Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / A Big Thanks and a Small Update
1 215 16 17 18 19 20 21 Previous Next  
Parent - By Rebel (****) Date 2012-01-15 12:41

> You have no clue about "semantic equivalence."

Made it a new topic due to its RE importance.
Parent - - By Venator (Silver) Date 2012-01-15 12:53
So, what did you take from Fruit? It is clear that the Crafty speed up must come from somewhere and I am 100% sure it is not from you.
Parent - - By bob (Gold) Date 2012-01-15 15:17
WHAT are you talking about?
Parent - By Venator (Silver) Date 2012-01-15 16:34
Did I write Chinese??

BTW, please answer this:
Parent - Date 2012-01-14 09:29
Parent - - By Homayoun_Sohrabi_M.D. (***) Date 2012-01-14 05:12
"CRAPty" ???

I am sure Bob is not too offended, in fact he'll be ok with anything except "more to come".
Parent - - By Rebel (****) Date 2012-01-14 08:14
I was in a poetic mood.

> I am sure Bob is not too offended, in fact he'll be ok with anything except "more to come".

There isn't more to come, after Mark's poor performance of yesterday I am done with the issue.
Parent - - By bob (Gold) Date 2012-01-14 15:31
Does explain a lot.  "pro deo" is well over 200 Elo BEHIND Crafty in the CCRL list.  I guess that makes your program "not worth a crapty"???
Parent - By Venator (Silver) Date 2012-01-14 15:36
95% of Crafty's elo points can be attributed to others. So it really can be considered a 'multi author's derivative' :lol:
Parent - - By AWRIST (****) Date 2012-01-13 11:54
Bob wrote: Again, I assume you are trying to sway the opinions of those that are NOT "technically competent" with such erroneous and distorted statements, and there you MIGHT win a vote or two.

This is still better than calling all who disagree with you "idiots and monkeys", dont you think so?
Parent - - By bob (Gold) Date 2012-01-14 04:20
I haven't called anyone a "monkey"...
Parent - - By Venator (Silver) Date 2012-01-14 10:22
"Idiots" and "stupid" are, however, two of your favourites when you are losing the argument.
Parent - - By AWRIST (****) Date 2012-01-14 12:37
Perhaps this helps if I elaborate, that Bob assumed that my room from where I type this must be full of monkeys. Bob perhaps never was in Gibraltar, but otherwise he couldnt assume that I would be the typist but not the monkeys. :lol:
Parent - - By Venator (Silver) Date 2012-01-14 14:08
Have you noted that Bob's behaviour, when he is losing the argument, is rather semantically equivalent!? :lol:
Parent - - By AWRIST (****) Date 2012-01-17 12:41
If you take Bob from the front, he is almost identical like from behind.
Parent - - By Alkelele (****) Date 2012-01-17 12:48
Nah, just semantically equivalent.
Parent - By AWRIST (****) Date 2012-01-17 13:14
Sorry for being too childish. b-o-b
Parent - - By bob (Gold) Date 2012-01-14 15:32
Better look again.  I get those insults directed at me FAR more than I direct them at others.  Yourself included...

I'm an adult.  I know how to ignore things that don't matter.
Parent - By Venator (Silver) Date 2012-01-14 15:34
Hyatt quote:

"I am completely unconcerned about the reverse-engineering that has been done. Seems like a fair way to "even the playing field" by forcing a secretive author to expose secrets... I'm not going to lose any sleep over this at all."

