn't there then only 2 threads get evaluated by MCTS?
Thanks for answering anymore here anyhow regards
Peter.
-
- Posts: 511
- Joined: Sun Apr 26, 2020 11:40 pm
- Full name: Brian D. Smith
Re: ShashChess
Thanks....yes, addressing to anyone really.peter wrote: ↑Mon Jan 17, 2022 8:18 amI don't mind your questions, but I'm probably the wrong guy, you asked, maybe a mistake, yet to answer again from my side, cause it was me saying I liked the easier handling:
Who says, that with one eligible UCI- option there can't be given enough change of pararmeter- setting as for a user- tool to adapt Shashin- theory- search and it (the search) yet does adapt for the rest of the positions on its own as well as before or even better?
Just my two cents, now waiting patiently for Andrea's answer, if he want's to give one anyhow, considering everybody (who is able to understand) can read the code and it's changes to get really exact answers on one's own regards
So, basically nothing has changed internally with the 'personalities' in that internally the engine still adapts to the old personalities (for lack of a better word) but one can only specify 'Tal'. I wonder why they kept that one as an option you can set and did away with the other options.