FIDE Rules on ICGA - Rybka controversy

[Harvey Williamson]

This in response to previous statements you made here (on this forum) Loop undoubtly being a Fruit clone.

I cannot find any such statement by myself. I can only find you using that phrase in your email to Jaap.

Let's see what we have -

1. On this forum - I made a cursory glance at Loop 2007 (the 64-bit version). It uses exactly the same PST as Fruit 2.1. The pawn eval uses the same scores. It shifts the mobility count as with Fruit (by 4,6,7,13) before multiplying, and then the arithmetic is the same (4 for N, 5 for B). I'm not that interested in going further.

2. From the Panel discussion - though with now Loop 2007 looming (which seems even more obviously a Fruit/Toga clone),

Then following the words of Jaap we read -

1. I spoke in Yokohama with mark watkins and gave him the opportunity to tell his story to the assembled chess porgrammers. Mark did so and informed us on the state of affairs. from his point of view it was not as clear as you stated but there were pointers.

2. i admit i was interested in true facts and i spoke to mark watkins. amir ban went into discussion with mark, and a few other people from the audience too. the oral formulation of a kind of conclusion was that there is not a convincing proof, moreover there was a discussion to what extent .... . in brief, it was not clear whether the monks were equal, they might be similar, but the weighing of the proportional part was too difficult for the "audience judges".

----------

So which one of you is telling the truth?

Was Jaap truthful when he stated, "from his [Mark Watkins] point of view it was not as clear as you stated but there were pointers." It obviously conflicts your earlier statements about LOOP, calling it a clone.

Here is the deal, I am more than willing to change the page that annoys you with the answer you give, but you can't be both right.

Perhaps Jaap misunderstood part of the discussion but my recollection of it is as Mark has posted.

[BB+]

It was bad enough in this thread when you were previously simply getting things wrong (eg, "fundamental right" to an appeal, Rajlich retracting his statement that Strelka contained Rybka code, old Glaurung containing Fruit-like method with infinite/ponder searches), but deliberately distorting people's words is even worse.

Anyway, on what may very well be last response to you, in this thread or otherwise...

I made a cursory glance at Loop 2007 (the 64-bit version). It uses exactly the same PST as Fruit 2.1. The pawn eval uses the same scores. It shifts the mobility count as with Fruit (by 4,6,7,13) before multiplying, and then the arithmetic is the same (4 for N, 5 for B). I'm not that interested in going further.

As should have been clear from the context, I was not interesting in going further [at that time] because what I had found by the first cursory inspection already made it clear that there was a lot of Fruit/LOOP overlap, and this sufficed for the time being. Later in that thread (which I will take the time to cite), I gave a (lightly) annotated disassembly of the relevant LOOP eval function.

2. From the Panel discussion - though with now Loop 2007 looming (which seems even more obviously a Fruit/Toga clone),

As you know, the Panel discussions are private. In any case, "More" does not mean "an undoubtedly 100% clone" to quote your words to Jaap. I am in full agreement (with myself) that the LOOP/Fruit overlap is more than the Rybka/Fruit overlap (which is the context of the above quotation). But I am also in sharp disagreement with your phrasing of LOOP being "an undoubtedly 100% clone", that for some reason you have attributed to me.

So which one of you is telling the truth?
Was Jaap truthful when he stated, "from his [Mark Watkins] point of view it was not as clear as you stated but there were pointers." obviously conflicts your earlier statements about LOOP, calling it a clone.

Yes, Jaap was truthful -- it is not my point of view that the Fruit/LOOP situation is "as clear as you stated", meaning "an undoubtedly 100% clone". Furthermore, I am telling the truth and there is no "obvious conflict", as I explain above. I think the only one who is not truthful is you, when you introduced this "an undoubtedly 100% clone" phrasing, and somehow then tried to palm it off as being from my mouth or writings.

To conclude:

• MarkW implied/said (in 2011 forum posts and private Panel discussions) that LOOP is more obviously a Fruit "clone" than Rybka,
  though made little if any quantification of this.
• Mr. Schröder wrote to Jaap and indicated that (contrary to Rybka) LOOP was "an undoubtedly 100% clone".
• Jaap correctly replied that MarkW (in his 2013 impromptu Yokohama presentation) did not think it was "as clear as you stated [ie 100% clone]".
• Jaap indicated that in my talk "there were pointers" [to LOOP sharing Fruit origins], namely that I sketched the evidence that was in the 2011 Forum post cited above.
• Jaap further (correctly) noted "moreover there was a discussion [with Amir Ban in particular] to what extent" LOOP was a clone, and any conclusion herein was not determined on-the-spot by the "audience judges" (after a 15-minute briefing), but would rather await a fuller analysis.
• Mr. Schröder, as Rajlich's representative, somehow confused all this, and asserted that MarkW had "changed his mind" regarding LOOP, writing to Levy: Mark Watkins remarks during the programmer meeting in Japan (2013) when he said, "The LOOP case is not so clear", he considered his previous statements as wrong and changed his mind.
• Mr. Schröder moreover (in this thread) erroneously proposes that Jaap's comments were in response to previous statements [MarkW] made here (on this forum) Loop undoubtly being a Fruit clone -- whereas the only use of this "undoubtedly" terminology was from Schröder himself.

Perhaps Jaap misunderstood part of the discussion but my recollection of it is as Mark has posted.

I think the misunderstanding is at least 90% that of Mr. Schröder.

[hyatt]

This in response to previous statements you made here (on this forum) Loop undoubtly being a Fruit clone.

I cannot find any such statement by myself. I can only find you using that phrase in your email to Jaap.

Let's see what we have -

1. On this forum - I made a cursory glance at Loop 2007 (the 64-bit version). It uses exactly the same PST as Fruit 2.1. The pawn eval uses the same scores. It shifts the mobility count as with Fruit (by 4,6,7,13) before multiplying, and then the arithmetic is the same (4 for N, 5 for B). I'm not that interested in going further.

2. From the Panel discussion - though with now Loop 2007 looming (which seems even more obviously a Fruit/Toga clone),

Then following the words of Jaap we read -

1. I spoke in Yokohama with mark watkins and gave him the opportunity to tell his story to the assembled chess porgrammers. Mark did so and informed us on the state of affairs. from his point of view it was not as clear as you stated but there were pointers.

2. i admit i was interested in true facts and i spoke to mark watkins. amir ban went into discussion with mark, and a few other people from the audience too. the oral formulation of a kind of conclusion was that there is not a convincing proof, moreover there was a discussion to what extent .... . in brief, it was not clear whether the monks were equal, they might be similar, but the weighing of the proportional part was too difficult for the "audience judges".

----------

So which one of you is telling the truth?

Was Jaap truthful when he stated, "from his [Mark Watkins] point of view it was not as clear as you stated but there were pointers." It obviously conflicts your earlier statements about LOOP, calling it a clone.

Here is the deal, I am more than willing to change the page that annoys you with the answer you give, but you can't be both right.

Everyone is being consistent here except YOU. The term "clone" has a distinct meaning. One example was the long-ago WMCCC event where someone entered a copy of genius. They hex-edited the thing to change the output, but it was a direct copy with no changes to the search or eval. That is a clone. An "identical copy". From Webster's:

one that appears to be a copy of an original form : duplicate <a clone of a personal computer>

Clearly that does NOT apply to loop. How can it be a copy if one uses bit boards and one does not? No one claimed Rybka 1.0 beta was a clone of fruit. Same reasons. Even the infamous earlier versions were not clones of Crafty because there were changes, even though there were many thousands of lines of code directly copied without any changes whatsoever.

We have been using the term derivative for years, it is the right term for all of these recent cases. And you want to quibble if Mark says "loop is not a clone". That does NOT mean nor imply it is either (a) not a derivative nor (b) it is a legal program by ICGA rules. You look silly when ALL you can try to do is twist the meaning of one single word to try to use it to further this pointless debate. And that is ALL you are doing here, quibbling about "not a clone" even though NOBODY believes loop is a clone of anything. But it is certainly not original. In fact loop and Rybka 1.0 beta have a LOT in common with how they were developed...

99% of this nonsense is caused by YOUR improper use of a very precise term, "clone". It does NOT belong in ANY of these discussions.

BTW, to Mark: You were pretty much right on when you briefly analyzed my statement somewhere above in all this nonsense. I don't think Rajlich could have done anything to change any of this, because the evidence is too overwhelming. And short of space aliens and such, there is little possibility of a rational explanation that would convince anyone he was innocent. David went far farther than a judge would have gone. In a court, he would have been asked ONE TIME for a defense. If he declined, that would be it, and it would NOT be grounds for an appeal since it was his choice. You can't appeal due to your own incompetence...

[Chris Whittington]

In practice we interpret carte blanche to mean that Ed and I brainstorm possible actions and make forum posts as we think fit. Before going beyond an idea for a plan and taking any action we consult Vas as to whether he agrees or not.

From this I infer that the breaching of the Panel privacy requirements was thus OK'd by Rajlich (cf. March 2015 publication on Rybka forum). Good to know, and indeed yet another reason why the ICGA should deny any appeal. [The purloining of said Panel discussions, and particularly subsequent "blackmail" attempts by Schröder therein, was a subject in Yokohama, and generally it seems that this type of behaviour is bound to ensure that the "Panel" is not going to be used in subsequent cases, at least not in the same form]. Should I also assume that various defamatory forum posts and/or web hostings by Schröder are also actionable as deriving from Rajlich?

Wrong on several counts.

No you can't infer anything other than a signed document or a formal appeal/approach/complaint to an external body is agreed by Vas Rajlich. Forum posts don't require Rajlich approval, those are independent postings as I and Ed have been making on forums since 1995 on and off.

Secondly you have no idea at all as to the source of the panel discussion leak. User Trotsky stated that he was sent a PM with a link into the sky from an unknown account which has never posted before or since. Could have been any disgruntled panel member. Accusations against Ed merely show your bias. Each shown piece of your bias detracts from the alleged impartiality status of COMP EVAL, a document involving weird metrications to incrimate Rajlich and depending on a curious choice of six comparison programs, chosen by you, some very weird and often incorrect categorisations of program elements, chosen by you, and some highly subjective little numbers, also chosen by you. But please feel free to continue showing dislike of Vas and anyone associated with him and cherry pickign internet quotes to justify actions against him and reason for not putting those right .... we will use the sample examples you have provided.

[Chris Whittington]

I'm not sure why you added the "however" sentence. Do you think it takes something away from the preceding sentence?

It was probably just me being cynical, having seen (certainly in academia) how organisations can often be certain to meet various rules and procedures while at the same time doing so just as a formality, with no real intent therein other than to keep up appearances (in the extreme, the accused will refer to these methods as "Kafkaesque", and indeed might "boycott" the process the same way that the presumptive losers in South American elections will "boycott" the polls so as not to lend legitimacy --- it is not always easy to differentiate between actual fraudulence and a self-fulfilling prophecy). Or to phrase it after-the-factly: even if the ICGA had done everything by-the-book, Rajlich and his representatives would simply find something else to chatter about. [On the other hand, I do agree that as a professional organisation, the ICGA should be (much) more attuned to formalities].

Levy's lifetime ban [properly the ICGA's lifetime ban, though I think my phrasing here is closer to reality] was quite a surprise to me when it was announced, and largely seemed like publicity to me (furthering his imitation of "Olympian" ideas), though as I already said back in 2011, if Rajlich isn't interested in competing in future events, as a practical matter the issue is essentially moot. Perhaps too my upbringing in the US leads me to think that a "lifetime ban" (or the more severe "permanent ban") is usually a misnomer, indeed the Wikipedia page on baseball bannings explicitly mentions: Terms such as "lifetime ban" and "permanent ban" are misnomers, as a banned person may be reinstated (i.e., have the ban removed) on the decision of the Commissioner of Major League Baseball, and in the case of Hall of Fame induction the ban can extend beyond a person's lifetime.

There are also ways of enforcing a "lifetime" ban without making it explicit. For instance, in 2008 Levy rejected the entry of KCC Paduk due to: past problems with [this] program in other computer Go tournaments, which dated back almost a decade (see the last paragraph before 1.1 here for more info and links to the evidence). Another example is Gian-Carlo Pascutto, who is a persona non grata over a dispute on whether he should have paid the commercial entry fee one year (the ICGA is demanding the difference of 250 euros I think, and he won't pay it, and thus would not be allowed to compete until this is resolved [he was partially interested in entering Sjeng in Tilburg 2011] -- undoubtedly a similar point would be made over Rajlich's 2010 prize monies).

So, if the suggestion that "the MAIN thing wrong" is that "it did NOT come from Vas", then I am still puzzled by that suggestion.

I am not going to put words in hyatt's mouth, but it seems that the "formal" issue of Schröder acting as Rajlich's representative could be considered the "MAIN" thing for the EC vis-a-vis Rajlich's appeal and the ICGA's possible ethical violations therein. However, I would think this would be remedied by the email of Oct 5. Levy, on the other hand, ignored any formalities, and already on Aug 27 stated (to Schröder) that no appeal was possible, and reiterated this to Rajlich on Oct 6. The stated reasons for the denial of appeal are predicated upon Rajlich's previous aloofness (and implicitly the lack of higher authority than the ICGA Board to which to address the appeal), but I might say that this is making the issue a bit terse: the side issues of repeated badgering (by Schröder) for the appeal, and moreover the lack of any citation of remotely viable grounds for the appeal -- both of these likely had some role in Levy's decision/wording.

Incidentally on the webpage for correspondence regarding the appeal one finds:

As evidence for that consider Mark Watkins remarks during the programmer meeting in Japan (2013) when he said, "The LOOP case is not so clear", he considered his previous statements as wrong and changed his mind.

This is simply false: I never said the quoted words (in any context AFAIK, and certainly not in comparison to Rybka/Fruit); furthermore they do not represent my opinion; and finally I consider my previous statements to be correct and have not "changed my mind." If these were put forth by Rajlich (through his representative) as grounds for an appeal [Oct 6 email to Levy], I would consider it to be a deliberate mischaracterisation.

And Levy should resign.

Wouldn't it have been easier to just buy out the ICGA election? It's got to be what, something like 50-100 votes at 40 euros a membership? Surely there's enough Rybka forum acolytes who would sign up for this... and as a bonus you could now revoke Hyatt's 1983/86 Championships! Sadly, the deadline for 2015 election candidates passed a few days before the EC decision arrived.

Ref: final parapraph - you really don't like anythign associated with Rybka, do you? Why the great bias? Doesn''t match with unbiased independent expert status, or do you not claim your work was of independent character?

There is no time limitation on resigning by the way.

[Chris Whittington]

It was bad enough in this thread when you were previously simply getting things wrong (eg, "fundamental right" to an appeal, Rajlich retracting his statement that Strelka contained Rybka code, old Glaurung containing Fruit-like method with infinite/ponder searches), but deliberately distorting people's words is even worse.

Anyway, on what may very well be last response to you, in this thread or otherwise...

I made a cursory glance at Loop 2007 (the 64-bit version). It uses exactly the same PST as Fruit 2.1. The pawn eval uses the same scores. It shifts the mobility count as with Fruit (by 4,6,7,13) before multiplying, and then the arithmetic is the same (4 for N, 5 for B). I'm not that interested in going further.

As should have been clear from the context, I was not interesting in going further [at that time] because what I had found by the first cursory inspection already made it clear that there was a lot of Fruit/LOOP overlap, and this sufficed for the time being. Later in that thread (which I will take the time to cite), I gave a (lightly) annotated disassembly of the relevant LOOP eval function.

2. From the Panel discussion - though with now Loop 2007 looming (which seems even more obviously a Fruit/Toga clone),

As you know, the Panel discussions are private. In any case, "More" does not mean "an undoubtedly 100% clone" to quote your words to Jaap. I am in full agreement (with myself) that the LOOP/Fruit overlap is more than the Rybka/Fruit overlap (which is the context of the above quotation). But I am also in sharp disagreement with your phrasing of LOOP being "an undoubtedly 100% clone", that for some reason you have attributed to me.

So which one of you is telling the truth?
Was Jaap truthful when he stated, "from his [Mark Watkins] point of view it was not as clear as you stated but there were pointers." obviously conflicts your earlier statements about LOOP, calling it a clone.

Yes, Jaap was truthful -- it is not my point of view that the Fruit/LOOP situation is "as clear as you stated", meaning "an undoubtedly 100% clone". Furthermore, I am telling the truth and there is no "obvious conflict", as I explain above. I think the only one who is not truthful is you, when you introduced this "an undoubtedly 100% clone" phrasing, and somehow then tried to palm it off as being from my mouth or writings.

To conclude:

• MarkW implied/said (in 2011 forum posts and private Panel discussions) that LOOP is more obviously a Fruit "clone" than Rybka,
  though made little if any quantification of this.
• Mr. Schröder wrote to Jaap and indicated that (contrary to Rybka) LOOP was "an undoubtedly 100% clone".
• Jaap correctly replied that MarkW (in his 2013 impromptu Yokohama presentation) did not think it was "as clear as you stated [ie 100% clone]".
• Jaap indicated that in my talk "there were pointers" [to LOOP sharing Fruit origins], namely that I sketched the evidence that was in the 2011 Forum post cited above.
• Jaap further (correctly) noted "moreover there was a discussion [with Amir Ban in particular] to what extent" LOOP was a clone, and any conclusion herein was not determined on-the-spot by the "audience judges" (after a 15-minute briefing), but would rather await a fuller analysis.
• Mr. Schröder, as Rajlich's representative, somehow confused all this, and asserted that MarkW had "changed his mind" regarding LOOP, writing to Levy: Mark Watkins remarks during the programmer meeting in Japan (2013) when he said, "The LOOP case is not so clear", he considered his previous statements as wrong and changed his mind.
• Mr. Schröder moreover (in this thread) erroneously proposes that Jaap's comments were in response to previous statements [MarkW] made here (on this forum) Loop undoubtly being a Fruit clone -- whereas the only use of this "undoubtedly" terminology was from Schröder himself.

Perhaps Jaap misunderstood part of the discussion but my recollection of it is as Mark has posted.

I think the misunderstanding is at least 90% that of Mr. Schröder.

hahahahahahahaha!!M and tou think it is 1 in 10,0000,0000000 gazillion that Rybka copied Fruit. What to believe with all this subjective numbery jiggery pokery ....?

[Chris Whittington]

I have the email in front of me and it states:

Yes, on October 1st you inquired to me about this email. It was about the time of your ChessVibes article, and I think I ignored it (I have no record of an outgoing email), as it was obvious you were deliberately interpreting Jaap's words to your own desire.

I am not responsible for your miscommunications especially not when feed with contradictive statements.

You are, however, responsible for deliberately twisting the words to your/Rajlich's advantage, quoting them incorrectly and out of context in an official(-ish) email to Levy.

There is another thing that bothers me, the Fritz Reul Loop case. Contrary to Rybka this is an undoubtedly 100% clone. ...

from his point of view it was not as clear as you stated

Here the phrase "as you stated" clearly referred to LOOP being "an undoubtedly 100% clone" as can be seen from your inquiry above. I agree with this, LOOP is not "100% clone" for the obvious reason that at the very least it is a bitboard rewriting, moreover it does have one or two new eval features (like a bonus for being on-tempo). And if one were to use the argumentation you use (or should I say "Rajlich uses", if you are indeed his representative?), then one could presumably enumerate up to "36 indisputable differences" in the evaluation (see page 58 of Rybka Reloaded PDF or your ChessVibes articles) including Fen parsing is different and Fruit has a 16*16 square mailbox, Rybka has an 8*8. [Note that Riis only had 10 "evaluation differences" in his original article (page 15-17), though I disputed all but at most 3-4 of these (page 7 of my rebuttal), and he later copied the first 21 of your enumeration into his follow-up in Entertainment Computing].

Let's review this again: At Yokohama, I gave a presentation to the 6 programmers, about clones in general, and mentioned LOOP. You later wrote to Jaap about LOOP, called it "an undoubtedly 100% clone", to which he replied that [MarkW said] from his point of view it was not as clear as you stated. You then relayed this to Levy as: As evidence for that consider Mark Watkins remarks during the programmer meeting in Japan (2013) when he said, "The LOOP case is not so clear", he considered his previous statements as wrong and changed his mind. See how you conveniently (as Rajlich's representative) ignored Jaap's words "as you stated", and put quotes around something I never said! My best recollection is that what I said is "LOOP is essentially a bitboard rewrite of Fruit, with a few minor changes."

The Fabien Loop complaint is now more than 2 years old and nothing yet has happened.

Part of the reason is that Reul has not (yet) been able to be contacted. Whereas Rajlich was sentenced in absentia only in the mind of Riis (and indeed, in the above article he repeats this error, even though both Levy and I corrected him in our rebuttals to his original ChessBase piece), Reul's case does seem that it will end up being that way.

La lucha continua.

Do you mean this, or La lucha continúa? The first means "The endless fight", while the second means "The fight continues"... Then again, I guess "I am not responsible for your miscommunications"...

hahahahahhahahah!!! sorry, I can't help if you lack social, political and historical reference points, your use of google translqte notwithstanding.

[Rebel]

Anyway, on what may very well be last response to you, in this thread or otherwise...

Or what?

Better get a grip on yourself, you are all worked up and take this personal while it isn't.

This is about truth finding.

I am entitled to my opinion that at the time of writing the ICGA was in suspect to wipe the LOOP case under the carpet because it involved an ICGA VIP. While Jaap can't be blamed for making a commercial deal with Nintendo (with LOOP) not realizing the wasp' nest he was entering with Reul he (in his mail) was downplaying the fraud Reul commited for his own interest at the expense of YOU. If you don't get it then I am sorry for you.

BTW, now that you spoken out, I have removed the text you took offense.

[hyatt]

In practice we interpret carte blanche to mean that Ed and I brainstorm possible actions and make forum posts as we think fit. Before going beyond an idea for a plan and taking any action we consult Vas as to whether he agrees or not.

From this I infer that the breaching of the Panel privacy requirements was thus OK'd by Rajlich (cf. March 2015 publication on Rybka forum). Good to know, and indeed yet another reason why the ICGA should deny any appeal. [The purloining of said Panel discussions, and particularly subsequent "blackmail" attempts by Schröder therein, was a subject in Yokohama, and generally it seems that this type of behaviour is bound to ensure that the "Panel" is not going to be used in subsequent cases, at least not in the same form]. Should I also assume that various defamatory forum posts and/or web hostings by Schröder are also actionable as deriving from Rajlich?

Wrong on several counts.

No you can't infer anything other than a signed document or a formal appeal/approach/complaint to an external body is agreed by Vas Rajlich. Forum posts don't require Rajlich approval, those are independent postings as I and Ed have been making on forums since 1995 on and off.

Secondly you have no idea at all as to the source of the panel discussion leak. User Trotsky stated that he was sent a PM with a link into the sky from an unknown account which has never posted before or since. Could have been any disgruntled panel member. Accusations against Ed merely show your bias. Each shown piece of your bias detracts from the alleged impartiality status of COMP EVAL, a document involving weird metrications to incrimate Rajlich and depending on a curious choice of six comparison programs, chosen by you, some very weird and often incorrect categorisations of program elements, chosen by you, and some highly subjective little numbers, also chosen by you. But please feel free to continue showing dislike of Vas and anyone associated with him and cherry pickign internet quotes to justify actions against him and reason for not putting those right .... we will use the sample examples you have provided.

This is poppycock. Ed stated on more than one occasion that he had made "a complete backup of the Wiki" even though he had formally left and stopped participating. He's published bits and pieces multiple times. So this "anonymous person" is pure bullshit.

BTW was your last statement made to Ed or to Mark? Because Ed has made a statement and then attributed it to mark, and then tried to use that statement (which Mark did not make as given) to suggest Mark had then changed his mind about loop. Mark just pieced together the conversation to show how it was distorted by leaving out context.

[hyatt]

Anyway, on what may very well be last response to you, in this thread or otherwise...

Or what?

Better get a grip on yourself, you are all worked up and take this personal while it isn't.

This is about truth finding.

I am entitled to my opinion that at the time of writing the ICGA was in suspect to wipe the LOOP case under the carpet because it involved an ICGA VIP. While Jaap can't be blamed for making a commercial deal with Nintendo (with LOOP) not realizing the wasp' nest he was entering with Reul he (in his mail) was downplaying the fraud Reul commited for his own interest at the expense of YOU. If you don't get it then I am sorry for you.

BTW, now that you spoken out, I have removed the text you took offense.

Please stop with the bullshit. We discussed the loop case. It was NEVER "about to be swept under the carpet". I told you that more than once. Sometimes things move slowly (loop). Sometimes they move ultra-slowly (Rybka). But they do generally move. I had told you David was trying to contact the author through every possible means. No responses at all. Which was not the same as the Rybka case where Vas just continually said "not now."