Author: Kolss
Date: 10:27:51 10/13/05
Go up one level in this thread
On October 13, 2005 at 12:35:36, Mike Hood wrote:
>On October 13, 2005 at 12:16:42, Kolss wrote:
>
>>On October 13, 2005 at 12:08:49, Mike Hood wrote:
>>
>>>There have been complaints about Chessbase's UCI (and Winboard) adaptors for a
>>>long time. Forgive me if I'm being stupid, but why doesn't someone just sit down
>>>and write a new adaptor than can be used to replace the current UCI.dll (Fritz
>>>8) or UCINet.dll (Fritz 9)? And if Chessbase objects to the distribution of a
>>>freeware UCI adaptor, sell it to them.
>>
>>
>>Hi,
>>
>>The problem is that no one - except, of course, a few Chessbase guys - knows the
>>Fritz interface. You can only write an adapter if you know *both* sides. The
>>current one takes input in "UCI format" and converts it into "Chessbase format",
>>whatever that is exactly. However, there seem to be some problems with it - we
>>("Ikarus team") have never really tested it properly though...
>>I hope this is more clarification than confusion!
>>
>>Best regards - Munjong.
>
>I suspected something like this when I posted my message. (Maybe I'm not as
>naive as I pretend to be). The best candidates for writing a new UCI-Chessbase
>adaptor would be someone who has already written a Chessbase-compatible engine,
>and there are a few some people who use this forum.
Hi again,
Well, I suppose that in principle that would be possible. E.g., we (or any of
the young talents or the Junior team or C. Théron or...) could theoretically
write an adapter UCI<->Ikarus-CB. You would then have to buy the Ikarus engine
from Chessbase to get the adapter Ikarus-CB<->Chessbase-native (called
"Ikarus.eng"). That would result in a double adapter
UCI<->Ikarus-CB<->Chessbase-native, which is a bit ugly, though not really new
if you think about the WB<->UCI<->Chessbase-native way. But remember that Ikarus
(or the other young talents or...)compiled for Chessbase runs as a dll; I don't
know enough about interface programming, is it possible to call a new process
from a dll?! Besides that, I don't think we would be allowed to do this anyway,
as we agreed to not distribute the Ikarus-CB interface.
I would propose as an alternative solution to create an ini-file for your
UCI-engine from which it reads all parameters (such as hashtable size) and then
ignores commands from the GUI.
By the way, has anyone (other young talents?!) tested whether this "bug" in the
Chessbase-UCI really affects playing strength measurably? I have never run any
serious tests, but a few hundred blitz games of our "native" against an
otherwise identical UCI-version (without the above-mentioned hack; and no, I did
not check whether the "bug" actually occurred at all - it wasn't even known back
then) failed to show anything close to significant. If I remember correctly
(must be more than two years ago), the UCI-version actually scored something
like +2 in 300 or so games. That is, of course, only a very crude test, but I
expect that it will be difficult to find a real strength penalty caused by the
"buggy" Chessbase-UCI adapter.
Best regards - Munjong.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.