Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / A Big Thanks and a Small Update
1 216 17 18 19 20 21 Previous Next  
Parent - - By bob (Gold) Date 2012-01-10 20:22
I think you can take his wording at face value.  If it is your "own code" you wouldn't even have to type it to call it original, now would you?  If he meant it as you suggest, then I certainly apologize.  But I personally take it as "if I type it, it is original at the source code level."  Not sure why he would add that "source code level" in any case...
Parent - - By Ugh (*****) Date 2012-01-10 20:40
Well, your own code would have to be typed out or generated by another program for the first time, wouldn't it not?

"source code level" differentiates between ideas level and algorithm level. Ie it is original source code but not necessarily an original idea etc.

Just read and engage, parsing as necessary. Shouldn't be too difficult, then you can speak Vas too. Maybe you might get to look in new ways. Not holding my breath though.
Parent - - By bob (Gold) Date 2012-01-10 22:26
Again, semantics.  "I am writing my own book".  What am I doing?  Typing something written by someone else, or writing something I am creating as I go?  "I am copying this article by hand."  What am I doing?  Typing something written by someone else, or creating as I go?  I attached "own code" to mean "that which I am now typing".  That is, "If I am typing it, it is mine, whether I am copying or creating."  If My interpretation was wrong, as I said, I apologize.  But I don't see anything that particularly convinces me one way or the other at the moment.  And in light of the "ALL versions of Rybka were 100% original including the early pre-fruit versions" I think my definition HAS to be the one that is correct, because those early versions were absolutely copies with zero percent chance of error.
Parent - - By John Sidles (*) Date 2012-01-10 22:58
Bob, what precisely are the "semantics" of your claiming for the ICGA the mantle of Ken Thompson (or any other computer science experts)?

Is there any means by which folks can  "semantically" verify assertions of near-unity on that panel?  For example, how can folks determine whether Ken Thompson drafted any substantial portion of that report, or performed any substantial portion of the analysis?

Your practice of mentioning names like "Ken Thompson" --- in a fashion that suggests much, yet asserts little, and even that little cannot be verified --- creates a disastrously confused state of affairs that substantially diminishes the ICGA's reputation.
Parent - - By bob (Gold) Date 2012-01-11 00:25
You are certainly free to contact Ken Thompson.  Frederick Friedel certainly asked him what he thought and apparently did not like his answer.  Why not just ask Ken, so that you get an answer from someone that you won't immediately try to discredit and claim is false? 

Ken looked at the evidence.  He voted that he believed copying was clearly done.  For more, contact him.  He is not hard to find.  I'm not going to post his email here on the web to earn him the attention of all the spambots...  But try doing what I suggested that Ed do, and actually RESEARCH something, rather than asking others to do it, and then not accepting what they say.  Solves a LOT of issues, when you think about it, right???
Parent - - By John Sidles (*) Date 2012-01-11 01:30
Bob, someone (maybe Ken himself?) has scrubbed all mention of Rybka from Ken's own web pages, and Wikipedia, and from every other source that Google can find.  And so (again) your post has suggested much, yet asserted little, and even that little cannot be verified.  

That is why, more and more, Ed Schroder's course of action seems like wisdom.
Parent - By bob (Gold) Date 2012-01-11 04:05
Did Ken EVER mention Rybka in his own web pages?  I can't see why.  The implication that "someone has scrubbed..."  suggests there was something there at some point in time.  Do you have any evidence to suggest that?  I can certainly ask him directly rather than by speculation.  I've not seen Ken discuss this in public, so I won't begin to speculate whether he never did, or whether he did and then removed the comments.  Certainly doesn't sound like him, to me...
Parent - By bob (Gold) Date 2012-01-12 18:07
I don't see the reply I posted to this, so either it is dangling somewhere else or I didn't post it properly.  In any case, If you want to know what Ken thinks, why not simply drop him an email?  He can be found easily enough through google.  I don't see where we claim "the mantle of Ken Thompson."  Ken joined the investigation.  And looked at the evidence.  And concluded copying was done and voted such.  I get inquiries of all sorts through the email system regularly.  I answer every email I receive except for obvious SPAM.  I'd bet Ken would do the same.  IF you ask.  If you just sit and wonder, you will never know.  If I tell you what he did or said, it is hearsay.  So go to the "horse's mouth".
Parent - By nine castles (**) Date 2012-01-12 17:34 Edited 2012-01-12 17:55
The book analogy is not as favorable as you think it is. Imagine:

In an alternate universe, Bob Hyatt, Fabien Letouzey and Vasik Rajlich are all prominent authors of mystery novels.

Bob and Fabien both write popular novels that are influential in the genre.

