FIDE Rules on ICGA - Rybka controversy

[Harvey Williamson]

But it would be nice if you would have told the full story of the lost source and not only half.

As I had not heard "this half" of the story, I would see no reason for Harvey to know about it either.

In December 2012 source code was recovered of an in between R1 - R2 version which was suggested as a base for an indepedent unbiased expert to close the matter once and for all. Levy did chicken out. Like Levy did chicken out to grant Rajlich an appeal via the same circus (Panel | Secratary | Board).

If Rajlich had mentioned this in his "appeal", perhaps Levy would have taken this into consideration after having received new pieces of evidence.

I should have been more clear on that yes, that source code was available, my bad. It's there between the lines.

• Good morning David,
  
  Two points.
  
  1. I don't understand your refusal to re-open the Rybka-Fruit case. In the past (the WCCC in Graz 2003) the ICGA banned a suspect program (the LIST program of Fritz Reul) because he refused to cooperate in an investigation. Later Reul did cooperate and Reul was reinstated after he handed over his source code to (expert) Dr. Chrilly Donninger who inspected the LIST source code and found it original. So if Reul was granted an appeal I don't understand your refusal for Rybka especially now that a sea of contra evidence is available.
  
  An alternative to the time consuming nature of an appeal is to follow the same procedure as the ICGA did in the LIST (Fritz Reul) case and ask an independent expert both sides agree on and his decision will be final. My suggestion would be Dr. Donninger again. As an intermediator I can lobby for the precedent you set in the LIST case with Vasik.

Nevertheless I think it wouldn't matter much. David was quite abrupt from the beginning. You might find my behaviour towards David vexatious, in reality it shows his unwillingness to a reasonal request with respect a precedent he set himself in the past with Reul.

I am guessing that email was from you to David not from Vas?

You are not guessing, you know Mark knows.

But now that you are here, isn't the request a reasonable one with the Reul case in mind?

I am afraid I don't know/remember as I don't live and breath this case 24/7 as you seem to. I have moved on and I am sure Vas has as well. The ICGA should look at the life sentence.

[Rebel]

I am afraid I don't know/remember as I don't live and breath this case 24/7 as you seem to. I have moved on and I am sure Vas has as well. The ICGA should look at the life sentence.

Well then, if you don't know or remember, why ask a suggestive question?

Context was very clear, wasn't it?

[hyatt]

The MAIN thing wrong with that request is that it did NOT come from Vas. In most court systems, ONLY the accused can request any sort of appeal or re-hearing. Not disinterested third parties. That was the first comment in the FIDE EC ruling if you recall? Whether it would do any good or not is unknown, but certainly requests from others is not going to attract any attention from the ICGA board relative to this decision of theirs. I would not expect them to listen if _I_ requested an appeal. There's only one person that can actually do this, and almost 4 years is more than a little late.

[Rebel]

The MAIN thing wrong with that request is that it did NOT come from Vas.

This is sooooo silly. The underlying implication is that I fabricated the whole thing. Fraud to be precisely.

Now I understand that I am not very popular with David but I don't buy it that I am no longer trustworthy and my word isn't good enough.

Do you really think the outcome would have been different (yes to appeal) if Vas had sent that mail?

Such a lame excuse.

[Harvey Williamson]

Do you really think the outcome would have been different (yes to appeal) if Vas had sent that mail?

It would have shown that Vas had some interest in an appeal.

[Rebel]

Do you really think the outcome would have been different (yes to appeal) if Vas had sent that mail?

It would have shown that Vas had some interest in an appeal.

So I have commited fraud.

Thank you Harvey.

[syzygy]

The MAIN thing wrong with that request is that it did NOT come from Vas. In most court systems, ONLY the accused can request any sort of appeal or re-hearing. Not disinterested third parties.

What a complete nonsense. It was submitted on behalf of Vas and that is sufficient. That it was also submitted in the name of other parties that were not considered to have an interest is irrelevant. (Actually those parties explained why they considered themselves to have an interest. This was not accepted by the EC, but that does not harm the case in any way. I guess one has to have had some legal training to understand such obvious things?)

I don't know if Vas signed a "power of attorney" or signed the letter himself, but if the EC had had any doubt it would certainly have asked for clarification.