Just tells it al, doesn't it?
Parent - - By Vasik Rajlich (Silver) Date 2012-01-11 21:30

there is no transitive property for cloning. Otherwise, I could take Now, decompile it, add some Robbolito code, and dare you to complain. :smile:

Parent - - By mjlef (***) Date 2012-01-12 20:42
Robbolito is hundreds of points stronger than Now, so you would do much better starting with 98% robbo and 2% Now! :-)
Parent - By Vasik Rajlich (Silver) Date 2012-01-12 22:39
Yes! The Robbo guys would then be terrified of accusing me of anything. :smile:

Parent - By AWRIST (****) Date 2012-01-13 11:48
No, Vas, that's way too specific! :lol:
Parent - - By Jury Osipov (**) Date 2012-01-12 19:20
I agree with Vas.
Strelka (1.0 beta, 1.8 and 2.0B) was a clone (or derivative) of Rybka 1.0 beta.
But mostly it is a clone of Fruit 2.1.
Let's compare the sizes of EXE:
Fruit 2.1 : 184 320 bytes,
Strelka 2.0B: 184 320 bytes,
Rybka 1.0 beta: 2 514 944 bytes.
Parent - - By Vasik Rajlich (Silver) Date 2012-01-12 22:40
Please Jury, not this again. :smile:

Parent - - By turbojuice1122 (Gold) Date 2012-01-13 03:18
Can someone with the executables confirm this with the file sizes?  It's primarily interesting because it shows an intention of making that file size, since it has parts of both programs in it, obviously, in order to play the same moves every time as Rybka.  In other words, it would show an intention to cause trouble. :smile:
Parent - By Vasik Rajlich (Silver) Date 2012-01-13 07:49
Or it would show a twisted sense of humor. :smile:

Parent - - By Gaмßito (****) Date 2012-01-13 05:19

Three questions off topic:

1. Is the new Strelka 5.1 a clone of Houdini 1.5a?
2. Which are the biggest changes in Strelka 5.1?
3. How much time you taked to decompile it?

Parent - - By Jury Osipov (**) Date 2012-01-13 17:14
Я вам ничего говорить не буду, иначе я опять что-нибудь скажу.

(Sorry, I don't know how to translate it).
Parent - - By nebulus (****) Date 2012-01-13 17:38
Something like: "The less I talk, the better."
Parent - - By Graham Banks (*****) Date 2012-01-13 20:43

> Something like: "The less I talk, the better."

Better for who? Everybody likes being evasive. Doesn't help at all.
Parent - By nebulus (****) Date 2012-01-13 21:07
Maybe it has something to do with Houdini. I don't know... ask him. More accurately he said: "I'm not going to say anything, otherwise I might again say something." But what is this something supposed to be? Indeed doesn't help much.
Parent - By Hannibal (**) Date 2012-01-13 21:28 Edited 2012-01-13 21:33

> 2. Each program must be the original work of the entering developers. Programming teams whose code is derived from or including game-playing code written by others must name all other authors, or the source of such code, in their submission details. Programs which are discovered to be close derivatives of others (e.g., by playing nearly all moves the same), may be declared invalid by the Tournament Director after seeking expert advice. For this purpose a listing of all game-related code running on the system must be available on demand to the Tournament Director.

Above is the rule you have been charged with breaking.  Could you elaborate on:

1) Your understanding of the word "derived from" in this context

2) Some examples of when and why a program would be accurately termed a "close derivative" of another
Parent - - By Permanent Brain (*****) Date 2012-01-17 22:32 Edited 2012-01-17 22:36
Xuse me folks, 1300+ :eek: postings in ONE thread, please let's stop the sickness. Vas' last posting was on January 13th, and now I guess we should let things cool down until something substantial happens.
Parent - - By Rebel (****) Date 2012-01-17 22:38
New record ?
Parent - - By Permanent Brain (*****) Date 2012-01-18 13:22

> New record ?

