Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Uci in Fritz 8 Gui is generally ok!

Author: Tord Romstad

Date: 03:49:24 01/13/04

Go up one level in this thread


On January 12, 2004 at 18:19:49, Mike S. wrote:

>Home come only *some* UCI engines are affected by that problem, and others not?
>I've just tried the UCI engines
>
>Aristarch,
>Pharaon,
>Ruffian,
>SOS
>and Yace
>
>in the Fritz GUI checking the free RAM with a memory tool, and with *all* of
>them the amount of RAM was allocated which I'd set in the F3 engine dialogue.
>Note that some of the strongest (non-chessbase) engines are among them.

What you and many others seem to have difficulties to grasp is that the
problem only appears *sometimes*, not *always*, and that it is difficult
to see any pattern in when it occurs (except that it seems to happen
only in the second and subsequent rounds of tournaments with multiple
engines).

Matthias and I are not the only ones who have noticed this problem,
by the way.

>So where may the reason be, when only some engines have this problem (and
>obviously not the strongest one's :-))? In Fritz, or in these engines rather?
>
>But if you intend to ruin your own reputation as a serious member of
>computerchess community, launching such unfounded accusations like above is a
>perfect way to achieve that. With that, you can follow in the footsteps of a
>number of persons with a certain history :-) probably some of them will gather
>together below of my posting, as always in such cases.
>
>Again: Many engines, i.e. the above mentioned, don't have that problem.

Or rather, you didn't see any problems in the limited experiments you
performed.  Engine authors who study log files of the communication
between the GUI and the engine in a long series of games can easily
spot problems which are unnoticed by ordinary users.

>If anybody would really consider that a major professional chess software
>publisher intends to make dirty hashize tricks against enginges which are
>*hundreds of elos weaker*, then I'd say even an average intelligent 12 year old
>would have a better sense for reality.

Engines like Deep Sjeng and Ruffian are most definitely not "hundreds
of elos weaker" than any of the CB native engines.

For the record:  I personally think this is just a bug in the CB GUIs,
and not an intentional attempt to cripple UCI engines.  But the problem
is very real, and I am very disappointed that CB does not seem to be
interested in fixing it.

If you are unaware of the history:  There is a good reason why a lot of
people suspect an attempt of cheating from CBs side.  This is not the
first time there is a serious problem with the support for non-native
engines.  A few years ago, there was a "Winboard adapter" for the CB
GUIs, which was notorious for its non-standard implementation of the
protocol, and which seriously crippled many of the strongest Winboard
engines.

Again, I don't think this was an attempt to cheat, but it is
disappointingly unprofessional when coming from a company like ChessBase.
I would personally prefer that they removed all support for UCI and/or
Winboard engines from their GUIs rather than that they keep using some
sort of buggy or non-standard interface.

Tord




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.