[BB+]

Well, Vas reached out to David, invited him for a phone call giving David his private number. David refused. Demanded email. Not very polite. Actually quite rude. Of course after the phone call the conversation would continue in email. Maybe Vas had something on his chest to say in private before continuing? Missed chance by David?

"Invited" a phone call, but "demanded" email... There would be good reasons for Levy to prefer not to do anything private (either with Rajlich or Letouzey). If you take your "court" analogy, does someone who is likely to judge a future case typically go around having private conversations on the topic with the accused? Should Levy have offered to have Rivello call him (after 1400 European time on weekdays) to privately discuss the EC case?

Anyway the FIDE EC was quite clear on the poor quality of the David - Vas email exchange, quoting :

Levy was of the opinion that one should first decide on the guilt, and then on the punishment. But as the EC pointed out, there was a necessity of (explicitly) informing Rajlich of the possible penalties/sanctions. However, given that Rajlich knew it was clear that the ICGA had no intention of handling things in a fair manner, I can't think that such a "right to be heard" would be seen by him as anything more than perfunctory.

Ah, that's what you mean. Have you noticed that in a version of Glaurung you find similar code?

Can you be a bit more specific? Here is the relevant part of Fruit 2.1 in protocol.cpp, with the following code (rather oddly) at the end of parse_go():

Code: Select all

   if (infinite || ponder) SearchInput->infinite = true;
   // search
   ASSERT(!Searching); ASSERT(!Delay);
   Searching = true;
   Infinite = infinite || ponder;
   Delay = false;
   search();
   search_update_current();
   ASSERT(Searching);
   ASSERT(!Delay);
   Searching = false;
   Delay = Infinite;
   if (!Delay) send_best_move();

As noted, "infinite||ponder" is redundantly calculated twice. Moreover Delay is set to false even though it was ASSERTed to be so three lines previous. Two quirks in ten lines of code. Strange that Rybka 1.0 copied them both (I refer you to page 16 of RYBKA_FRUIT for the disassembly). Looking at Glaurung 2.2, I don't see anything of this sort, the relevant function is called wait_for_stop_or_ponderhit(). The older 1.2.1 also doesn't seem to have anything like this (the function uci_go() calls think(), which handles time management, and then root_search() has the same wait_for_stop_or_ponderhit() as in 2.2). Whereas the Fruit/Rybka similarities are clear, I don't see any "similar code" in the Glaurung versions I could find.

I should have been more clear on that yes, that source code was available, my bad. It's there between the lines.

It would be nice if which version of the source code in question was made clear too... (I had thought this meant Rybka 4, which might make a difference regarding the "lifetime" aspect of the ban if Rajlich wished to later compete, but would be irrelevant for previous matters). Also, as far as I can tell, that email is from 2013-Aug-27, while the email from Rajlich (to the ICGA) appointing you as his representative is dated 2013-Oct-05. Why would you think the ICGA would look seriously at your "contra evidence" before then? Anyway, as you noted MarkL labelled Rybka Reloaded as "rhetoric", while my own terminology would perhaps not be so polite. When I spoke with Aske Plaat and Jaap in Yokohama, I got the impression that Tilburg was aware of your work, and that someone had concluded it was at best irrelevant.

Did I miss anything?

Yes. For instance, copying the multiplication by 5 (one line after the infamous 0.0). See also 4.1.3 of my RECAP document, regarding BadThreshold and flag usage. The point is not that there is a "large" or "small" amount of copying, but rather that these are at a very low level ("code", to the extent that it is possible w/o the Rybka 1.0 source), which both points to something like copy-paste in these segments and further leads one to have some reason to make an investigation of other similarities between Fruit 2.1 and Rybka 1.0.

[BB+]

The MAIN thing wrong with that request is that it did NOT come from Vas. In most court systems, ONLY the accused can request any sort of appeal or re-hearing. Not disinterested third parties.

What a complete nonsense. It was submitted on behalf of Vas and that is sufficient. That it was also submitted in the name of other parties that were not considered to have an interest is irrelevant. (Actually those parties explained why they considered themselves to have an interest. This was not accepted by the EC, but that does not harm the case in any way. I guess one has to have had some legal training to understand such obvious things?)