Vasik reads their books and enjoys them, recognizes their talent and the value of their creative contributions, and a year or so later publishes a novel incorporating some themes influenced by prior works in the genre. He acknowledges his debt to prior works in interviews, but does not give co-author credit since he came up with the plot, the storyline, and wrote most of the characters, although he (again) admits that none of it could have been quite as good had he not consumed and enjoyed work done before him - as any fan of the genre would have.

Bob and Fabien flip out about "originality" as if each mystery novel had to have been written in a vacuum and only virgins to the genre can write an original work, not anyone who has been living and breathing it, as any passionate author would have before he wrote anything original in what is now an established genre.

Get it?

Bob, when you were involved with Cray Blitz, it was much easier to be completely original with both code and ideas. You had to be, because nobody had done much of anything before you. You are like Arthur Conan Doyle in my analogy, and Vas is Agatha Christie: everyone writing crime or mystery novels once read Sherlock Holmes, and everyone who wrote a great novel after Doyle also contributed a great deal of their own creative work. Nobody doing work like this could not be influenced by your work, unless you expect that professional chess programmers should never read crafty. So when you talk about literal, exact code duplication, that is a charge everyone should take seriously -- akin to literal, direct line-by-line plagiarism in a mystery novel. But when you talk about duplication of "ideas", nobody can or should take that seriously -- that is like reusing the "least likely suspect" plot twist. Everyone does it - and while I don't agree with everything Riis wrote, I think he did persuasively argue that the same thing is now happening to Rybka.

Edit: See my post below for more about this. I think the comparison of Rybka->IPPOLIT lineage vs Fruit->Rybka lineage may be a good way of teasing out our opinions.
Parent - - By Rebel (****) Date 2012-01-12 23:27

> Again, semantics.


That will be the last ICGA myth to be debunked, soon.

Starring Crafty 22.2

A preview:

Much of the ICGA semantics is based on the MG/EG evaluation technique of Fruit, copied by you Bob Hyatt starting in 2008 and introduced in version 22.2, a lot of thanks for doing so.

Engines using the MG/EG evaluation technique automatically become semantic equal on many places in EVAL.

Two Crafty examples: http://www.top-5000.nl/temp.htm
Parent - By Ugh (*****) Date 2012-01-12 23:30
Difficult to know whether to laugh or cry. What a mess.
Parent - - By bob (Gold) Date 2012-01-13 02:01 Edited 2012-01-13 02:06
Let's examine your latest lie/distortion/nonsense.

First, Crafty's trapped bishop code pre-interpolation scoring, which was added in 22.2:

    if (square == sqflip[side][A7] && SetMask(sqflip[side][B6]) & Pawns(enemy))
      score -= bishop_trapped;
    else if (square == sqflip[side][B8] &&
        SetMask(sqflip[side][C7]) & Pawns(enemy))
      score -= bishop_trapped;
    else if (square == sqflip[side][H7] &&
        SetMask(sqflip[side][G6]) & Pawns(enemy))
      score -= bishop_trapped;
    else if (square == sqflip[side][G8] &&
        SetMask(sqflip[side][F7]) & Pawns(enemy))
      score -= bishop_trapped;

So, what does that do?  Checks for a repeated pattern 4 times.  Bishop at a7, pawn at b6, bishop at b8, pawn at c7.  Bishop at h7, pawn at g6.  Bishop at g8, pawn at f7.  This gets done TWICE, once for each color, the sqflip just changes a7 to a2 when it is black to move...

In 22.2, we turned that into this:
    if (square == sqflip[side][A7] && SetMask(sqflip[side][B6]) & Pawns(enemy)) {
      score_eg -= bishop_trapped;
      score_mg -= bishop_trapped;
    } else if (square == sqflip[side][B8] &&
        SetMask(sqflip[side][C7]) & Pawns(enemy)) {
      score_eg -= bishop_trapped;
      score_mg -= bishop_trapped;
    } else if (square == sqflip[side][H7] &&
        SetMask(sqflip[side][G6]) & Pawns(enemy)) {
      score_eg -= bishop_trapped;
      score_mg -= bishop_trapped;
    } else if (square == sqflip[side][G8] &&
        SetMask(sqflip[side][F7]) & Pawns(enemy)) {
      score_eg -= bishop_trapped;
      score_mg -= bishop_trapped;
    }

