New engine: Xiphos

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

pferd
Posts: 134
Joined: Thu Jul 24, 2014 2:49 pm

Re: New engine: Xiphos

Post by pferd »

AndrewGrant wrote:Not everyone uses a deterministic approach.

He seeds with system time https://github.com/milostatarevic/xipho ... ash.c#L124

So that is nothing to worry about, IMO.
You are right, using a constant seed value for srand made the program completely deterministic. It's just a bit confusing at first.
Ras
Posts: 2487
Joined: Tue Aug 30, 2016 8:19 pm
Full name: Rasmus Althoff

Re: New engine: Xiphos

Post by Ras »

In util.c, there is the function time_in_ms() which returns an int. Maximum is about 25 days in case int is 32 bits. I'd expect occasional time issues depending on the uptime of the system. Using int64_t for anything time related in milliseconds would get around that.

Another thing that might bite depending on compiler and platform is the data sharing across threads without volatile and memory barriers, e.g. the shared_search_info structure.

However, very clean coding style! :-)
ThatsIt
Posts: 991
Joined: Thu Mar 09, 2006 2:11 pm

Re: New engine: Xiphos

Post by ThatsIt »

Hi !

THX for the engine.

Any plans to implement ponder ?

Best wishes,
G.S.
(CEGT team)
Damir
Posts: 2801
Joined: Mon Feb 11, 2008 3:53 pm
Location: Denmark
Full name: Damir Desevac

Re: New engine: Xiphos

Post by Damir »

The engine is still in the early stages of development, am sure Milos will include it in his next version. In the meantime you should be happy that you got a new engine to play with, instead of making requests of what is missing already.... :o :o
ThatsIt
Posts: 991
Joined: Thu Mar 09, 2006 2:11 pm

Re: New engine: Xiphos

Post by ThatsIt »

Damir wrote:The engine is still in the early stages of development, am sure Milos will include it in his next version. In the meantime you should be happy that you got a new engine to play with, instead of making requests of what is missing already.... :o :o

That wasn't a request of mine, it was just a question !

Best wishes,
G.S.
(CEGT team)
mt
Posts: 29
Joined: Tue Feb 27, 2018 7:00 am

Re: New engine: Xiphos

Post by mt »

A minor update:
I've fixed a couple of bugs related to time control. Most significantly, in the version 0.1 the incremental time control seems to be broken (the patched version should be ~30 Elo points stronger).
This is not an official release, and for the repeated time control the difference is not significant. Also, this version is more generous with the time it leaves to the GUI, so hopefully, it avoids forfeiting on time.

Binaries: https://github.com/milostatarevic/xipho ... /tag/0.1.1
User avatar
Guenther
Posts: 4605
Joined: Wed Oct 01, 2008 6:33 am
Location: Regensburg, Germany
Full name: Guenther Simon

Re: New engine: Xiphos

Post by Guenther »

mt wrote:A minor update:
I've fixed a couple of bugs related to time control. Most significantly, in the version 0.1 the incremental time control seems to be broken (the patched version should be ~30 Elo points stronger).
This is not an official release, and for the repeated time control the difference is not significant. Also, this version is more generous with the time it leaves to the GUI, so hopefully, it avoids forfeiting on time.

Binaries: https://github.com/milostatarevic/xipho ... /tag/0.1.1
A version w/o popcount for older 64 bit hardware would be nice.

Guenther
https://rwbc-chess.de

trollwatch:
Chessqueen + chessica + AlexChess + Eduard + Sylwy
User avatar
Graham Banks
Posts: 41416
Joined: Sun Feb 26, 2006 10:52 am
Location: Auckland, NZ

Re: New engine: Xiphos

Post by Graham Banks »

Guenther wrote:A version w/o popcount for older 64 bit hardware would be nice.

Guenther
Agreed.
gbanksnz at gmail.com
mt
Posts: 29
Joined: Tue Feb 27, 2018 7:00 am

Re: New engine: Xiphos

Post by mt »

Graham Banks wrote:
Guenther wrote:A version w/o popcount for older 64 bit hardware would be nice.

Guenther
Agreed.
Will include it in the next (official) release.
Thanks everyone for all the responses and the help debugging some of the issues. :)