Page 1 of 1

Marginal on strike

PostPosted: 14 Jan 2009, 07:55
by Olivier Deville
Marginal on strike

Marginal, UCI engine by Alexander Turikov, is playing in OpenWar 5th Edition. The GUI used is Winboard_F, + Polyglot for UCI engines.

Marginal played without any problems in the test tournaments, and in the real tourney as well until round 19.

And now, it just thinks forever after the opening moves are made. I have not updated a single thing in my comp, so this is really a mystery for me why it was workking before, and it does not work anymore now.

Marginal still works under Arena in UCI mode. Please have a look, if you have some free time. Any workaround is welcome...

Since ChessAlex/Marginal website seems to be gone, I have put Marginal here :

http://www.open-aurec.com/Marginal.rar

The author seems to have vanished, no answer so far from him.

Olivier

Re: Marginal on strike

PostPosted: 14 Jan 2009, 08:26
by Guenther Simon
Olivier Deville wrote:Marginal on strike

Marginal, UCI engine by Alexander Turikov, is playing in OpenWar 5th Edition. The GUI used is Winboard_F, + Polyglot for UCI engines.

Marginal played without any problems in the test tournaments, and in the real tourney as well until round 19.

And now, it just thinks forever after the opening moves are made. I have not updated a single thing in my comp, so this is really a mystery for me why it was workking before, and it does not work anymore now.

Marginal still works under Arena in UCI mode. Please have a look, if you have some free time. Any workaround is welcome...

Since ChessAlex/Marginal website seems to be gone, I have put Marginal here :

http://www.open-aurec.com/Marginal.rar

The author seems to have vanished, no answer so far from him.

Olivier


Hi Olivier,

I did test Marginal the very first time a few days ago, that was when I still
had Vista Home Premium SP1 installed on my new Quad. Here I also
noticed that Marginal always used 99.x% of all of its time for the first
move in a tc moves/time. It worked correctly in increment mode though!
Did you perhaps change the WB_F version lately? It is possible that the
time control commands where somehow changed, but I have no time to check
it until tonight. You could compare an older WB version for the
behaviour, of course it could still be a tc bug in Marginal, but it would be
strange, as it worked for you before.

Guenther

Re: Marginal on strike

PostPosted: 14 Jan 2009, 08:38
by Graham Banks
Guenther Simon wrote:
Olivier Deville wrote:Marginal on strike

Marginal, UCI engine by Alexander Turikov, is playing in OpenWar 5th Edition. The GUI used is Winboard_F, + Polyglot for UCI engines.

Marginal played without any problems in the test tournaments, and in the real tourney as well until round 19.

And now, it just thinks forever after the opening moves are made. I have not updated a single thing in my comp, so this is really a mystery for me why it was workking before, and it does not work anymore now.

Marginal still works under Arena in UCI mode. Please have a look, if you have some free time. Any workaround is welcome...

Since ChessAlex/Marginal website seems to be gone, I have put Marginal here :

http://www.open-aurec.com/Marginal.rar

The author seems to have vanished, no answer so far from him.

Olivier


Hi Olivier,

I did test Marginal the very first time a few days ago, that was when I still
had Vista Home Premium SP1 installed on my new Quad. Here I also
noticed that Marginal always used 99.x% of all of its time for the first
move in a tc moves/time. It worked correctly in increment mode though!
Did you perhaps change the WB_F version lately? It is possible that the
time control commands where somehow changed, but I have no time to check
it until tonight. You could compare an older WB version for the
behaviour, of course it could still be a tc bug in Marginal, but it would be
strange, as it worked for you before.

Guenther


Same problem with Marginal here. I gave up on it.

Re: Marginal on strike

PostPosted: 14 Jan 2009, 10:50
by H.G.Muller
Probbly it interprets the Windows tick count as a signed number, confusing it when it is negative. This wraps around in 2^32 = 4G msec = 4M sec = 50 days. So it would work correctly for 25 days, and then be broken for the next 25 days, etc...

Re: Marginal on strike

PostPosted: 14 Jan 2009, 14:16
by Olivier Deville
Thanks all for the replies.

I made a fresh installation of Marginal from the copy I got by e-mail, and now it works again.

And yes, it supports only incremental time control.

Olivier

Re: Marginal on strike

PostPosted: 12 Jun 2009, 19:06
by pimidrez
Olivier Deville wrote:Thanks all for the replies.

I made a fresh installation of Marginal from the copy I got by e-mail, and now it works again.

And yes, it supports only incremental time control.

Olivier


HI OLIVER!!

i will like to know if anyone have the source code of marginal, cause I work in VB to and I will like to take a look at the source..

thanks!!

Re: Marginal on strike

PostPosted: 12 Jun 2009, 19:12
by Olivier Deville
pimidrez wrote:
Olivier Deville wrote:Thanks all for the replies.

I made a fresh installation of Marginal from the copy I got by e-mail, and now it works again.

And yes, it supports only incremental time control.

Olivier


HI OLIVER!!

i will like to know if anyone have the source code of marginal, cause I work in VB to and I will like to take a look at the source..

thanks!!


Unfortunately not.

Alexander Turikov deleted his website and even his e-mail address. He lost all interest for computer chess :(

Olivier

Re: Marginal on strike

PostPosted: 12 Jun 2009, 20:32
by Ron Murawski
Olivier Deville wrote:Marginal on strike

...
Since ChessAlex/Marginal website seems to be gone, I have put Marginal here :

http://www.open-aurec.com/Marginal.rar

The author seems to have vanished, no answer so far from him.

Olivier


Hi Olivier,

Marginal is also available from SDChess:
http://www.sdchess.ru/download_engines.htm

I wish old ChessAlex engine by the same author was available!

Ron