Fruit/Rybka timeline

[BB+]

Here is a timeline that I derived for various important dates with Fruit/Rybka (until the whole thing broke loose on CCC; there are better sources for later events).

Code: Select all

Mar     2003: Rajlich starts Rybka development.
Nov  30 2003: He makes his first CCC post.
Dec  28 2003: He mentions Rybka for the first time on CCC.
Dec  31 2003: He mentions Rybka uses Kaufman's material imbalances.
Jan End 2004: Rybka participates in CCT6.
Feb   6 2004: He asks Nalimov for permission to use his code.
              [He posts many times this day, including about Rybka eval]
Mar  10 2004: Fruit 0.98a is trickling around, gets mentioned by Dann Corbit.
              [can be downloaded from his site, I think just as an executable]
Mar  15 2004: Rajlich posts about Fruit.
Mar  17 2004: Fruit 1.0 (with source code) is released.
Mar  27 2004: Rajlich posts NPS numbers for 64-bit Rybka 1.3.
Mar  31 2004: Rajlich enters Rybka 1.4 (then 1.5) into ChessWar.
              [Over the next week, various bug-fixed versions are submitted]
Nov  24 2004: Rajlich enters Rybka 1.6 into ChessWar.
Dec  24 2004: Fruit 2.0
Apr   5 2005: Interview of Letouzey with Frank Quisinsky
Jun  17 2005: Fruit 2.1 is released.
Jun  20 2005: Rajlich comments on Fruit 2.1
Aug/Sep 2005: Rajlich quits his job, develops Rybka full-time
Nov  30 2005: Rajlich announces that Rybka 1.0 will be ready on Friday
Dec   5 2005: Rajlich notes he has 200 requests for the Beta, says it will
              be downloadable until midnight tonight. Commercial release
              will happen soon (Dec 16, moved up to 11th), cost of 34 euros.
Dec 11/12 05: Mehrmann posts, and Rajlich responds, about how much Fruit
              is in Rybka ("The Rybka source code is original")
Dec  16 2005: Rajlich re-asks for Nalimov's permission (no longer amateur)
Dec  20 2005: Interview with Frank Quisinsky (and others)
              Implies he gained 20 Elo from going back/forth thru Fruit sources
Dec  20 2005: Enrico Carrisco questions one element of the interview, saying
                "I can say that HIARCS shouldn't be listed as benefiting
                 from Fruit 2.1's open source code, however. Clearly, less
                 developed engines and amateurs will benefit from such code
                 more than long-term and highly developed projects."
Dec  23 2005: Uri Blass notes that Rybka "node counts" can decrease.
Dec 27-30 05: Rybka competes and wins in Paderborn.
Dec/Jan 05/6: Various Rybka 1.0.x versions appear
Jan   1 2006: Enrico Carrisco inquires about "nodes" in Rybka.
              Rajlich responds: "What is the Hiarcs definition of node?"
Feb   8 2006: Carrisco talks more about NPS; Rajlich calls his theories "wild"
Feb  13 2006: Peter Skinner is so annoyed by commercial Rybka postings
              on TalkChess that he resigns as CCT8 director.
Feb 24/25 06: Rybka wins CCT8 (with Skinner directing)
              [Note that Rybka had connection problems, but that Carrisco,
               as the HIARCS operator did not claim the forfeit win]
Mar  16 2006: Rybka 1.1 is released (already 75 Elo more than 1.0)
May 1/5 2006: Rybka 1.2 is released, the last being 1.2f
              Rybka has gained about 100 Elo over 1.0 Beta by now
May/Jun 2006: Rajlich competes in the 2006 WCCC (under his own name,
              presumably to avoid the commercial entry fee?)
Jun  10 2006: Rybka 2.0 Beta (MP only) is released
Jul  26 2006: Rybka 2.1 is released
Nov  10 2006: Rybka 2.2 is relased
Dec  14 2006: Rybka 2.2n2 is released
Jan   ? 2007: Larry Kaufman is hired.
Feb 12/15 07: Rybka 2.3 is released.
early?  2007: Someone pretends to be LK, and internal Rybka info is leaked.
May  4? 2007: Strelka appears
Jun 13/18 07: Rybka 2.3.2a is released, Rybka competes in the Amsterdam WCCC
              It is about 40-50 Elo stronger than 1.2f on equal hardware
Jan   8 2008: Strelka 2.0B with source code is released
Apr ~19 2008: Dann Corbit contacts Letouzey about Strelka/Fruit similarities,
              failing to mention Rybka.
Apr End 2008: Rick Fadden posts on Rybka Forum (as ExperiMental), asking
              Rajlich to correct the Rybka 1.0 node counts
Apr/May 2008: Rick Fadden disassembles Rybka 1.0 Beta, comparing it to Strelka
Aug   5 2008: Rybka 3 is released, about 90-100 Elo better than Rybka 2.3.2a
Aug ~15 2008: Zach Wegner (joined by Christophe Theron) begins to ask rather
              specific questions about the Rybka/Fruit overlap.
              [Some of this seems now hidden at CCC?]

[BB+]

I forgot to prepend the following: Oct 1 2002, Crafty 19.0 is released.

[MoldyJacket]

Thank you very much, I found this timeline extremely helpful in wading through all the articles, documentation and discussions (which I’m still doing).

Forgive me for asking but I noticed you refer to yourself as a mathematician at the University of Sydney in the Dr Søren Riis critical analysis document, yet you are listed as a research fellow on staff. Are you being modest in not taking the doctorate mantle?

[hyatt]

Thank you very much, I found this timeline extremely helpful in wading through all the articles, documentation and discussions (which I’m still doing).

Forgive me for asking but I noticed you refer to yourself as a mathematician at the University of Sydney in the Dr Søren Riis critical analysis document, yet you are listed as a research fellow on staff. Are you being modest in not taking the doctorate mantle?

"research fellow" and "doctor" are not exclusive. My daughter earned her Ph.D. and took a position here at UAB for a couple of years, called a "post-doc fellowship." Another student (my advisee) took a position elsewhere after finishing his Ph.D. and his title was "research fellow". Typically, a "research fellow" does not teach, but is devoted to research (and obtaining funding for same). A post-doc is similar. A "teaching fellow" is primarily a teacher. All could have PhD degrees, or none might. (except for the post-doc of course, that certainly implies you have one).

[BB+]

I received by PhD in mathematics from Georgia in 2000. Your question is relevant though, as one of my colleagues (also a "Research Fellow" I think) recently got his PhD after being (mostly technically) a grad student for 20 years. The term "Research Fellow" in this context means largely that I am supported by a research grant (or Magma user fees), and that I don´t typically teach any classes. I guess I could complain that Dr. Riis did not call me "Dr. Watkins", but then such honorifics tend to get a bit clumsy in a sentence such as: "The unanimous ICGA Board decision was made by Dr. Levy, Prof. Bjornsson, Prof. Iida, Prof. Coulom, and Prof. van den Herik." It is also not always easy to know whether an "Associate Professor" in the US should be called a "Professor" in the European sense. The lesser term of "Assistant Professor" (usually pre-tenure) typically does not equate to this.

[MoldyJacket]

Thanks Dr. Watkins.