SAME CODE, except that every score -= reduction is duplicated, once for MG, once for EG.  Is that equivalent to Fruit 2.1?  Let's take a peak:

   if ((board->square[A7] == WB && board->square[B6] == BP)
    || (board->square[B8] == WB && board->square[C7] == BP)) {
      *opening -= TrappedBishop;
      *endgame -= TrappedBishop;
   }

   if ((board->square[H7] == WB && board->square[G6] == BP)
    || (board->square[G8] == WB && board->square[F7] == BP)) {
      *opening -= TrappedBishop;
      *endgame -= TrappedBishop;
   }

   if ((board->square[A2] == BB && board->square[B3] == WP)
    || (board->square[B1] == BB && board->square[C2] == WP)) {
      *opening += TrappedBishop;
      *endgame += TrappedBishop;
   }

   if ((board->square[H2] == BB && board->square[G3] == WP)
    || (board->square[G1] == BB && board->square[F2] == WP)) {
      *opening += TrappedBishop;
      *endgame += TrappedBishop;
   }

Similar?  Yes.  Although the code is not formulated the same way.  So those are NOT "semantically equivalent"  Not even close.  First, notice my code only addresses the 7th rank from the side to move perspective, while Fruit explicitly calls out all 8 of the squares.  Not just 4 as in Crafty.  So that is not the same.  Let's advance a bit with Crafty.  To the current versions, 23.x...  say 23.4 as the most recent:

    t = bishop_outpost[side][square];
    if (t) {
      if (t > 0) {
        if (!(mask_no_pattacks[enemy][square] & Pawns(enemy))) {
          if (pawn_attacks[enemy][square] & Pawns(side)) {
            t += t / 2;
            if (!Knights(enemy) && !(Color(square) & Bishops(enemy)))
              t += bishop_outpost[side][square];
          }
          score_eg += t;
          score_mg += t;
        }
      }
/*
************************************************************
*                                                          *
*   Check to see if the bishop is trapped at a7 or h7 with *
*   a pawn at b6 or g6 that has trapped the bishop.        *
*                                                          *
************************************************************
*/
      else {
        if (square == sqflip[side][A7]) {
          if (SetMask(sqflip[side][B6]) & Pawns(enemy)) {
            score_eg -= bishop_trapped;
            score_mg -= bishop_trapped;
          }
        } else if (SetMask(sqflip[side][G6]) & Pawns(enemy)) {
          score_eg -= bishop_trapped;
          score_mg -= bishop_trapped;
        }
      }
    }

You have to include ALL of that code, because the initial test to see if this is an "interesting square" uses that early "outpost test."  If t (the outpost flag) is > 0, then we score this as an outpost bishop.  If it is zero, we don't do anything with this block.  Else if it is < 0, it is a "anti-outpost" which is a "ugly" square where it can be trapped, namely a2/h2/a7/h7. 

Fruit STILL tests for g8, b8.  See those "trapped squares" in my code?  Nope.  Wait a minute.  Fruit detects trapped bishops on the 6th rank as well.  Here is the REST of the fruit code you dishonestly snipped:

static void eval_pattern(const board_t * board, int * opening, int * endgame) {

   ASSERT(board!=NULL);
   ASSERT(opening!=NULL);
   ASSERT(endgame!=NULL);

   // trapped bishop (7th rank)

   if ((board->square[A7] == WB && board->square[B6] == BP)
    || (board->square[B8] == WB && board->square[C7] == BP)) {
      *opening -= TrappedBishop;
      *endgame -= TrappedBishop;
   }

   if ((board->square[H7] == WB && board->square[G6] == BP)
    || (board->square[G8] == WB && board->square[F7] == BP)) {
      *opening -= TrappedBishop;
      *endgame -= TrappedBishop;
   }

   if ((board->square[A2] == BB && board->square[B3] == WP)
    || (board->square[B1] == BB && board->square[C2] == WP)) {
      *opening += TrappedBishop;
      *endgame += TrappedBishop;
   }

   if ((board->square[H2] == BB && board->square[G3] == WP)
    || (board->square[G1] == BB && board->square[F2] == WP)) {
      *opening += TrappedBishop;
      *endgame += TrappedBishop;
   }

   // trapped bishop (6th rank)

   if (board->square[A6] == WB && board->square[B5] == BP) {
      *opening -= TrappedBishop / 2;
      *endgame -= TrappedBishop / 2;
   }

   if (board->square[H6] == WB && board->square[G5] == BP) {
      *opening -= TrappedBishop / 2;
      *endgame -= TrappedBishop / 2;
   }

   if (board->square[A3] == BB && board->square[B4] == WP) {
      *opening += TrappedBishop / 2;
      *endgame += TrappedBishop / 2;
   }

   if (board->square[H3] == BB && board->square[G4] == WP) {
      *opening += TrappedBishop / 2;
      *endgame += TrappedBishop / 2;
   }

   // blocked bishop

So, you dishonestly snipped the bishop trapped code in Fruit, leaving just 6 lines you want to claim are equivalent.  You used a version of my trapped bishop code which determines where the bishop is trapped or not, from a version that existed PRIOR to Fruit.  You lie about the "match" not even looking at code from the past 3 years of Crafty, rather looking at one that was right after the fruit interpolation approach was used.  The code doesn't match.  The basic operation is different.  And Crafty hasn't "done it like that since the end of 2008.  Over 3 years ago.  Easy to verify by using any source you want to see when 19.2 came along, which was the FIRST version to test b8/g8... 

