Page 1 of 1

CCRL Testing IvanHoe

Posted: Fri Mar 09, 2012 3:53 pm
by kingliveson
The world must truly be coming to an end. :)

http://www.computerchess.org.uk/ccrl/404.live/
http://computerchess.org.uk/ccrl/4040.live/

Originally posted on Chess2u.com

Re: CCRL Testing IvanHoe

Posted: Fri Mar 09, 2012 6:15 pm
by ernest
kingliveson wrote:The world must truly be coming to an end. :)
Why? :shock:

Re: CCRL Testing IvanHoe

Posted: Fri Mar 09, 2012 6:37 pm
by ernest
You mean, they are finally testing IvanHoe, that's it? ;)

BTW, couldn't see if they are using Robbobases...

Re: CCRL Testing IvanHoe

Posted: Fri Mar 09, 2012 10:47 pm
by Adam Hair
ernest wrote:You mean, they are finally testing IvanHoe, that's it? ;)

BTW, couldn't see if they are using Robbobases...
Brent Magnusson, who is doing the 6 CPU testing, uses Robbobases. I am not certain if the other IvanHoe testers are using them.

Re: CCRL Testing IvanHoe

Posted: Fri Mar 09, 2012 10:58 pm
by Adam Hair
kingliveson wrote:The world must truly be coming to an end. :)

http://www.computerchess.org.uk/ccrl/404.live/
http://computerchess.org.uk/ccrl/4040.live/

Originally posted on Chess2u.com
Maybe so :)

I am sure this has been asked and answered somewhere. But, could you answer something for me? IvanHoe 999946h shows up as IvanHoe 9.46b in Fritz. Could you explain why the internal name and the name of the executable differs?

Thanks,
Adam

Re: CCRL Testing IvanHoe

Posted: Sun Mar 11, 2012 1:56 am
by kingliveson
Adam Hair wrote: I am sure this has been asked and answered somewhere. But, could you answer something for me? IvanHoe 999946h shows up as IvanHoe 9.46b in Fritz. Could you explain why the internal name and the name of the executable differs?

Thanks,
Adam
It is a way of shortening the version string. Without the change, 999946h would appear as IvanHoe VERSION 999946h-RobboLib, and 999947c, IvanHoe-Beta VERSION 999947c.

9. = 9999
## = last 2 digits
b = beta

IvanHoe999946a and IvanHoe99946h both appear as IvanHoe 9.46b. This apparently is an issue for Fritz GUI because it stores UCI configuration using name/version rather than the executable. It is easily resolved if one insists on using an older version which probably has more bugs, by simply renaming the .uci file. Depending on version of Windows (hidden files/folders may need to be enabled), the configuration files are located in:

C:\Users\yourname\AppData\Roaming\ChessBase\Engines.UCI
C:\Documents and Settings\yourname\Application Data\ChessBase\Engines.UCI

It probably would have been wiser to leave out the 'b' and use the sub-version lettering where it mattered. The first 12 public releases did not have letters to differentiate main version. That said, blame the developers for having this peculiar backwards versioning. :P

Re: CCRL Testing IvanHoe

Posted: Sun Mar 11, 2012 2:51 am
by Adam Hair
kingliveson wrote:
Adam Hair wrote: I am sure this has been asked and answered somewhere. But, could you answer something for me? IvanHoe 999946h shows up as IvanHoe 9.46b in Fritz. Could you explain why the internal name and the name of the executable differs?

Thanks,
Adam
It is a way of shortening the version string. Without the change, 999946h would appear as IvanHoe VERSION 999946h-RobboLib, and 999947c, IvanHoe-Beta VERSION 999947c.

9. = 9999
## = last 2 digits
b = beta

IvanHoe999946a and IvanHoe99946h both appear as IvanHoe 9.46b. This apparently is an issue for Fritz GUI because it stores UCI configuration using name/version rather than the executable. It is easily resolved if one insists on using an older version which probably has more bugs, by simply renaming the .uci file. Depending on version of Windows (hidden files/folders may need to be enabled), the configuration files are located in:

C:\Users\yourname\AppData\Roaming\ChessBase\Engines.UCI
C:\Documents and Settings\yourname\Application Data\ChessBase\Engines.UCI

It probably would have been wiser to leave out the 'b' and use the sub-version lettering where it mattered. The first 12 public releases did not have letters to differentiate main version. That said, blame the developers for having this peculiar backwards versioning. :P
Your explanation triggered my memory :) .
Thanks for the information.