New UCI 2 version of Monarch

[Steve Maughan]

Geschrieben von:/Posted by: Steve Maughan at 13 April 2004 03:02:07:

After a two year absence there is a new version of Monarch. The new version has been completely re-written in C (was Delphi). The new version is UCI compatible (versions 1 & 2!) but not currently Winboard compatible. For those interested you can download it at:
www.stevemaughan.com/monarch
I'd estimate that the current strength is ~1950 ELO on decent hardware. Hopefully this will increase over the coming months!
Have fun,
Steve Maughan


http://www.stevemaughan.com/monarch

[Olivier Deville]

Geschrieben von:/Posted by: Olivier Deville at 13 April 2004 08:09:26:
Als Antwort auf:/In reply to: New UCI 2 version of Monarch geschrieben von:/posted by: Steve Maughan at 13 April 2004 03:02:07:

After a two year absence there is a new version of Monarch. The new version has been completely re-written in C (was Delphi). The new version is UCI compatible (versions 1 & 2!) but not currently Winboard compatible. For those interested you can download it at:
www.stevemaughan.com/monarch
I'd estimate that the current strength is ~1950 ELO on decent hardware. Hopefully this will increase over the coming months!
Have fun,
Steve Maughan

Very good news !! Thanks Steve for updating this very interesting engine
Olivier



ChessWar

[Igor Gorelikov]

Geschrieben von:/Posted by: Igor Gorelikov at 13 April 2004 11:50:52:
Als Antwort auf:/In reply to: New UCI 2 version of Monarch geschrieben von:/posted by: Steve Maughan at 13 April 2004 03:02:07:

After a two year absence there is a new version of Monarch. The new version has been completely re-written in C (was Delphi). The new version is UCI compatible (versions 1 & 2!) but not currently Winboard compatible. For those interested you can download it at:
www.stevemaughan.com/monarch
I'd estimate that the current strength is ~1950 ELO on decent hardware. Hopefully this will increase over the coming months!
Have fun,
Steve Maughan

TaskInfo shows that Monarch ALWAYS uses 8-9 Mb of RAM instead of the hash value I set.
Conditions: Windows 98 SE, under Polyglot, under Arena.
Igor

[Steve Maughan]

Geschrieben von:/Posted by: Steve Maughan at 13 April 2004 12:15:31:
Als Antwort auf:/In reply to: Hash allocation geschrieben von:/posted by: Igor Gorelikov at 13 April 2004 11:50:52:

TaskInfo shows that Monarch ALWAYS uses 8-9 Mb of RAM instead of the hash
value I set. Conditions: Windows 98 SE, under Polyglot, under Arena.

Igor,
Thanks - I'll take a look.
Steve

[Fabien Letouzey]

Geschrieben von:/Posted by: Fabien Letouzey at 13 April 2004 12:20:27:
Als Antwort auf:/In reply to: Re: Hash allocation geschrieben von:/posted by: Steve Maughan at 13 April 2004 12:15:31:

TaskInfo shows that Monarch ALWAYS uses 8-9 Mb of RAM instead of the hash
value I set. Conditions: Windows 98 SE, under Polyglot, under Arena.

Igor,
Thanks - I'll take a look.
Steve

Note that it could be my fault as well.
PolyGlot sends UCI options before receiving "uciok".
It does send "uci" first though.
Fabien.

[David Dahlem]

Geschrieben von:/Posted by: David Dahlem at 13 April 2004 22:39:56:
Als Antwort auf:/In reply to: Hash allocation geschrieben von:/posted by: Igor Gorelikov at 13 April 2004 11:50:52:

After a two year absence there is a new version of Monarch. The new version has been completely re-written in C (was Delphi). The new version is UCI compatible (versions 1 & 2!) but not currently Winboard compatible. For those interested you can download it at:
www.stevemaughan.com/monarch
I'd estimate that the current strength is ~1950 ELO on decent hardware. Hopefully this will increase over the coming months!
Have fun,
Steve Maughan

TaskInfo shows that Monarch ALWAYS uses 8-9 Mb of RAM instead of the hash value I set.
Conditions: Windows 98 SE, under Polyglot, under Arena.
Igor

Monarch is uci, and Arena supports uci. Perhaps the problem is caused by using an adapter when an adapter is not necessary.
Regards
Dave

[Steve Maughan]

Geschrieben von:/Posted by: Steve Maughan at 14 April 2004 01:50:39:
Als Antwort auf:/In reply to: Re: Hash allocation geschrieben von:/posted by: Fabien Letouzey at 13 April 2004 12:20:27:

Fabien,

Note that it could be my fault as well.
PolyGlot sends UCI options before receiving "uciok".

This could be the problem. Monarch waits for "isready" before initializing.
Regards,
Steve

[Fabien Letouzey]

Geschrieben von:/Posted by: Fabien Letouzey at 14 April 2004 10:22:24:
Als Antwort auf:/In reply to: Re: Hash allocation geschrieben von:/posted by: Steve Maughan at 14 April 2004 01:50:39:

Note that it could be my fault as well.
PolyGlot sends UCI options before receiving "uciok".

This could be the problem. Monarch waits for "isready" before initializing.

That's unrelated IMO.
PolyGlot sends "uci", then the options, then "isready" (of course!).
The (unrelated) trick is that it does not wait for "uciok" before sending the options. Does this cause a problem?
Fabien.

[Igor Gorelikov]

Geschrieben von:/Posted by: Igor Gorelikov at 14 April 2004 10:33:01:
Als Antwort auf:/In reply to: Re: Hash allocation geschrieben von:/posted by: David Dahlem at 13 April 2004 22:39:56:

Conditions: Windows 98 SE, under Polyglot, under Arena.
Igor

Monarch is uci, and Arena supports uci. Perhaps the problem is caused by using an adapter when an adapter is not necessary.
Regards
Dave

To make it clear: Monarch first is running under Polyglot plus Winboard, then under Arena (without Polyglot).
Sorry, if my text was dubious.
Igor

[Steve Maughan]

Geschrieben von:/Posted by: Steve Maughan at 14 April 2004 14:46:10:
Als Antwort auf:/In reply to: Re: Hash allocation geschrieben von:/posted by: Fabien Letouzey at 14 April 2004 10:22:24:

Fabien,

Note that it could be my fault as well.
PolyGlot sends UCI options before receiving "uciok".

This could be the problem. Monarch waits for "isready" before initializing.

That's unrelated IMO.
PolyGlot sends "uci", then the options, then "isready" (of course!).
The (unrelated) trick is that it does not wait for "uciok" before sending the >options. Does this cause a problem?

Probably not but from readng the UCI protocol I had the impression that the engine should be initialized after the isready command. Is this correct?
Regards,
Steve

[Fabien Letouzey]

Geschrieben von:/Posted by: Fabien Letouzey at 14 April 2004 15:01:34:
Als Antwort auf:/In reply to: Re: Hash allocation geschrieben von:/posted by: Steve Maughan at 14 April 2004 14:46:10:

Fabien,

Note that it could be my fault as well.
PolyGlot sends UCI options before receiving "uciok".

This could be the problem. Monarch waits for "isready" before initializing.

That's unrelated IMO.
PolyGlot sends "uci", then the options, then "isready" (of course!).
The (unrelated) trick is that it does not wait for "uciok" before sending the >options. Does this cause a problem?

Probably not but from readng the UCI protocol I had the impression that the engine should be initialized after the isready command. Is this correct?
Regards,
Steve

probably not?
I was asking whether this causes a problem with your engine or not.
Yes of course what I do is "incorrect" (reread my previous posts in this thread).
But I don't think any engine minds for the moment.
Fabien.