You are beyond a doubt the most devious and dishonest person I have met in this regard.  You lie at every opportunity.  You post a "comparison on your web page" that shows source code that pre-dates Fruit.  here is an example:

Here is the code from 19.0/19.1:

  if (WhiteBishops) {
    if (WhiteBishops&mask_A7H7) {
      if (WhiteBishops&SetMask(A7) && SetMask(B6)&BlackPawns)
        score-=BISHOP_TRAPPED;
      else if (WhiteBishops&SetMask(H7) && SetMask(G6)&BlackPawns)
        score-=BISHOP_TRAPPED;
    }
  }

Here is the code from 19.2:

if (WhiteBishops) {
    if (WhiteBishops&mask_WBT) {
      if (WhiteBishops&SetMask(A7) && SetMask(B6)&BlackPawns)
        score-=BISHOP_TRAPPED;
      else if (WhiteBishops&SetMask(B8) && SetMask(C7)&BlackPawns)
        score-=BISHOP_TRAPPED;
      else if (WhiteBishops&SetMask(H7) && SetMask(G6)&BlackPawns)
        score-=BISHOP_TRAPPED;
      else if (WhiteBishops&SetMask(G8) && SetMask(F7)&BlackPawns)
        score-=BISHOP_TRAPPED;
    }
  }

What changed?  I added the b8/g8 squares.  When was that code released?  19.2 was released LONG before Fruit, which came along in June of 2005.   I can not say specifically without a bit of effort that is pointless, but 18.x was released in 2003, so that certainly pre-dates Fruit.  Ergo, hard to copy something that didn't even exist, eh?  So, if anything was copied by anybody, perhaps Fabien liked my idea of b8/g8 and used that.  Unfortunately for YOUR case, my bishop code looked completely different by 3 years ago, starting with 23.0 which was released in March 2009.  So you show OLD code, claim that it shows I copied something, yet you stupidly choose a version that was released way prior to fruit.  And you stupidly choose a version that looks nothing like any crafty of the past 3 years.  And you try to make a point by purely dishonest means?

"dishonest liar" is the ONLY term that fits.  I'll leave it at that, since I think things are quite clear.  Your web page contains nothing but either distortions or else outright lies.  I am sure you are proud of it...
Parent - - By Rebel (****) Date 2012-01-13 08:19
Read again: Much of the ICGA semantics equivalence (SE) is based on the MG/EG evaluation technique of Fruit.

In principle it has nothing to do with Crafty, xxx, yyy.

Crafty 22.2 is used as an example to show how SE slips into EVAL (all over) when using the MG/EG evaluation technique.

Don't you see the humor of this all ?

PST issue - Fruit Bishop PST introduced in Crafty 22.2

SE issue - Fruit MG/EG semantics introduced in Crafty 22.2

Very funny, you giving Vas another unburden piece of evidence.
Parent - - By bob (Gold) Date 2012-01-13 12:52
NONE of the ICGA evidence is "based on the interpolated scoring".  ALL of the ICGA evidence is based on semantic equivalence of ALL the code in a block of the evaluation.  Not just that Rybka uses two values.  That is yet ANOTHER lie.  Keep piling them up.
Parent - - By Rebel (****) Date 2012-01-13 14:54
NONE of the ICGA evidence is "based on the interpolated scoring"

Thank you.

We will see when we remove the MG/EG concept and what stays then, the visual compare suddenly will look less SE.
Parent - By bob (Gold) Date 2012-01-14 04:08
Not if you remove it from BOTH...

I still don't believe you have a clue about semantic equivalence.  I really am beginning to believe you don't have a clue about ANYTHING other than dishonesty and distortion...  There you are a practicing master.
Parent - - By bob (Gold) Date 2012-01-11 17:31
Finally had some time to do a bit of research to refute the above explanation for comments vas made about what makes for "original code."  Here is a DIRECT QUOTE from the rybka forum, that seems both crystal clear AND damning:

    > Regarding Strelka/IPPOLIT: as the author(s) seem to have typed their own source code (or code to generate this), how are they not "original" under your definition?

I doubt that all of that code was typed by hand. If it was, then sure, it's "original at the source code level".

So it seems MY interpretation of his comments was dead accurate.  "you type it, it is original."

what a complete crock.

Now, let's see how you try to spin that.  (Here is a link to that specific post: http://rybkaforum.net/cgi-bin/rybkaforum/topic_show.pl?pid=390718#pid390718 )  This ought to be good for a few laughs.
Parent - - By nine castles (**) Date 2012-01-12 17:42 Edited 2012-01-12 17:58
I think this might be a good way of getting at it. Alluding to my mystery novel analogy above, there are a couple things an author could do that would make the work "unoriginal" or "derivative", instead of merely "influenced by":

1. Vas reads Bob's novel and notices that at the end, it is revealed that the old granny witness to the murder (the last person anyone would suspect of being the killer) was actually the killer. This is very surprising to the reader and offers a dramatic punch. He adds a similar trope to his book.

Not plagiarism, still an original work, since the entire plot and character design was his own, as well as all the details in between.

2. Vas reads Bob's novel and notices a few turns of phrase that are especially well written. He "borrows" these phrases and has some of his characters say them too in what is otherwise an original storyline.

Definitely unethical, and almost certainly copyright infringement, but not nearly as bad as #3:

3. Vas reads Bob's novel and thinks it's very good and treats it like a "foundation". He begins with the complete manuscript and starts by renaming some of the characters. He tightens up the storyline, adds some of his own ideas for sub-plots, rewrites and improves a bunch of the dialogue and makes other substantial changes that are his own creative ideas. By the time he's finished, he has a much better novel than the one Bob wrote, but many pieces of Bob's novel are still intact and in a few critical places where Vas didn't change anything. Wrongly thinking that his creative contributions make it his own work, or perhaps naively thinking that due to the extent of the reworking he's done that nobody will notice, he publishes the novel under his own name.

This is more or less the definition of a "derivative work" and under US copyright law, Bob owns the copyright to Vas' book.

--
I hope this illustrates my approach to this, and I hope it helps, because sometimes it's unclear what we are talking about. Do you mean to say that all three are unethical, when it comes to computer chess programming? Do you think Vas is guilty of all three? A lot of the time, it sounds like you are talking only about #1, when we talk about "ideas", and this is definitely the weakest case you could make. If you think you can prove #2 or #3, that's really all you should be talking about, imo.

So on the IPPOLIT case, it sounds like Vas thinks IPPOLIT along the lines of #3: the authors started with Rybka and made a few changes, but their creative contributions were (in his view) minimal compared to what they started with and it's still basically his "story". Based on what he's said, he seems to think that the influence of Fruit and Crafty on Rybka is more like #1, where he borrowed thematic tropes and good ideas and worked them into an otherwise creatively original narrative. Personally, I don't know enough about this to make a judgment either way, but hopefully this will make it easier for everyone to communicate about these issues.
Parent - - By bob (Gold) Date 2012-01-12 18:03
I believe that what Vas did is closer to your last example.  He simply copied the fruit eval, then modified it to use bitboards (changed the names of characters and locations, but the plot remains identical).  And over time he added more.  Have you ever read Clive Cussler?  Take one of his books.  He always starts, before the real story is told, way back in history somewhere.  And plants some "seeds" (Amelia Earhart's plane in his book Sahara, for example) that appear to be "fluff" until some key point in the story where suddenly that detail plays an important role.

Suppose he left that "preamble" off and just started in on the main story line.  The books would still work just fine, because at the right moment, the "discovery" could still be explained, without the early introduction.  Now suppose someone copies his book, and adds the long introduction explaining what happened and why, to set the stage for the discover (as Cussler actually does).  Is that new?  Nope.  Better?  Probably. 

In short, Vas copied fruit code, translated it to a new board representation, which did NOT add any new and original "stuff".  Then over a period of several years he DID add to that original story.  But at least through 2.3.2a, if you read his "story" you continually think "looks just like the story Fruit I read last year, but with a few changes here and there."  Better story?  Of course.  Original story?  not even close.
Parent - - By nine castles (**) Date 2012-01-12 18:18
All right, fair enough. I think you are at minimum charging something that is worthy of serious sanctioning and it's good to be able to focus on that point. Maybe some day I'll catch up on the technical knowledge to understand the ICGA report and evaluate the evidence for myself. It's useful to know that you think you have clear evidence that while Rybka did play vastly stronger chess than it's competitors, the added creative contributions from Vas were in whole or in substantial part in successive modifications on top of an existing codebase and not integration of good ideas (#1 above) as Vas says it is.

