bug-gnubg
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Bug-gnubg] Windows development/test build released


From: Michael Petch
Subject: Re: [Bug-gnubg] Windows development/test build released
Date: Sat, 16 Mar 2013 21:34:33 -0600
User-agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:17.0) Gecko/17.0 Thunderbird/17.0

On 2013-03-16 14:00, Neural Gnat wrote:
> I've just re-analysed a 1000-game money session that I did about a week 
> ago with 2012's World Class versus Casual. This new test version has 
> found 1839 doubtful moves, 304 bad moves and 247 very bad moves, 
> knocking the mainstream version down from Supernatural to World Class 
> (-4.0).
> 
> The question is, how do you determine which of those opinions are 
> correct? Dare I mention XG? ;o)

I think if you have one version play the other in a large number of
money sessions, and can show a noticeable improvement in matches won for
one version over the other, then one can probably make the determination
that one is better than the other.

Philippe Michel did make some posts to this mailing list about it.

http://lists.gnu.org/archive/html/bug-gnubg/2012-09/msg00008.html
http://lists.gnu.org/archive/html/bug-gnubg/2013-01/msg00013.html
http://lists.gnu.org/archive/html/bug-gnubg/2013-01/msg00017.html

It might be an interesting experiment to run the Depreli positions
through the new version to see how they compare with XG2 and the
previous version of GNUBG.

> 
> Another question is, how do I get these two versions playing each other? 
> I tried the "socket" players a few years ago but, with no instructions, 
> no result and no feedback from GnuBg, I soon gave up.
> 

You can use the GUI for part of this but if you are automating you will
probably want to use the command line interface (CLI). The CLI is far
more convenient for automation. Assuming Windows platform the command
line interface is called gnubg-cli and can be found in the same
directory as the main GNUBG executable.

Launch the GUI for one version of the bot. Set up the parameters for
player 0 via settings menu/players. Set player 0 to GNUBG and set the
playing strength you want. SPECIAL NOTE: Make sure that no player name
has SPACES in it! (Trust me this caught me of guard in testing). Close
the GUI and launch the command line version of that bot. Enter this on
the command line:

external 127.0.0.1:9999

The number after the colon can be any free port number. If you are
running both versions bots on the same computer you can use the loopback
ip address of 127.0.0.1. If you want to listen on your external IP for
that computer then replace 127.0.0.1 with your computer's IP.

The external command will put that copy of GNUBG into listen mode and
will look for connection requests from another bot.

Now open up the GUI for another version of GNUBG (ie the test version).
Start a new match. Select "Modify Player Settings...". Select Player 1
as being GNUBG. Click the "Socket" radio button. In the socket box enter
the IP address and port number just as you entered them on the external
command earlier (ie: 127.0.0.1:9999). For player 0, set up GNUGB but set
the skill level you want this copy of the bot to play at. The current
bot should start playing against the remote bot that was listening.

If you want the second copy of bot to use the CLI, use the GUI to set up
player 0 (the strength you wish this bot to play at) and then exit the
GUI version. Launch the CLI of the second bot (newer version) and issue
this command on the command line:

set player 1 external 127.0.0.1:9999
new match <matchlength>

Where match length is 0 for money session, and any other number is a
match to specified length.

The bot should start playing automatically.

Using the GUI is convenient for setting up playing strength. This can be
done from the command line but isn't at straight forward. If you are
serious about automating bot vs bot play, it is much easier to write
scripts under Linux IMHO.

-- 
Michael Petch
CApp::Sysware Consulting Ltd.
OpenPGP FingerPrint=D81C 6A0D 987E 7DA5 3219 6715 466A 2ACE 5CAE 3304



reply via email to

[Prev in Thread] Current Thread [Next in Thread]