Computer Chess Biggest Liar

General discussion about computer chess...
Post Reply
User avatar
thorstenczub
Posts: 592
Joined: Wed Jun 09, 2010 12:51 pm
Real Name: Thorsten Czub
Location: United States of Europe, germany, NRW, Lünen
Contact:

Re: Computer Chess Biggest Liar

Post by thorstenczub » Thu Oct 13, 2011 8:43 am

Prima wrote: And the crime continues...
here in germany, when a thief steals something in a shop, he is put into a special room and waits until the police comes. then the police writes down everything and thats it.

a judge will do a punishment. the shop gets the stolen stuff back. the thief has to pay a punishment and is not allowed to ever enter the shop again.

if you want to make a comparison:

the police, the judge, the shop , the thief...

the police and the judge are not the same people.

By your philosophy, society can complain on the crimes they observed / know for a fact happened, all they want, but nothing gets to be resolved. And the crime continues...
as long as people are hungry or ill they will steal.

you cannot stop thiefs by telling them not to steal.

the best way to stop stealing is to increase living conditions of all.

never the judge or the police or even the politician should be the same force.

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

Re: Computer Chess Biggest Liar

Post by Hood » Thu Oct 13, 2011 3:42 pm

[quote="Jeremy Bernstein"]

But ok, point taken. There are delicate flowers reading this forum, so I'll stop there. ;)

Jeremy

It has nothing to the delicateness.

I think that evryone of us wants to be secured from being offened without a proof or by impoliteness of someone.
There are not persons on the forum who are or can be judges of someone.
It is clear so no one is allowed to do such judgement.

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

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

Jeremy Bernstein
Site Admin
Posts: 1226
Joined: Wed Jun 09, 2010 7:49 am
Real Name: Jeremy Bernstein
Location: Berlin, Germany
Contact:

Re: Computer Chess Biggest Liar

Post by Jeremy Bernstein » Thu Oct 13, 2011 4:07 pm

Hood wrote:
Jeremy Bernstein wrote:
But ok, point taken. There are delicate flowers reading this forum, so I'll stop there. ;)

Jeremy

It has nothing to the delicateness.

I think that evryone of us wants to be secured from being offened without a proof or by impoliteness of someone.
There are not persons on the forum who are or can be judges of someone.
It is clear so no one is allowed to do such judgement.

Rgds
Hood
I am afraid that I disagree. This is a community, and a community sets its own standards. Until recently, one of the standards of this community was a strong aversion to cloning. Now that the previous top engine has been exposed as a derivative and the current top engine is a clone, the hard line against clones has softened: apparently it's most important to (many) people to get the strongest software possible, regardless of what is done/what is stolen/who is defrauded/what amount of dishonesty is exerted to achieve it.

If the community cannot provide guidance to its members based on important standards and principles, there's no point in having them at all. And by extension, there's little point to the community -- if you want law of the jungle, you can get that anywhere.

jb

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

Re: Computer Chess Biggest Liar

Post by Hood » Thu Oct 13, 2011 6:32 pm

Jeremy Bernstein wrote: I am afraid that I disagree. This is a community, and a community sets its own standards.

jb
I am sorry I doubt on the base of observed quarrels if it is a community .
Even if it is community where anyone can offend someone their life will be short.

The problem is - do we accept the fact that publishing Ippo code was a real revolution and world has changed or not.
If we have not accepted that fact we will fall in the jungle of suspicions, invigilations, quarrels etc. and it will be the end.
The power will be consumed for prosecuting not for developing.

I do not see the other choice then abolishement of that what is connected with publishing Ippo codes. Other way is the way to nowhere.

i.e.
1)++i; i++;
2)i+=2;
3)++i;++i;

are differrent but the result is the same.

Optimizing compiler shall convert that 3different sets to the same set of assembler instructions.

Disassembling such a code will show that the source code is identical but it was not.
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 Oct 13, 2011 6:39 pm

No, disassembling will show that the SEMANTICS are identical. The semantic behavior is to add 2 to I. Nobody looks at a single statement, we looked at blocks of code. The chance of two different people incrementing I by 2 is fairly high. The chance of two different people producing the same block of code, AND, incrementing I by 2 at the exact same place is vanishingly small.

Disassembling can't show anything about "the source." One CAN compare a source program to a binary to see if they are semantically equivalent. If so, the chances that the binary was developed independently of the source in question is essentially zero.

Jeremy Bernstein
Site Admin
Posts: 1226
Joined: Wed Jun 09, 2010 7:49 am
Real Name: Jeremy Bernstein
Location: Berlin, Germany
Contact:

Re: Computer Chess Biggest Liar

Post by Jeremy Bernstein » Thu Oct 13, 2011 6:57 pm

Hood wrote:i.e.
1)++i; i++;
2)i+=2;
3)++i;++i;

are differrent but the result is the same.

Optimizing compiler shall convert that 3different sets to the same set of assembler instructions.

Disassembling such a code will show that the source code is identical but it was not.
Compiling those statements in gcc 4 with optimizations will yield this:

Code: Select all

:
	pushq	%rbp
	movq	%rsp, %rbp
	xorl	%eax, %eax
	popq	%rbp
But we're talking about dozens of identical or nearly identical assembly instructions, not a simple increment. An increment isn't functionality, it's an operation. Operations strung together yield algorithms. And algorithms strung together yield complete programs. We're talking about lengthy sequences of identical or nearly identical algorithms which can be found in Robbo and in Houdini. Your point simply doesn't make any sense, it's a reductio ad absurdum.

jb

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

Re: Computer Chess Biggest Liar

Post by Hood » Thu Oct 13, 2011 7:31 pm

'reductio ad absurdum' was a method of Poirot detective :-).
test your compilation against: i++; and l++ :-)

It makes a deep sense. The example was short and trivial to demostrate the problem.
As we know copying, using of ideas is not forbidden and writing a similar code is ok.

Lets assume that someone is adapting other idea writing his own code which has to (shall) be similar. Programing is using standard schemes. So ... We do have similar source code compiled and optimized. It is getting more similar to origin after compilation.

only disasembling is not a prove. It is showing sth but it is not 100% prove.
Smolensk 2010. Murder or accident... Cui bono ?

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

Jeremy Bernstein
Site Admin
Posts: 1226
Joined: Wed Jun 09, 2010 7:49 am
Real Name: Jeremy Bernstein
Location: Berlin, Germany
Contact:

Re: Computer Chess Biggest Liar

Post by Jeremy Bernstein » Thu Oct 13, 2011 7:50 pm

Hood wrote:'reductio ad absurdum' was a method of Poirot detective :-).
test your compilation against: i++; and l++ :-)

It makes a deep sense. The example was short and trivial to demostrate the problem.
As we know copying, using of ideas is not forbidden and writing a similar code is ok.

Lets assume that someone is adapting other idea writing his own code which has to (shall) be similar. Programing is using standard schemes. So ... We do have similar source code compiled and optimized. It is getting more similar to origin after compilation.

only disasembling is not a prove. It is showing sth but it is not 100% prove.
Here's an anecdote: I work for a small software company, maybe 10 full-time engineers during crunch times, like now. When we're not just before a release, more like 7 full-time, and with comings and goings, let's say I've had about 15-20 co-programmers in the last 12 years. Yet I can pop open a random source file from our 3 GB source code repository, look at a random function and tell you with probably 85% accuracy which of my current or former co-workers wrote it, or last edited it.

Because each one of us has a unique 'fingerprint' to the way we think about code: and I'm not talking about comments, punctuation, spelling, variable names or similar. I'm talking about form, structure, organization, how algorithms are encapsulated, how long/short/specific/general the code blocks are. How error checking is performed. How data structures are conceived and even how the data is packed within. Even just how the code flows on the page is different. And this after 12 years of intensive cooperation and mutual growth with one another.

And I can tell you with 100% certainty: given any problem which isn't ass-simple to solve, we would each solve it in a completely different way -- in a way which would under no circumstances yield identical or near-identical assembly output. I think that you have only very general knowledge about this stuff, and there's no particular reason why you need to believe me. But anyone on this forum who has programmed with others knows that I'm telling the truth.

jb

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 Oct 13, 2011 10:02 pm

He might well not understand your example, so here's one everyone can relate to.

1. Take a book by Matt Reilly. Tear the cover off. White out the title. Give it to someone that reads a lot of fiction. You won't get past chapter one before figuring out who wrote the thing. Reilly has the fastest-pace books I have ever read.

2. Take a book by Clive Cussler. Do the same. Cussler always starts off with some historic "fabrication" that seems irrelevant until 1/2-2/3 of the way thru his story where it suddenly becomes a critical linchpin that makes the "great escape" possible.

3. The same is true for many other authors. Once you have read several books by a given author, you can read other books and pick out his in several ways. The cast of characters. The way the story line is developed. The way the characters speak/act.

This is just as true of a programmer. Programming languages offer a rich set of choices in how to express an algorithm. So rich that two students won't write the same basic algorithm (say a sort) in the same way. Much less a chess program or a big part of a chess program. The way a person programs is almost as unique as their finger print or retinal scan or DNA.

To think two different programmers (Vas and Fabien) would write so much code that is EXACTLY the same is simply a ridiculous idea in the first place. And the last place. And anywhere in between.

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

Re: Computer Chess Biggest Liar

Post by Prima » Fri Oct 14, 2011 12:01 am

thorstenczub wrote:as long as people are hungry or ill they will steal.

you cannot stop thiefs by telling them not to steal.

the best way to stop stealing is to increase living conditions of all.

never the judge or the police or even the politician should be the same force.
This has to be yet another bizarre excuses supporting thievery.
Of course society can't stop thieves from stealing. However, there are penalties for stealing [crimes], as a means of deterrent. These deterrents should give anyone a "clue", if not from a morally/ethically obligation.
The best way to stop stealing is to instill discipline/morals/ethics in the youths [your children] from birth. When these children grow up, they will not depart from the good code of conduct - for better or worst living conditions. But frivolous reasons OR personal "human rights" for committing them will not obviate enforcement of the law/ethics/morals. Stealing is stealing. That's "a bad" thing, and there are no excuses.
thorstenczub wrote:if you want to make a comparison:

the police, the judge, the shop , the thief...

the police and the judge are not the same people.
No, but the judge and the police force have a common goal: law. Still, a police officer can exercise his/her discretion, relative to who committed the crime, to either take the suspect to jail or not, based on factors such as evidences, facts, suspects' behaviour and criminal/non-criminal history etc. Ditto a judge's ruling. Just because judges and the police forces are not literally the same enforcement unit does NOT make it acceptable for RH, or anyone for that matter, to run amok and do whatever they please. Otherwise every jail / prison in all countries, world-wide, will be filled up and billions more to go. Do you think countries worldwide have laws & police forces just so people get to be lawless & break them without consequences?

Just curios....
1. what's your take on current Fruit / Crafty => Rybka/Vas case?
2. Was it okay for VR to plagiarized Fruit / Crafty codes without due credit to their respective authors, claiming "Rybka is 100% original at the source code level", but we found that to be a lie?
3. Was it OK for the ICGA to ban Vas?
4. Do you think Bob Hyatt etc, should not have dug up the truth?
5. What do you think about Fabien Letouzey and the FSF considering the legal avenue?

Post Reply