I think many of Rybka's fans feel that they cannot acknowledge plagiarism (if they might otherwise be persuaded) without discounting Vas' contributions and advancements to computer chess, and this has been a source of cognitive dissonance for me as well. You can see that in the Riis article: I think he takes such great pains to extol Rajlich's programming genius and distance from his peers because when you hear "plagiarist" you typically think of someone who did nothing but sign his name to someone else's work. Scenario #3 above shows how someone can be a creative genius and put out a brilliant work while at the same time being a plagiarist. That may be the scenario we have. Thanks for running with me on this.
Parent - By bob (Gold) Date 2012-01-12 20:13
I don't think there is any difference between "successive modifications on top of an existing code base" and "integration of good ideas" so long as you qualify that as "integration of good ideas on top of an existing code base."  Not one person has denied that he made major improvements.  But copying someone else's code is wrong.  Claiming it as your 100% original work is wrong.  Entering it in tournaments that expressly require original code is wrong.
Parent - By Venator (Silver) Date 2012-01-14 09:27
95% of Crafty's elo can be contributed to others. What a crock.
Parent - - By AWRIST (****) Date 2012-01-09 01:41
Bob wrote 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

I am totally siding John who also has looked through your bias, Bob. Your naivety is incredible regarding your smartness and also experience in computerchess programming. The problem is that you have no sense for the relevance of judicial questions.

Here in the quote you miss the old argument from my side, that without examining also other commercial players just as a controlgroup measurement, and I am absolutely sure that you know this saort of requirement in stats and then also justice. So I ask you and all readers, how can it be that such an expert like Bob behaves like such a mega fool, who advocated the outsingling of Vas and his Rybka while ignoring the reality of all other programs.

I know well Bob's answer, he repeated it often enough. Sure, he said, if someone would lead him towards a minimum of existing evidence of another case he, Bob, would research it. But Heavens. A scientist like Bob must have understood that he was forced to control his research before such outside observers like me would have the time to show Bob in such a basic wrong - without such controlling of his own methods in the examining of Rybka.

I agree with John. And I suppose this is why no court case will be happening against Vas. Because there is absolutely no case at hand. What we have is the private one-sided conviction of Bob Hyatt and his overflooding propaganda. But in court the attorneys must examine both sides of the medal not just the biased side of a Bob Hyatt. Bob, this is a shame for all scientists. In case of Deep Blue 2 it was a similar case of violation of scientifical ethics. What will be case number three?
Parent - - By bob (Gold) Date 2012-01-09 05:31
What "judicial question" is there here?  "did he violate ICGA rule 2 or not?"  The answer is, quite simply, a resounding "yes".

I am NOT going to re-visit the debate about investigating every other program that enters.  There are hundreds of them.  It is beyond anybody's realistic ability to do because of the time required...

If the FSF pursues this, they will certainly prevail, because the copying evidence is large in scope, and accurate with regard to quality.
Parent - - By Banned for Life (Gold) Date 2012-01-09 05:54
What "judicial question" is there here?  "did he violate ICGA rule 2 or not?"  The answer is, quite simply, a resounding who cares?

If the FSF pursues this, they will certainly prevail, because the copying evidence is large in scope, and accurate with regard to quality.

The FSF is pursuing this, remember? And guess what, they won't prevail because none of your copying evidence is worth the paper it's printed on. But feel free to keep us posted on their progress! :yell:
Parent - - By bob (Gold) Date 2012-01-09 05:57
Obviously a bunch of tournament participants "care".  And just as obviously that group could give a squat whether or not YOU care...  The ICGA tournament is, and always has been, organized for the PROGRAMMERS.  Not a clown such as yourself that doesn't even understand the basic idea of competition using one's own work and such.
Parent - - By Banned for Life (Gold) Date 2012-01-09 06:21
The ICGA tournament has become a competition amongst second rate engine developers. The 2011 event went unnoticed by the rest of us until two weeks after its conclusion. Chessbase dropped its sponsorship and refused to publicize the event or even report on the results. With a little luck, we won't hear about next years competition at all. Good riddance!
Parent - - By sockmonkey (***) Date 2012-01-09 06:32
Wow, and I thought that this was a thread for Vas to answer questions. Did we really need another thread for his bloviating proxies?
Parent - By Banned for Life (Gold) Date 2012-01-09 06:48
Sure, why restrict a good thing? :smile:
Parent - - By bob (Gold) Date 2012-01-10 18:13
Did you REALLY expect any answers?  I guess I did too, at some level.  But it was NOT to be, obviously.  Unless you were only interested in when Rybka 5 might show up.
Parent - - By Ugh (*****) Date 2012-01-10 18:27
I think you got plenty of answers.

Just like David Levy got answers such as "all that code isn't in Rybka, it's just somebody's imagination". Which is a) true and b) sufficient. Why say more?
Parent - - By bob (Gold) Date 2012-01-10 19:52
Yeah, I got answers like "discussing 1.6.1 is not a topic for today..."  It won't EVER be a topic for discussion, because it is utterly hopeless to try to defend that mess...
Parent - By AWRIST (****) Date 2012-01-10 20:18
no, he answered you that this were his private engines that shouldnt be of interest for the crowd. Perhaps it helps you if I tell you that Vas didnt make money with the crafty turkey stuff in it. After he began without you, he went right on milky way. Now you are alone with Deville, but with over 400 Elo less. That's creativity, not copying. Copying whom?? Some weakies?

Why cant you apologize to Vass for all the wrong you've done to him in those 6+ years?

Even Charlie Sheen made a break in his TV-Show! Why not Bob Hyatt too? :cool:
Parent - By Venator (Silver) Date 2012-01-11 10:57 Edited 2012-01-11 10:59
Using your 'probably it went this way' scientific way of concluding, it is now pretty obvious and probable that this is the way the whole charade went:

You clearly stated several times that 'many pages of IDENTICAL Fruit-Rybka code' were in the ICGA report, side by side comparisons. Then, to your horror, you found out that this was a terrible mistake and that in fact the code in the ICGA report was just some imaginary code to make it look like Fruit.

This VERY nicely explains your endless postings here in this forum. Because if the evidence was clear cut and crystal clear, leaving no doubt at all, you would not be here trying to defend it and confuse matters.

So it is very PROBABLE that you are here in this forum to cover up your mistakes, adding lie upon lie, trying to rephrase your statements, hoping that if you would repeat them a thousand times, people would believe you and leave the matter alone.

In reality you dug the hole you created for yourself deeper and deeper, with the ironical high light the copied Fruit bishop PST in Crafty.
Parent - - By OleM (**) Date 2012-01-13 22:05
That depends on what you mean by defend. IF you mean trying to prove that R1.6.1 wasn't a copy of Crafty, then you may very well be right. I wouldn't be qualified to judge on that. But if you by defend mean giving a reasonable explanation for using Crafty as a base for experimentation, then perhaps. Although I don't like how you call people idiots or whatnot, I do believe you would recognize your own engine. In other words I believe you when it comes to wether Rybka 1.6.1 was a clone of Crafty or not. But when this was discovered you should have had the integrity to withdraw from the investigation, because you were no longer unbiased. No one with personal interests like that can be. I think you would have gained a lot more respect, and the ICGA a lot more credibility. You just never puts a victim in the jury.
Parent - - By bob (Gold) Date 2012-01-14 03:59
I can't imagine any "reasonable explanation."

1.  Events that use ICGA rule 2 would CLEARLY exclude such a program.

2.  Crafty's license agreement explicitly forbids using a program containing parts of Crafty in ANY computer chess competition except for one where you hold the entire thing for yourself, all the games (all opponents) are played on your computer, etc.  Entering other events like CCT / ChessWar is clearly not acceptable.

3.  Simply copyright logic would lead ANYONE to the conclusion that this program (rybka 1.4 - 1.6.1) simply goes WAY too far...

Why did I need to be "unbiased"?  Do you think the cops are "unbiased" when they investigate a case.  Do you think a prosecutor (or defense attorney) is "unbiased" when they prepare / present their cases?  The unbiased component comes from the Jury.  In this case, the Jury was the ICGA board.  NOT the ICGA panel.  Your last statement is telling, because it seems you think the panel WAS the jury.  Didn't happen.  The board made their deliberations and decisions in private, without any input from us beyond the report everyone has seen and the associated evidence.
Parent - By Venator (Silver) Date 2012-01-14 13:33
The unbiased component comes from the Jury.  In this case, the Jury was the ICGA board.

Any jury member in computer chess matters who has connections to you, is biased in the first place. That's one fact you clearly understand (but will deny immediately) and no matter how hard you will try to prove that the jury was unbiased: no, it was not.

Therefore any court would send this procedure into the dustbin immediately.
Parent - - By OleM (**) Date 2012-01-14 13:35
There is a reason why the police have an internal affairs group to investiga cases where cops are involved. And you would never be selected as a prosecutor in court if you were the victim of the case. You know this just as well as I do. There are of course degrees of bias, that is just human nature. I am not saying you shouldn't be allowed to speak your mind, or provide evidence (e.g code) or advise. I think we can agree that those who voted can be compared to the jury of a court case. In this case the problem is that the same people are also the prosecutors. Now don't tell me you don't see this. You can of course claim this isn't a relevant comparison, but you yourself keeps referring to how legal justice works, so that would be a bit hypocritical. The ICGA should have made a clear distinction of who should have done what in the process. Meaning that some people should be investigating and presenting the evidence, and another group should vote. Then I don't think anyone could have attacked the process in itself, and the attention could be given to the evidence instead. As it is both parts, both Vas AND the ICGA HAS lost some credibility, even if you deny that, to the general public.

As I said I believe you when it comes to R1.4-R1.6.1 being clone/derivative/copy (use whatever word is most precise) of Crafty as no one would know your code better than yourself. But as someone else has pointed out (Even Zach Wegner did this), that is not relevant to the case (if Vas broke rule 2 in ICGA competitions). Whatever it tells about what kind of person Vas is, is another matter completely.
Parent - - By bob (Gold) Date 2012-01-14 14:53
I will say it again.  The police/prosecution do not "recuse themselves" because they have been past victims of a similar crime.  Might happen if they "know" the person personally, as that could bias them.  But I have never met Vas, and likely never will.  If you feel the process was flawed, that is your opinion, and your right to have that opinion.  When we STARTED the investigation, there was NO KNOWLEDGE that he had copied Crafty.  In fact, we did not discover that until the investigation was nearly completed, maybe March or so of last year (I could likely find exactly when.  Actually I just looked at my email and March appears to be the correct time-frame.)  We were already well into the process of writing our report summarizing the evidence.  Until March, I had no idea this had happened.  Would have been quite difficult to recuse myself when the panel was formed, since this was not known until we were nearly done.

As you see, statements made here are not always the most factual you might find.  It is not as if I knew about this in 2009 or something.

As Paul Harvey might say, were he still alive, "Now you know the REST of the story..."
Parent - By OleM (**) Date 2012-01-14 21:32
The process wasn't perfect, that's for sure. I think in any case like this you should evaluate every aspect afterwards to see if things could have been done better to make sure that processes are transparent and fair, and secure as little bias as possible. And that is no matter the outcome. To have 34 of Vas' best friends do this investigation would have been just as bad of course.
I can accept the fact that the Crafty/R1.6.1 discovery was made so late in the investiagation that it was really no practical chance for you to step back.
Parent - By Venator (Silver) Date 2012-01-11 09:57
The ICGA tournament is, and always has been, organized for the PROGRAMMERS.

Currently second rated programmers, who prefer to interpret rules to their own advantage.
Parent - - By AWRIST (****) Date 2012-01-09 08:42
There he goes. David Levy lied. So Bob Hyatt cannot abstain, he must also lie. He writes:

I am NOT going to re-visit the debate about investigating every other program that enters.  There are hundreds of them.  It is beyond anybody's realistic ability to do because of the time required...

I was talking about control groups for scientific purposes. And the _only_ group that counts is that of the very few of the direct _commercial_ competition. Not hundreds of amateurs.
Parent - - By bob (Gold) Date 2012-01-10 18:06
So JUST the commercial programs can be derivatives?  Where did you come up with that idiocy from?

Until Rybka, NO commercial derivatives have been detected.  Several non-commercial ones have, however.  Make that dozens.
Parent - By AWRIST (****) Date 2012-01-10 20:22
Bob, I beg you. Hint: methodology. That choice comes first, then we can talk about results. I didnt offer any easyytalks about who could be poisoned of the other commercials. Kommodo comes to mind where Don nhas always admitted with Larry, that they profited of the Rybka publication by criminals. (cough)
Parent - - By AWRIST (****) Date 2012-01-13 12:04
Bob wrote: I STILL find it ludicrous to consider Ken Thompson as (a) incompetent

I still find it ludicrous to consider Einstein as incompetent, but he still advised in favor of the bomb.

Ken Thompson, or nobody has seen the source code of Rybka. Unless a brilliant computer scientist isnt a clairvoyant, he has nothing to say that is valid about such original code.
Somehow one must protect these names against your assaults, Bob. Get real.
Parent - By Chess_Rambo (***) Date 2012-01-13 14:41

> I still find it ludicrous to consider Einstein as incompetent, but he still advised in favor of the bomb.


And did it work? :razz:

> Somehow one must protect these names against your assaults, Bob. Get real.


Somehow one must protect these names against your assaults, Rolf. Get real. :razz:
Parent - - By bob (Gold) Date 2012-01-14 04:19
We have ALL seen the asm that comes from Rybka.  That is MORE than good enough for this comparison.
Parent - - By AWRIST (****) Date 2012-01-14 12:32
I tend to disagree. But please you may also take a look into this extremely interesting forum of talkchess where it was actually stated (please use the search funktion) that by no means R1 looked like Fruit.
What's up?
Parent - By bob (Gold) Date 2012-01-14 15:26
Incompetence...
Parent - By AWRIST (****) Date 2012-02-05 10:55
Please dont mention this Dershowitz, he's pro torture, so, he's more like a criminal himself. In Germany we have a professor Wolfsohn at the university of the Military and he's also pro torture. You get the drift. Evil are those who think that normally one would expect this shouldnt happen at all.
Of course we agree on Rybka against ICGA under David Levy.
Up Topic Rybka Support & Discussion / Rybka Discussion / A Big Thanks and a Small Update
1 216 17 18 19 20 21 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill