Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Rybka author rehabilitated but apologist Dr Søren Riis fails to make the case that Rybka is guilty as charged
20 views
Skip to first unread message
raylopez99
Jan 3, 2012, 11:44:19 AM1/3/12
to
An interesting article posted below on Chessbase, 2nd part of three.
In it, Dr. Riis, an apologist for Rybka programmer Rajlich, tries to
make the case Rajlich was unfairly banned by the ICGA.
He fails.
Why? Because being banned by the ICGA (the sponsor of computer chess
tournaments) is NOT the same as being guilty of copyright
infringement. It's clear that probably Rajlich is not guilty of
copyright infringement, since his Rybka borrowed from open source
"Fruit", as well as open source "Crafty", which are in the public
domain. What is a violation of ICGA however --and the reason Rybka
was banned--is not the legal rules of copyright but the HOUSE RULES of
plagiarism. The house rules by ICGA state: (WCCC Rule 2, related to
program originality).
"Each program must be the original work of the entering
developers. Programming teams whose code is derived from or including
game-playing code written by others must name all other authors, or
the source of such code, in their submission details. Programs which
are discovered to be close derivatives of others (e.g., by playing
nearly all moves the same), may be declared invalid by the Tournament
Director after seeking expert advice. For this purpose a listing of
all game-related code running on the system must be available on
demand to the Tournament Director."
Two violations of this rule by Rybka's author occurred. First, Rybka's
Rajlich did NOT give credit to Fruit IN THE SUBMISSION DETAILS. He
gave it publicly, but that is not the same. He has to say so in the
submission details.
Second, he did NOT hand over the source code to the Tournament
Director. Instead, the ICGA had to spend 18 months or so reverse
engineering Rybka to find similarities with Fruit. In one instance,
they even found where a float 0.0 was mistakenly converted to an int 0
but Rajlich apparently left the 'dot' notation in: so it was "0.",
which compiled as a float. In any event, it's not that Rajlich did
not do original work on Rybka, nor is it that he legally is guilty of
copyright infringement, nor that he did not give credit to Fruit and
Crafty publicly, rather, he did not play by the rules and satisfy the
two conditions above.
Case closed. Rybka's author is guilty of violating ICGA house rules
on computer chess programs. It's true that no doubt both Dr. Hyatt
and the other rival computer chess programmers were jealous of Rybka's
success (or rather the offspring of Fruit's success, since they were
derivatives of Fruit), but that does not make their conclusions less
sound.
RL
http://www.chessbase.com/newsdetail.asp?newsid=7807
A Gross Miscarriage of Justice in Computer Chess (part two)
03.01.2012 – In this part Dr Søren Riis of Queen Mary University in
London
From the initial release of Rybka 1.0 Beta until early 2011 the
general consensus in the computer chess community was that Rybka was
an original program, even though it had undoubtedly learned a great
deal from the leading open source programs in existence in 2005, Fruit
and Crafty. This consensus was based on Rybka’s position on the Elo
rating lists and steadily increasing strength, its search behavior,
unique features, etc.
Rajlich conceded from the very beginning that he had studied Fruit’s
open source code very closely and learned a great deal from it. In an
interview from 2005, right after Rybka 1.0 Beta was released, Rajlich
acknowledged the computer chess community’s as well as his own debt to
Fruit:
Yes, the publication of Fruit 2.1 was huge. Look at how many
engines took a massive jump in its wake: Rybka, HIARCS, Fritz, Zappa,
Spike, List, and so on. I went through the Fruit 2.1 source code
forwards and backwards and took many things.
Rajlich later publicly praised the work of Fabien Letouzey:
I don't want to get too specific about which ideas from Fruit I
think are really useful, but they fall into two categories:
1) Very specific tricks, mostly related to search.
2) Philosophy of the engine (and in particular of the search).
Fruit could really hardly be more useful along both of these
dimensions. Fabien is a very good engineer, and also has a very clear
and simple conception of how his search should behave.
Finally, Rybka 1.0 Beta’s Readme file gave credit to Fruit in a
“Special Thanks” section:
…for Fruit, which shattered a number of computer chess myths,
demonstrated several interesting ideas, and made even the densest of
us aware of fail-low pruning.
By definition, plagiarism only happens when credit to sources is not
given, which was never the case with Rybka. These acknowledgements
were widely known at the time Rybka entered WCCC tournaments and,
rather obviously, Rybka’s debt to Fruit steadily diminished over the
years as program improvements superseded the original program code in
Rybka 1.0 Beta.
In it, Dr. Riis, an apologist for Rybka programmer Rajlich, tries to
make the case Rajlich was unfairly banned by the ICGA.
He fails.
Why? Because being banned by the ICGA (the sponsor of computer chess
tournaments) is NOT the same as being guilty of copyright
infringement. It's clear that probably Rajlich is not guilty of
copyright infringement, since his Rybka borrowed from open source
"Fruit", as well as open source "Crafty", which are in the public
domain. What is a violation of ICGA however --and the reason Rybka
was banned--is not the legal rules of copyright but the HOUSE RULES of
plagiarism. The house rules by ICGA state: (WCCC Rule 2, related to
program originality).
"Each program must be the original work of the entering
developers. Programming teams whose code is derived from or including
game-playing code written by others must name all other authors, or
the source of such code, in their submission details. Programs which
are discovered to be close derivatives of others (e.g., by playing
nearly all moves the same), may be declared invalid by the Tournament
Director after seeking expert advice. For this purpose a listing of
all game-related code running on the system must be available on
demand to the Tournament Director."
Two violations of this rule by Rybka's author occurred. First, Rybka's
Rajlich did NOT give credit to Fruit IN THE SUBMISSION DETAILS. He
gave it publicly, but that is not the same. He has to say so in the
submission details.
Second, he did NOT hand over the source code to the Tournament
Director. Instead, the ICGA had to spend 18 months or so reverse
engineering Rybka to find similarities with Fruit. In one instance,
they even found where a float 0.0 was mistakenly converted to an int 0
but Rajlich apparently left the 'dot' notation in: so it was "0.",
which compiled as a float. In any event, it's not that Rajlich did
not do original work on Rybka, nor is it that he legally is guilty of
copyright infringement, nor that he did not give credit to Fruit and
Crafty publicly, rather, he did not play by the rules and satisfy the
two conditions above.
Case closed. Rybka's author is guilty of violating ICGA house rules
on computer chess programs. It's true that no doubt both Dr. Hyatt
and the other rival computer chess programmers were jealous of Rybka's
success (or rather the offspring of Fruit's success, since they were
derivatives of Fruit), but that does not make their conclusions less
sound.
RL
http://www.chessbase.com/newsdetail.asp?newsid=7807
A Gross Miscarriage of Justice in Computer Chess (part two)
03.01.2012 – In this part Dr Søren Riis of Queen Mary University in
London
From the initial release of Rybka 1.0 Beta until early 2011 the
general consensus in the computer chess community was that Rybka was
an original program, even though it had undoubtedly learned a great
deal from the leading open source programs in existence in 2005, Fruit
and Crafty. This consensus was based on Rybka’s position on the Elo
rating lists and steadily increasing strength, its search behavior,
unique features, etc.
Rajlich conceded from the very beginning that he had studied Fruit’s
open source code very closely and learned a great deal from it. In an
interview from 2005, right after Rybka 1.0 Beta was released, Rajlich
acknowledged the computer chess community’s as well as his own debt to
Fruit:
Yes, the publication of Fruit 2.1 was huge. Look at how many
engines took a massive jump in its wake: Rybka, HIARCS, Fritz, Zappa,
Spike, List, and so on. I went through the Fruit 2.1 source code
forwards and backwards and took many things.
Rajlich later publicly praised the work of Fabien Letouzey:
I don't want to get too specific about which ideas from Fruit I
think are really useful, but they fall into two categories:
1) Very specific tricks, mostly related to search.
2) Philosophy of the engine (and in particular of the search).
Fruit could really hardly be more useful along both of these
dimensions. Fabien is a very good engineer, and also has a very clear
and simple conception of how his search should behave.
Finally, Rybka 1.0 Beta’s Readme file gave credit to Fruit in a
“Special Thanks” section:
…for Fruit, which shattered a number of computer chess myths,
demonstrated several interesting ideas, and made even the densest of
us aware of fail-low pruning.
By definition, plagiarism only happens when credit to sources is not
given, which was never the case with Rybka. These acknowledgements
were widely known at the time Rybka entered WCCC tournaments and,
rather obviously, Rybka’s debt to Fruit steadily diminished over the
years as program improvements superseded the original program code in
Rybka 1.0 Beta.
MikeMurray
Jan 3, 2012, 11:47:17 AM1/3/12
to
raylopez99
Jan 3, 2012, 11:48:49 AM1/3/12
to
On Jan 4, 12:47 am, MikeMurray <mikemur...@despammed.com> wrote:
> On Tue, 3 Jan 2012 08:44:19 -0800 (PST), raylopez99
>
> <raylope...@gmail.com> wrote:
> On Tue, 3 Jan 2012 08:44:19 -0800 (PST), raylopez99
>
>
> Good post.
Thanks Mike. Hope you have a good year chess wise and otherwise.
RL
raylopez99
Jan 3, 2012, 11:57:51 AM1/3/12
to
On Jan 4, 12:44 am, raylopez99 <raylope...@gmail.com> wrote:
Oops, I had to change the title: NOT guilty as charged. In other
words, Rybka is guilty.
It's also interesting in the second article linked to Chessbase that
the "Fruit derivatives" (including Houdini) all made rapid advances
while the "old stalwarts" like Craft and Fritz made only modest
improvements after Fruit was released as open source. What this means
(and speaking as a programmer myself) is that the old source code by
Hyatt (Crafty) and the Fritz programmer(s) was too 'baked in' to
radically change. Unless you write in an OOP (object oriented
programming) program language like C++ or C# or Java, which most chess
programmers do not (they prefer the 'closer to the machine' language
of C, which is not OOP) it is hard to change source code easily. So
the programmers who took advantage of the new algorithms introduced by
Fruit were the new guys, who wrote their programs from scratch, and
that included Rybka's author. The old guard, like Hyatt, could not
take advantage of Fruit's new algorithms without ripping up all the
years of work that they had done before. So they tinkered with their
code rather than rewrite it--and thus the real advantages of Fruit
fell to the new guard, like Rybka et al. That is clear from just
eyeballing the Elo vs time graph in the Chessbase link in the OP.
RL
Oops, I had to change the title: NOT guilty as charged. In other
words, Rybka is guilty.
It's also interesting in the second article linked to Chessbase that
the "Fruit derivatives" (including Houdini) all made rapid advances
while the "old stalwarts" like Craft and Fritz made only modest
improvements after Fruit was released as open source. What this means
(and speaking as a programmer myself) is that the old source code by
Hyatt (Crafty) and the Fritz programmer(s) was too 'baked in' to
radically change. Unless you write in an OOP (object oriented
programming) program language like C++ or C# or Java, which most chess
programmers do not (they prefer the 'closer to the machine' language
of C, which is not OOP) it is hard to change source code easily. So
the programmers who took advantage of the new algorithms introduced by
Fruit were the new guys, who wrote their programs from scratch, and
that included Rybka's author. The old guard, like Hyatt, could not
take advantage of Fruit's new algorithms without ripping up all the
years of work that they had done before. So they tinkered with their
code rather than rewrite it--and thus the real advantages of Fruit
fell to the new guard, like Rybka et al. That is clear from just
eyeballing the Elo vs time graph in the Chessbase link in the OP.
RL
Offramp
Jan 4, 2012, 4:44:51 AM1/4/12
to
I heard that Rybka was secretly on the telephone to a human.
Taylor Kingston
Jan 4, 2012, 1:49:47 PM1/4/12
to
On Jan 3, 8:47 am, MikeMurray <mikemur...@despammed.com> wrote:
>
> Good post.
http://chessbase.com/newsdetail.asp?newsid=7811
Today's installment (part 3 of 4) insists that the whole case
against Rajlich hinges on a single extraneous period (.). As a COBOL
programmer, I know how easy it is to type an extra period
accidentally, and what far-reaching consequences it can have.
>
> Good post.
http://chessbase.com/newsdetail.asp?newsid=7811
Today's installment (part 3 of 4) insists that the whole case
against Rajlich hinges on a single extraneous period (.). As a COBOL
programmer, I know how easy it is to type an extra period
accidentally, and what far-reaching consequences it can have.
Eric Hallsworth
Jan 5, 2012, 3:12:56 AM1/5/12
to
The ChessBase report is presented as an independent study by an
accredited expert, whereas Dr Rils is a Moderator on the Rybka forum no
less! And his study has been prepared with the assistance of the Rybka
author Vasik Rajlich. Hardly impartial at all.
It is interesting to see ChessBase refuting something they never even
reported on. Their "Gross Miscarriage of Justice...." articles say that
'it was widely reported on in the global media', but they never
mentioned it at all during the 6 months of suspicion, consideration,
ICGA trial and judgement! Not a word until now, 6 months later.
I wonder why?!
My guess is that they hope to distribute a Rybka 5 in the future.
Another guess. I note that they also failed to mention altogether
another major Computer Chess event, the Annual World Championship for
2011. This was won by Junior (WCC) and Hiarcs (WCSC). Therefore I doubt
if ChessBase will be distributing the forthcoming NEW Junior 13 and
Hiarcs 14 engines.
Best regards - Eric
In article <6ce8035d-a46e-4f36...@z25g2000vbs.googlegroup
s.com>, Offramp <alane...@gmail.com> writes
With best wishes from Eric Hallsworth
45 Stretham Road, Wilburton, Cambs CB6 3RX, England
Publisher of Selective Search, the UK's only Computer Chess Magazine, est. 1985.
web pages... http://www.elhchess.demon.co.uk
Also e-mail: erichal...@gmail.com
Our dream: ManunitedwithGod
accredited expert, whereas Dr Rils is a Moderator on the Rybka forum no
less! And his study has been prepared with the assistance of the Rybka
author Vasik Rajlich. Hardly impartial at all.
It is interesting to see ChessBase refuting something they never even
reported on. Their "Gross Miscarriage of Justice...." articles say that
'it was widely reported on in the global media', but they never
mentioned it at all during the 6 months of suspicion, consideration,
ICGA trial and judgement! Not a word until now, 6 months later.
I wonder why?!
My guess is that they hope to distribute a Rybka 5 in the future.
Another guess. I note that they also failed to mention altogether
another major Computer Chess event, the Annual World Championship for
2011. This was won by Junior (WCC) and Hiarcs (WCSC). Therefore I doubt
if ChessBase will be distributing the forthcoming NEW Junior 13 and
Hiarcs 14 engines.
Best regards - Eric
In article <6ce8035d-a46e-4f36...@z25g2000vbs.googlegroup
s.com>, Offramp <alane...@gmail.com> writes
>I heard that Rybka was secretly on the telephone to a human.
--
With best wishes from Eric Hallsworth
45 Stretham Road, Wilburton, Cambs CB6 3RX, England
Publisher of Selective Search, the UK's only Computer Chess Magazine, est. 1985.
web pages... http://www.elhchess.demon.co.uk
Also e-mail: erichal...@gmail.com
Our dream: ManunitedwithGod
MikeMurray
Jan 5, 2012, 7:42:55 PM1/5/12
to
Hallsworth's post in this thread resonates with me.
None
Jan 5, 2012, 8:21:36 PM1/5/12
to
On Jan 5, 7:42 pm, MikeMurray <mikemur...@despammed.com> wrote:
> > Today's installment (part 3 of 4) insists that the whole case
> >against Rajlich hinges on a single extraneous period (.). As a COBOL
> >programmer, I know how easy it is to type an extra period
> >accidentally, and what far-reaching consequences it can have.
>
> Everything I've read about this would indicate it's a more than that.
> Hallsworth's post in this thread resonates with me.
At what frequency?
> > Today's installment (part 3 of 4) insists that the whole case
> >against Rajlich hinges on a single extraneous period (.). As a COBOL
> >programmer, I know how easy it is to type an extra period
> >accidentally, and what far-reaching consequences it can have.
>
> Everything I've read about this would indicate it's a more than that.
> Hallsworth's post in this thread resonates with me.
Taylor Kingston
Jan 5, 2012, 8:15:00 PM1/5/12
to
On Jan 5, 4:42 pm, MikeMurray <mikemur...@despammed.com> wrote:
> On Wed, 4 Jan 2012 10:49:47 -0800 (PST), Taylor Kingston
>
> On Wed, 4 Jan 2012 10:49:47 -0800 (PST), Taylor Kingston
>
expert on this matter and have no definite opinion; I just wanted to
note that a single misplaced period/decimal can be /very/ significant
in a computer program.
MikeMurray
Jan 6, 2012, 9:36:45 AM1/6/12
to
On Thu, 5 Jan 2012 17:21:36 -0800 (PST), None <joesc...@gmail.com>
wrote:
I can't read the dial -- the vibrations you know.
wrote:
stan
Jan 6, 2012, 4:53:21 PM1/6/12
to
MikeMurray wrote:
> On Thu, 5 Jan 2012 17:21:36 -0800 (PST), None <joesc...@gmail.com> wrote:
>>On Jan 5, 7:42 pm, MikeMurray <mikemur...@despammed.com> wrote:
< snip >
> On Thu, 5 Jan 2012 17:21:36 -0800 (PST), None <joesc...@gmail.com> wrote:
>>On Jan 5, 7:42 pm, MikeMurray <mikemur...@despammed.com> wrote:
>>> Everything I've read about this would indicate it's a more than that.
>>> Hallsworth's post in this thread resonates with me.
>>
>>At what frequency?
>
> I can't read the dial -- the vibrations you know.
FWIW, the Riis articles are very clearly opinion and not even remotely
objective news reporting. The most damning point in the affair is that
the Rybka author made no defense - then or now. If Chessbase could get
the source to answer the charges then they would have some interesting
news, but on the other hand nothing short of source code will resolve
the issue and I'm not expecting a source release at this point.
Tony M
Jan 7, 2012, 5:00:15 PM1/7/12
to
On Tue, 3 Jan 2012 08:44:19 -0800 (PST), raylopez99
<raylo...@gmail.com> wrote:
>
>Why? Because being banned by the ICGA (the sponsor of computer chess
>tournaments) is NOT the same as being guilty of copyright
>infringement. It's clear that probably Rajlich is not guilty of
>copyright infringement, since his Rybka borrowed from open source
>"Fruit", as well as open source "Crafty", which are in the public
>domain.
Crafty and Fruit, although open source, are not public domain, and are
>
>Why? Because being banned by the ICGA (the sponsor of computer chess
>tournaments) is NOT the same as being guilty of copyright
>infringement. It's clear that probably Rajlich is not guilty of
>copyright infringement, since his Rybka borrowed from open source
>"Fruit", as well as open source "Crafty", which are in the public
>domain.
subject to copyright.
JA
Jan 7, 2012, 7:00:11 PM1/7/12
to
On 03/01/2012 4:44 PM, raylopez99 wrote:
> Rybka's
> Rajlich did NOT give credit to Fruit IN THE SUBMISSION DETAILS. He
> gave it publicly, but that is not the same. He has to say so in the
> submission details.
That sounds like a technical infringement to me - scarcely grounds for
> Rybka's
> Rajlich did NOT give credit to Fruit IN THE SUBMISSION DETAILS. He
> gave it publicly, but that is not the same. He has to say so in the
> submission details.
the opprobium being heaped on his head. Why was it necessary for a
committee of 34 experts to spend months reverse-engineering Rybka to
prove what was publicly admitted?
This whole affair smells of a witch-hunt. The fact is, Rybka was better
than the other engines, including Fruit. Whatever Rajlich got from
Fruit was evidently only a part of the story.
JA
raylopez99
Jan 10, 2012, 2:03:00 PM1/10/12
to
On Jan 8, 8:00 am, JA <apple3f...@gmail.com> wrote:
> On 03/01/2012 4:44 PM, raylopez99 wrote:
> > Rybka's
> > Rajlich did NOT give credit to Fruit IN THE SUBMISSION DETAILS. He
> > gave it publicly, but that is not the same. He has to say so in the
> > submission details.
>
> That sounds like a technical infringement to me - scarcely grounds for
> the opprobium being heaped on his head.
That is true. But rules are rules. If you write "WHY NOT HAVE NAZIS
> On 03/01/2012 4:44 PM, raylopez99 wrote:
> > Rybka's
> > Rajlich did NOT give credit to Fruit IN THE SUBMISSION DETAILS. He
> > gave it publicly, but that is not the same. He has to say so in the
> > submission details.
>
> That sounds like a technical infringement to me - scarcely grounds for
> the opprobium being heaped on his head.
KILL JEWS? IN WWII CIVILIANS WERE TARGETED, UNFORTUNATELY" in most
moderated forums, even as a joke, or as a statement of fact (indeed
civilians were targeted, and in Poland were most Jews were killed the
Polish lost more people, Jews and Gentiles alike, than any other
country proportionately, so you could argue that what happened to the
Polish Jews, while unfortunate, was not that unique by WWII standards
in Poland), you will be banned by the moderator. Unfair? Yes. But
those are the house rules. No mention of Nazis or the Holocaust--it's
verboten. Likewise here I agree--as did MORE THAN HALF THE VOTING
COMMITTEE OF THE ICGA (i.e., more than half voted for acquittal of
Rybka's author)---that the punishment did not fit the crime--since
most programmers borrow heavily. But it's not true that the ICGA
simply railroaded Rybka's author. He did borrow code--as did many
others--but you know what: speeding on US highways even by 1 mph is
still a technical infraction even if everybody around you is speeding
by a lot more.
> Why was it necessary for a
> committee of 34 experts to spend months reverse-engineering Rybka to
> prove what was publicly admitted?
>
my mind) that he lost the source code.
> This whole affair smells of a witch-hunt. The fact is, Rybka was better
> than the other engines, including Fruit. Whatever Rajlich got from
> Fruit was evidently only a part of the story.
that did not make his getting and failure to acknowledge any less of
an infraction. BTW it's funny that Fruit's author himself
participated in the witch hunt. I bet it's because Fruit's author had
regrets over releasing Fruit as open source and seeing people improve
on it, and make money off of it. That must be galling.
RL
raylopez99
Jan 10, 2012, 2:04:31 PM1/10/12
to
On Jan 5, 2:49 am, Taylor Kingston <ttk5...@gmail.com> wrote:
>
> http://chessbase.com/newsdetail.asp?newsid=7811
>
> Today's installment (part 3 of 4) insists that the whole case
> against Rajlich hinges on a single extraneous period (.). As a COBOL
> programmer, I know how easy it is to type an extra period
> accidentally, and what far-reaching consequences it can have.
I saw that as a strawman argument. And your analogy is inapt: in
>
> http://chessbase.com/newsdetail.asp?newsid=7811
>
> Today's installment (part 3 of 4) insists that the whole case
> against Rajlich hinges on a single extraneous period (.). As a COBOL
> programmer, I know how easy it is to type an extra period
> accidentally, and what far-reaching consequences it can have.
fact, the extra period--if you read the article you'd know this--had
NO effect, contrary to your analogy.
RL
0 new messages