I don't know if Vas signed a "power of attorney" or signed the letter himself, but if the EC had had any doubt it would certainly have asked for clarification.

I'm not quite sure what the argument is, but I thought [as evident from later parts of his post] that hyatt was talking about the (Aug 27) request by Schröder to the ICGA regarding an appeal et al. Given that (to my knowledge) Rajlich only sent an email to the ICGA regarding Schröder's capacity as his representative on Oct 5, I am not sure why the ICGA should accept anything Schröder says ("on behalf" of Vas or otherwise) before that? Maybe you two are talking about different things?

Regarding the EC, there has been some doubt expressed that Rajlich's emendation was accepted, but the EC decision says that it was "integrated", and I can attest it was sent to the ICGA and they responded to it, so I can't imagine the EC itself ignored it.

Not sure the laws where you live but here an appeal is granted whether a criminal or civil case.

Operation of the High Court (of Australia): There is no automatic right to have an appeal heard by the High Court and parties who wish to appeal must persuade the Court in a preliminary hearing that there are special reasons to cause the appeal to be heard. Note again there is a distinction between the "right to appeal", and the right to have an appeal be heard.

Incidentally, one of the more famous "vexatious" litigants (David Eastman) in the Australian court system had the habit of firing attorney A, necessitating A showing up in court to say that his instructions were withdrawn, then Eastman would represent himself for a day, hire B, fire B after a few days, re-hire A, who then needed to be filled in on what had happened in the interim, etc. Delay. Delay. Delay. See #34 (page 15 of the PDF, internal page 10) of this. [Eastman is now awaiting retrial, as explained in that PDF, the special advisor (Board of Inquiry) accepted Paragraph 5 of his appeal, namely that one of the forensic investigators was too personally attached to the case -- the full High Court did not accept the advisor's full suggestion (quashing and pardon), but instead ordered a retrial upon quashing].

[Chris Whittington]

The MAIN thing wrong with that request is that it did NOT come from Vas. In most court systems, ONLY the accused can request any sort of appeal or re-hearing. Not disinterested third parties.

What a complete nonsense. It was submitted on behalf of Vas and that is sufficient. That it was also submitted in the name of other parties that were not considered to have an interest is irrelevant. (Actually those parties explained why they considered themselves to have an interest. This was not accepted by the EC, but that does not harm the case in any way. I guess one has to have had some legal training to understand such obvious things?)

I don't know if Vas signed a "power of attorney" or signed the letter himself, but if the EC had had any doubt it would certainly have asked for clarification.

Quite so. A person's representative can sign for him.

Representative status: I am Vas Rajlich's representative/advisor and have been since the start of this affair. I sent and signed (as did Vas Rajlich) the "can we appeal to FIDE" document which FIDE listed as a Complaint and asked for further information. FIDE explicitly accept this status, they commented on it in their report.

Ed is also Vas Rajlich's representative/advisor and has been for a long time. Ed can state when this was formally confirmed, I forget. But certainly before the FIDE documents were sent in.

In fact we both have more than just representative status, we have a carte blanche freedom to act on his behalf. I asked Vas if he wanted to continue this carte blanche now that the favourable result from FIDE was in and the forums had settled on a general viewpoint from that, and he confirmed it again (myself and Ed). 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. If he doesn't, then we don't go down that path. If he does, he is involved as he needs to be (see below).

Involvement of Vas: Vas is heavily involved enough, if you want to metricate the involvement, to have written to me 445 emails during this process. The most recent over the last few days. I guess there are a similar number to Ed, and many are group emails.
Vas has the capacity to completely switch off this whole matter and get on with his own other work. He has a family to look after. He does not wish to to be writing documents or forum posts or arguing. He prefers that he has other volunteers to do this for him, and he trusts that we liaise with him, as we do. He has the final word. He has rid himself of all the stress. He would like his name cleared.

In my opinion, the endless demands from Hyatt-ICGA that Vas appear in person, or write in person, and can only make an appeal in person, and is condemned for not appearing before the biased Secretariat in person to be grilled, are time wasters and decoys. This is proven by the eventual statement of Levy that "no appeal would be allowed, whether it was made by Vas in person or not", following endless statements by both Secretariat members that an appeal would be entertained IF, and ONLY IF, it came from Vas in person.

La lucha continua.