Computer Chess Biggest Liar

General discussion about computer chess...
kranium
Posts: 55
Joined: Mon Aug 02, 2010 10:49 pm
Real Name: Norman Schmidt

Re: Computer Chess Biggest Liar

Post by kranium » Sun Oct 23, 2011 3:08 pm

from Houdini_15a_w32.exe search:

Code: Select all

    sub_53C12C((int)"bestmove %s ponder %s\n", v14);
    }
    else
    {
      sub_53C12C((int)"bestmove NULL\n", v17);
    }
    v15 = sub_53C895();
    fflush((FILE *)v15 + 1);
  }
  result = timeEndPeriod(1u);
  byte_14C69F7 = 1;
  return result;
}
from Robbolito 0.09 search():

Code: Select all

    output_best_move();
    timeEndPeriod(1);
    }

Tony Mokonen
Posts: 16
Joined: Mon Jun 14, 2010 7:11 pm

Re: Computer Chess Biggest Liar

Post by Tony Mokonen » Wed Oct 26, 2011 3:08 am

kranium wrote:In last versions of Chesslogik Robbolito (e series, g3, and 0.09), which Sentinel and I released open-source under GPL on 12/31/2009 (or earlier)
we implemented (what I consider to be) a very unusual idea:

We linked to a Microsoft library via
#pragma comment(lib,"winmm.lib")
in Robbolito.h

winmm.lib is a 'multimedia' library, with sound routines for wave, midi, mci, etc. programming...quite unusual for a console chess application

Sentinel's idea was to improve time accuracy, and there were 3 very interesting functions we wanted to use:
timeBeginPeriod
timeEndPeriod
timeGetTime
I can name one other program that uses these functions - my own crappy little program Tony's Chess. I Googled for high resolution timer functions, and chose to use this set of functions. However, I am not sure if they are the best choice. According to http://msdn.microsoft.com/en-us/library ... 85%29.aspx

"This function affects a global Windows setting. Windows uses the lowest value (that is, highest resolution) requested by any process. Setting a higher resolution can improve the accuracy of time-out intervals in wait functions. However, it can also reduce overall system performance, because the thread scheduler switches tasks more often. High resolutions can also prevent the CPU power management system from entering power-saving modes. Setting a higher resolution does not improve the accuracy of the high-resolution performance counter."

Hood
Posts: 200
Joined: Thu Jun 10, 2010 2:36 pm
Real Name: Krzych C.

Re: Computer Chess Biggest Liar

Post by Hood » Thu Nov 03, 2011 8:49 pm

kingliveson wrote:
I have no idea what you are talking about. We are taking about a bold-face liar.
If someone is lying 1,2 times it does not mean he shall be called a liar. Were you always telling a truth in your life?

Rgds
Hood.
Smolensk 2010. Murder or accident... Cui bono ?

There are not bugs free programms. There are programms with undiscovered bugs.
Alleluia.

hyatt
Posts: 1242
Joined: Thu Jun 10, 2010 2:13 am
Real Name: Bob Hyatt (Robert M. Hyatt)
Location: University of Alabama at Birmingham
Contact:

Re: Computer Chess Biggest Liar

Post by hyatt » Thu Nov 03, 2011 9:21 pm

I don't think the case being addressed was one of just "one or two lies."

Prima
Posts: 328
Joined: Tue Dec 14, 2010 6:12 am

Re: Computer Chess Biggest Liar

Post by Prima » Thu Nov 03, 2011 10:30 pm

Hood wrote:
kingliveson wrote:
I have no idea what you are talking about. We are taking about a bold-face liar.
If someone is lying 1,2 times it does not mean he shall be called a liar. Were you always telling a truth in your life?

Rgds
Hood.
And where do you get this inference from? If someone steals just ONCE, that makes the person a thief. If someone lies just ONCE, that makes the person a liar. If someone rapes just once, that person is a rapist. If you kill once, that makes you a murderer...

However, whether AND/OR because such acts & the likes have been done by others, is an entirely different/separate prospect and SHOULD NOT be misconstrued as the norm or "acceptable behaviour".

Right is right, just as wrong is wrong.

hyatt
Posts: 1242
Joined: Thu Jun 10, 2010 2:13 am
Real Name: Bob Hyatt (Robert M. Hyatt)
Location: University of Alabama at Birmingham
Contact:

Re: Computer Chess Biggest Liar

Post by hyatt » Fri Nov 04, 2011 12:41 am

In this "new and improved morality" the idea seems to be that wrong is not wrong if others are also doing wrong. Doesn't matter whether we are talking majority, minority, or just a very small group. As long as you have "someone" doing the same thing you are doing, it can't be wrong. Not sure where this came from, however...

Hood
Posts: 200
Joined: Thu Jun 10, 2010 2:36 pm
Real Name: Krzych C.

Re: Computer Chess Biggest Liar

Post by Hood » Fri Nov 04, 2011 3:44 pm

Prima wrote:
Hood wrote:
kingliveson wrote:
I have no idea what you are talking about. We are taking about a bold-face liar.
If someone is lying 1,2 times it does not mean he shall be called a liar. Were you always telling a truth in your life?

Rgds
Hood.
And where do you get this inference from? If someone steals just ONCE, that makes the person a thief. If someone lies just ONCE, that makes the person a liar. If someone rapes just once, that person is a rapist. If you kill once, that makes you a murderer...

However, whether AND/OR because such acts & the likes have been done by others, is an entirely different/separate prospect and SHOULD NOT be misconstrued as the norm or "acceptable behaviour".

Right is right, just as wrong is wrong.
I do not agree, if someone made a mistake it does not mean that he will repeat that and shall be responsible and penalised during all his life!

Somebody is being the liar while he (she) is lying but while she(he) is telling the truth he(she) is not being the liar.
So we can not judge the people on the base of one or 2 facts for all life.

Rgds
Hood
Smolensk 2010. Murder or accident... Cui bono ?

There are not bugs free programms. There are programms with undiscovered bugs.
Alleluia.

Hood
Posts: 200
Joined: Thu Jun 10, 2010 2:36 pm
Real Name: Krzych C.

Re: Computer Chess Biggest Liar

Post by Hood » Fri Nov 04, 2011 3:53 pm

It has nothing to the new morality but to the reality.

If someone told one time true is he a true speaker? Not :-)

If someone told one time false is he the liar? Not. :-) either.

One or some facts are not deciding for all the life. People are changing themselves so such estimation as liar or not liar is mistaken. True speaker may become a liar and the liar may become a true speaker.

You can only tell that someon is lying and is beina a liar at that time not for always (is liar)
or is telling the true and is being a true speaker at that time not always.

Rgds
Hood

hyatt wrote:In this "new and improved morality" the idea seems to be that wrong is not wrong if others are also doing wrong. Doesn't matter whether we are talking majority, minority, or just a very small group. As long as you have "someone" doing the same thing you are doing, it can't be wrong. Not sure where this came from, however...
Smolensk 2010. Murder or accident... Cui bono ?

There are not bugs free programms. There are programms with undiscovered bugs.
Alleluia.

Prima
Posts: 328
Joined: Tue Dec 14, 2010 6:12 am

Re: Computer Chess Biggest Liar

Post by Prima » Fri Nov 04, 2011 7:45 pm

Hood wrote:I do not agree, if someone made a mistake it does not mean that he will repeat that and shall be responsible and penalised during all his life!

Somebody is being the liar while he (she) is lying but while she(he) is telling the truth he(she) is not being the liar.
So we can not judge the people on the base of one or 2 facts for all life.

Rgds
Hood
Intentional or mistake DOES NOT excuse the nature of the crime and/or unacceptable behaviour. To put it in perspective: I drive inebriated and caused a fatal accident. Though I did not intend to kill someone, should it excuse the fact I killed someone, albeit by mistake, or more factually, by means of MY reckless, irresponsible behaviour? No. I'll be charged with manslaughter. And rightfully so. It has absolutely NOTHING to do with the probability that I will "mistakenly" cause another fatal accident via drunk driving. This is NOT the case of Robert Houdart relative to Houdini & RobboLito 0.85g3 (GPL) and Vas Rajlich, relative to Rybka.

Intentional act is quite obvious on the premeditation of the wrong-doer and usually receives the highest punishment. This is the category Robert Houdart is in.

The connotation that Robert Houdart made a "mistake" is demonstrably easy to prove/disprove, if not laughable. When one does something by mistake, the assumption is the person did not intend on the "unacceptable" act. In Robert Houdart's case (and Vas Rajlich), the unacceptable act is the "lies". So when light is shed on the matter and the person finds him/herself in the same situation again, a different & right course of action is to be expected to avoid the "previous mistake", right?

Ask Robert Houdart, again, if Houdini is based, code-wise, on RobboLito 0.85g3 et.al. Assuming he made a "mistake" & is repentant of it and doesn't evade the issue, he should be truthful relative to Houdini's origins/code make-up and NOT make the previous "mistake" of lying --- AGAIN.

By the way, judging someone is quite different from stating something based on proven facts. Driver-1 caused a fatal accident by means of drunk driving, while Driver-2 intentionally drove her car to hit & kill her ex-boyfriend. Both killed someone. One did it by mistake (fact). The other did it intentionally (fact) and based on evidence, we can conclude that Driver-2 WAS intentionally nefarious. Just because she is provably nefarious does not mean she was judged based on "personal feelings" about her.

kranium
Posts: 55
Joined: Mon Aug 02, 2010 10:49 pm
Real Name: Norman Schmidt

Re: Computer Chess Biggest Liar

Post by kranium » Sat Nov 05, 2011 9:20 pm

Houdini wrote:
Frank Quisinsky wrote: Peter:
Robbolito is GPL
Only the first unclear Igorit sources are PD
Frank,
Not for the first time, I'm afraid that you are very badly informed.

For a Robbolito source code without any license, you have to look no further than the Ippolit web site: http://ippolit.wikispaces.com/RobboLito .

You will find Robbolito version 0.085f1 with Italian/Albanian (?) file names, variables etc. I have no idea who created this release, but I doubt that it was Norman. It may be the person that identifies himself as "Roberto Pescatore" (= "Bobby Fischer" - so probably not his real name) on the Ippolit web site.

The so-called "GPL-licensed" version Norman released later is virtually identical to this 0.085f1 version, if you inspect the source codes you will find no changes other than translating files and variables to English and reorganizing spaces.

Cheers,
Robert
??
Sorry Robert, apparently Frank isn't 'badly informed' at all, perhaps you should check your facts before posting such nonsense....
of course, i can understand you would certainly like everyone to believe your particular 'spin' on the subject.

But Robbolito 0.085g3 and 0.09 compared to Robbolito version 0.085f1 are like night and day...

here are some of the changes implemented in 0.09 alone

-The source code readability greatly improved, more translations added and updated, indentation and white space usage unified and standardized
-All new features easy adjustable during compilation time by setting appropriate switches in robbolito.h
-Alignment of hash structures improved
-New, improved bit scan functions
-Improved rand functions (Mersenne Twist) implemented
-New version of popcnt function as an option
-Cache prefetching added
-New smooth scaling of null move, inspired by Dann Corbit's idea, and modified for Robbolito
-Improved time management especially in time controls without increment and tournament,
-Time management var Increase_Depth communicates with search to ensure proper search depth is reached before returning from search
-Move on PonderHit

--------------------------------
RobboLito 0.09 is greatly improved, and stronger than any original wikispaces Ippolit release...

here's an image of the longest running 1 CPU rating list from a well-known Russian CC site:
Image

Take a close look at places 6 and 7...

(PS -Robbolito version 0.085f1 doesn't even make the list...)

Post Reply