I guess so. I don't recall having seen 1,000+ postings in the same thread, on any computer chess message board, before.
Parent - By Uly (Gold) Date 2012-01-18 21:57
Sorry, I had to branch some threads because it was a big mess.
Parent - By Lukas Cimiotti (Bronze) Date 2012-01-17 22:42
Very good idea. In fact Vas told me he'd only come here for a few days - which are over now. I think he won't be back before Rybka 5 beta testing starts.
Parent - Date 2012-02-05 04:02
- - By John Sidles (*) Date 2012-01-08 22:09
Based upon my own training and personal experience in medical research, in both 2008 and 2011 I gave explicit common-sense advice to both Bob Hyatt and the ICGA

My advice now is the same as in 2008: (1) the ICGA's judicial procedures are so gravely flawed that no-one caught up in them should even attempt to defend themselves, and therefore (2) the ICGA ought to amend its flawed procedures *before* its leaders engage in any further discussions of guilt-versus-innocence.

In answer to a rebuttal from Bob Hyatt, I referred Bob to an article by Alan Dershowitz titled The Criminal Trial Is Not About Justice for the Victim, in which Prof. Dershowitz explains why in all judicial proceedings (both civil and criminal, both formal and informal) the quality of the process has precedence over the quality of the justice.

It is very regrettable to see that instead of the ICGA taking steps to address these procedural flaws, and embrace these principles of justice, the ICGA leadership has evolved further toward gross nontransparency and gravely flawed decision-making.

Perhaps there is no remedy, except to (eventually) elect new ICGA leaders who better appreciate the principles of justice that Prof. Dershowitz talks about. This would be the best outcome that I can foresee.

Therefore, I wish Vasik Rajlich every success and enjoyment in further developing Rybka, and I hope too that a common-sense appreciation of the procedural foundations of justice and the virtues of transparency comes to the ICGA leadership.
Parent - - By bob (Gold) Date 2012-01-08 22:40
We do not see any "procedural flaws".  We chose the best people we could find to investigate this.  The data was presented as clearly and concisely as possible.  It was even presented by three different documents.  Zach's report, Mark's report, and the ICGA technical summary.

The ICGA is a private organization.  If you want to hire outside investigators to go back over what we found, I'd be happy to point you to some, or to put you in touch with a couple of well-known compiler guys that could point you to the right people to do a make-over of the semantic analysis we did.  You will have to pay them, of course.  And that was our conundrum, how to get the best people with no budget to do so.  Programmers already pay to join the ICGA.  The ICGA uses that to pay for the quarterly journal it produces.  There is no giant pool of money to hire independent contractors.  And there is no need.  We have had MANY investigations in the past.  Several programs have been banned, or booted, depending on when they were caught.  The ICGA is free to investigate these complaints however it wants.  I don't see how it could have been done in any other way.  There are not many computer chess people on the planet, particularly those intimately familiar with compiler optimization, x86 assembly language, semantic analysis, and with significant experience in working on chess programs so that the various bits and pieces of an engine make sense and are understood. 

I STILL find it ludicrous to consider Ken Thompson as (a) incompetent;  He developed Unix for heaven's sake.  And the various versions of the belle chess program / machine.  Not to mention C and other things.  (b) biased.  Based on what?  He hasn't competed since the 80's.  (c) a rubber-stamper.  I've known Ken since the 70's, and he is ANYTHING BUT a sign and forget type.  There were plenty of others.  When this investigation started about 5 years ago, I was ALSO quite skeptical of the claims.  I started looking.  And the more I looked, the less I liked what I saw.  A few months of examination and I was pretty well convinced, and the more time passed, the MORE convinced I became.  Nothing exculpatory was found.  Nada.  But the evidence showing copying continued to grow.  And grow.  And finally a large group of programmers filed a formal protest with the ICGA early in 2011 and the formal process started.

So calling the process "flawed" is just a deflection of attention, and is actually grossly inaccurate.  Calling the panel "biased" suggests that 30+ people were ALL biased and had some mystery axe to grind vis a vis Rybka.  Saying the ICGA was "non-transparent" is simply an unfounded statement.  We actively discussed the evidence, even before the ICGA was formally involved.  The report and all evidence developed has been made public.  Part of the discussions were kept private, at the request of those taking part, just like jury deliberations are not made public, only their final verdict.  So how could things be more transparent?  And on the other side, talking about being "opaque" how do you characterize Vas' refusal to even discuss this with the ICGA, even after the report was finished where could write any sort of rebuttal he wanted, without having to interact directly with the panel he if so chose.  I am aware of no "secrets or hidden information" of any kind.  I do not believe that most on the panel would allow that to transpire, in fact.  So I have no idea what you are talking about there.  The ICGA rules have been discussed over a period of 40 years.  Nothing secret about them.  Nothing ambiguous about them.  They are just as general as the general thread of the US constitution, and have been interpreted consistently for years as other programs were investigated and proven guilty as well.  Most here simply are completely unfamiliar with the ICGA, its purpose, the purpose of the rules, or even the concept of "original program" as it applies to the ICGA in general, and tournament rule 2 specifically.  And so we see all the speculation about what the ICGA should be doing, what the rule should allow, what the investigative process should and should not do, all in what would properly be termed an intellectual vacuum regarding computer chess.  SR's report harped on one thing, over and over.  Rule 2.  One does not get to violate a rule just because one disagrees with it.  Every OTHER competitor has been aware of rule 2.  They are aware of what happens to those that violate the rule and get caught.  They are aware of the reasons the rule is used.  So it is pretty funny to see someone like SR, who has never written a chess program (nor any other kind of program I suspect based on the misstatements he has made) tell us what kind of rule we should be using, and why OUR interpretation of basic concepts like ideas vs actual source code are both wrong (in one breath) and then outdated (in his next rail against it).  WE, the competitors, made the rules to make the tournament into what we wanted it to be.  A non-chess-programmer telling us what our rules should be is quite similar to a home framing sub-contractor telling a neurosurgeon which points he should cauterize first, where the incision should be made, etc.  He has no clue.  Neither do the non-competitors regarding rule 2.

If the programmers did not like rule 2, they would either change it or not participate in events that use it.  To date, neither has happened.  The rules have been amended from time to time.  But nothing in the present case warrants this, other than we will likely try to make it easier on future panels by having some sort of source code repository for competing versions, so that the reverse-engineering approach does not need to be used.
Parent - - By John Sidles (*) Date 2012-01-09 00:09
I've served on several juries, and had any judge of any court ever issued instructions similarly confused as Bob's, we jurists would have immediately voted to acquit.  And we would so vote for one over-riding reason (which necessarily trumps all considerations of evidence): foggy procedures cannot foster justice.  Thus I understand Ed Shroder's actions.
Parent - By Nelson Hernandez (Gold) Date 2012-01-09 00:46
Oh dear, Bob will soon rehash his epic adventures as a jury foreman now.
Parent - - By bob (Gold) Date 2012-01-09 05:35
What, EXACTLY is confusing to you.  I have had a judge lecture the jury I served on about the concept of "reasonable doubt".  Most of us already understood the concept.  Would a reasonable person have a reasonable doubt that the events occurred as presented?  Certainly not in this case. 

There was no "foggy procedure" here....
Parent - - By John Sidles (*) Date 2012-01-09 12:05
Bob, by now it is evident to everyone that the fraction of folks who entertain "reasonable doubts" regarding the ICGA's procedures in the Rajlich/Rybka matter, is sufficiently great as to hang any jury.  Isn't that fact plainly evident to all? 

As Peter Skinner, who is the organizer of CCT14, noted earlier this week: "Many of the people that I have contacted on the matter do not disapprove Rybka participating. In fact most welcome it."

Thus a jury of Rajlich/Rybka peers has voted "innocent".  And it is evident that the chess-playing public largely agrees.

Bob, isn't it now clear (to everyone) that the latter two juries are the ones that should prevail? 

Given that there are well-accepted administrative options, aren't the two "innocent" juries the ones that the ICGA should now be heeding?
Parent - - By bob (Gold) Date 2012-01-09 17:38
I do not agree, no.  A very few highly prejudiced people might "claim" that reasonable doubt exists that Vas copied code, but you have to be both highly prejudiced, as well as completely unwilling to look at evidence carefully.  In that case, I suppose he might get a hung jury if he were in a criminal trial.  In a civil case, no chance, because you only need a majority of jurors to agree, not a unanimous decision.
Parent - - By John Sidles (*) Date 2012-01-09 17:58
Bob, I have trouble reconciling the claims of your post with assertions like Riis' "Not even half of the original committee of 34 voted for a guilty verdict." 

It is likely that these issues will be much-discussed in connection with CC14, and it would surprise me very much if the CC14 contestants themselves endorsed the ICGA/Rybka ban with anything like the solid supermajority that your post would suggest.

Bob, do you have any early impressions as to how the CC14 contestants themselves have come to regard the ICGA decision?
Parent - - By bob (Gold) Date 2012-01-10 18:01
I think it is as near perfect agreement as one might come up with.  Some initially (myself included) thought the life-time ban was a bit over the top, but after seeing Vas' refusal to offer anything in defense, and in light of some incredibly ridiculous statements over the past few days "if someone TYPES code, then it is original at the source code level."  Think that would be true if I had a listing and I just typed it in by hand, making a PERFECT copy?  I don't.  Apparently Vas does.  I don't know what more there is to say since he will not discuss the evidence, period...

BTW, your first statement should read "just about half of the original 34 actually PARTICIPATED.  We had no stipulation that they must take part, something we will likely change for the next one.  Friedel, for example, was a member and did not post a thing nor cast a vote.  Does that change anything?  Would it have mattered if we had polled each applicant to determine if they would participate actively and exclude those that said no, so that we had 14 out of 14 voting "we believe Vas copied code"?  Are US elections invalid when sometimes < 50% of the qualified voters turn out?

If you read Riis' article carefully, you can see a lot of Ed/Chris in the background, no doubt.  And every possible distortion has been applied.  Clearly the comment about 14 out of 34 (which is CLEARLY WRONG by the way, since we did point out that there were a few secret ballots cast) is an intentional distortion trying to imply that 20 did NOT agree.  That is NOT the case.  20 simply did NOT participate at all.
Parent - - By Ugh (*****) Date 2012-01-10 18:22
Hyatt wrote: I think it is as near perfect agreement as one might come up with.  Some initially (myself included) thought the life-time ban was a bit over the top, but after seeing Vas' refusal to offer anything in defense, and in light of some incredibly ridiculous statements over the past few days "if someone TYPES code, then it is original at the source code level."  Think that would be true if I had a listing and I just typed it in by hand, making a PERFECT copy?

As usual you are MISQUOTING, aka changing the data to make a cheating point.

Vas actually wrote "if someone types their OWN code, then it is original ...."

which is entirely true. If you type your own code it is obviously original. If you type somebody else's code then it is obviously not original.

By OWN code, I think you can assume he generated the code in his own brain using his own thought processes and ideas and knowledge of programming.

But Hyatt, in desperation to try to win, modifies the quote. Typical.
Parent - By John Sidles (*) Date 2012-01-10 18:41
Yes, Bob's misquote is regrettable.

Similarly, earlier in this thread, Bob claimed for the ICGA panel the mantle of Ken Thompson, and yet now we learn that it's odds-on that any given ICGA panel member never participated and/or voted at all.

What's up with that?  Who participated and who voted?  What rules of evidence were followed?  If rules of evidence were modified mid-course, by what formal procedure were such modifications agreed upon? 

Any justice system that cannot provide good answers to questions like these, manifestly is no kind of justice system at all.

That is the common-sense reason, which has become plainly evident to many chess fans, why Ed Schroder's decision was entirely appropriate.
Up Topic Rybka Support & Discussion / Rybka Discussion / A Big Thanks and a Small Update
1 215 16 17 18 19 20 21 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill