A Question for Our Sponsor..IPPO Links OK or Not??

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Locked
User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 10:14 pm
Location: Munster, Nuremberg, Princeton

Re: Why always attacking the Champion? (Psychological questi

Post by Rolf » Mon Jun 28, 2010 2:47 pm

bob wrote: As far as "If he might be correct" I have looked at the code. I've already verified that this is a duck. No point in trying to waste any time to attempt to convince me it is a pig in a duck's suit.
Two or three years ago you wrote that you had NOT looked up the code in detail, alone because of the time that would have cost. Like Sven I ask you now if you claim having looked at what Zach had shown to you? If yes, then I can skip this too because it's just not correct science.

Also just as addition, Bob, you are so busy with nice metaphors, the pigs in duck's suit, so you need the personification of what you are talking about. But these anon jerks are totally ok for you although you dont even know their nature. Perhaps they are a herd of monkeys or kangoroos under LSD! :lol:
-Popper and Lakatos are good but I'm stuck on Leibowitz

bob
Posts: 20922
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Why always attacking the Champion? (Psychological questi

Post by bob » Mon Jun 28, 2010 4:15 pm

Sven Schüle wrote:
bob wrote:As far as "If he might be correct" I have looked at the code. I've already verified that this is a duck. No point in trying to waste any time to attempt to convince me it is a pig in a duck's suit.
Just to confirm: does that mean you have looked in detail into the disassembled R1 binary?

In case you have, what have you seen when comparing that to the Fruit 2.1 source? Which parts of R1 did you look at: only those listed by Zach, or more (which)?

In case you haven't, and I misunderstood your remark, what else did you mean by "I have looked at the code"?

Sven
I looked at what Zach has posted. I looked at other small pieces when this first started. As I have said, previously, _any_ copied code would be a no-no when the author says there was _no_ copied code. To disprove never, you need one example. To disprove all, you need a near-infinite number of examples. Here we are working on "none" where most are twisting it into "all".

bob
Posts: 20922
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Why always attacking the Champion? (Psychological questi

Post by bob » Mon Jun 28, 2010 4:19 pm

Rolf wrote:
bob wrote: As far as "If he might be correct" I have looked at the code. I've already verified that this is a duck. No point in trying to waste any time to attempt to convince me it is a pig in a duck's suit.
Two or three years ago you wrote that you had NOT looked up the code in detail, alone because of the time that would have cost. Like Sven I ask you now if you claim having looked at what Zach had shown to you? If yes, then I can skip this too because it's just not correct science.

Also just as addition, Bob, you are so busy with nice metaphors, the pigs in duck's suit, so you need the personification of what you are talking about. But these anon jerks are totally ok for you although you dont even know their nature. Perhaps they are a herd of monkeys or kangoroos under LSD! :lol:
We looked at specific code as a group. All it takes is _one_ example of copied code to settle the issue, and there are several examples. Doesn't matter how trivial you think they are, how unimportant you think they are. When someone says "I copied _no_ code" and you can show they copied _some_ code, that is enough. Yes, a group could go thru the entire binary, convert it back to C and compare to fruiit. But does it matter _how_ guilty one is? I've never seen that in a court. Just "guilty" or "innocent". That was the intent here. Nobody would want to expend the effort to compare the entire binary, when examples of copied code have already been discovered. What part of that is so hard to understand?

Sven
Posts: 3902
Joined: Thu May 15, 2008 7:57 pm
Location: Berlin, Germany
Full name: Sven Schüle
Contact:

Re: Why always attacking the Champion? (Psychological questi

Post by Sven » Mon Jun 28, 2010 5:44 pm

bob wrote:
Sven Schüle wrote:
bob wrote:As far as "If he might be correct" I have looked at the code. I've already verified that this is a duck. No point in trying to waste any time to attempt to convince me it is a pig in a duck's suit.
Just to confirm: does that mean you have looked in detail into the disassembled R1 binary?

In case you have, what have you seen when comparing that to the Fruit 2.1 source? Which parts of R1 did you look at: only those listed by Zach, or more (which)?

In case you haven't, and I misunderstood your remark, what else did you mean by "I have looked at the code"?

Sven
I looked at what Zach has posted. I looked at other small pieces when this first started.
So what Zach has posted is "the code" for you. Up to now I had thought you had participated in the detailled analysis itself, I was mistaken in assuming that. You verified his publication by looking at his publication. Good. Let's assume this is science. "Looks like a duck - is a duck."

But since you know the Zach examples, which of these points would serve "best" from your viewpoint to prove that code was literally copied from Fruit 2.1 to Rybka 1.0 beta? Just one example is sufficient in the beginning.

I'll help you with the link to Zach's comparison.

Maybe Zach can provide the related disassembled code (not published yet) after you made your first choice.

As an aside, please don't get me wrong (this applies to all readers here of course): despite what someone has written in this forum recently about me, I do *not* say "Zach is a liar". That is completely false. His work regarding Fruit-R1 comparison is surely an important and impressive work. What I am attacking is *neither* the person *nor* his capabilities, both is foreign to my nature. I am addressing *only* the results in this given case, and also the way they are interpreted by you.

So let's start, play the first ball ... (I'm serious about this!)
bob wrote:As I have said, previously, _any_ copied code would be a no-no when the author says there was _no_ copied code. To disprove never, you need one example. To disprove all, you need a near-infinite number of examples. Here we are working on "none" where most are twisting it into "all".
I'll comment on this part later on.

Sven

Terry McCracken
Posts: 16464
Joined: Wed Aug 01, 2007 2:16 am
Location: Canada

Re: Why always attacking the Champion? (Psychological questi

Post by Terry McCracken » Mon Jun 28, 2010 5:48 pm

bob wrote:
Rolf wrote:
bob wrote: As far as "If he might be correct" I have looked at the code. I've already verified that this is a duck. No point in trying to waste any time to attempt to convince me it is a pig in a duck's suit.
Two or three years ago you wrote that you had NOT looked up the code in detail, alone because of the time that would have cost. Like Sven I ask you now if you claim having looked at what Zach had shown to you? If yes, then I can skip this too because it's just not correct science.

Also just as addition, Bob, you are so busy with nice metaphors, the pigs in duck's suit, so you need the personification of what you are talking about. But these anon jerks are totally ok for you although you dont even know their nature. Perhaps they are a herd of monkeys or kangoroos under LSD! :lol:
We looked at specific code as a group. All it takes is _one_ example of copied code to settle the issue, and there are several examples. Doesn't matter how trivial you think they are, how unimportant you think they are. When someone says "I copied _no_ code" and you can show they copied _some_ code, that is enough. Yes, a group could go thru the entire binary, convert it back to C and compare to fruiit. But does it matter _how_ guilty one is? I've never seen that in a court. Just "guilty" or "innocent". That was the intent here. Nobody would want to expend the effort to compare the entire binary, when examples of copied code have already been discovered. What part of that is so hard to understand?
Why do you bother explaining yourself to Rolf? What do you gain from it?
I actually took the time in the past to read his diatribes but now I think what's the point? It's the same thing over and over and over again.
Terry McCracken

bob
Posts: 20922
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Why always attacking the Champion? (Psychological questi

Post by bob » Mon Jun 28, 2010 6:03 pm

Sven Schüle wrote:
bob wrote:
Sven Schüle wrote:
bob wrote:As far as "If he might be correct" I have looked at the code. I've already verified that this is a duck. No point in trying to waste any time to attempt to convince me it is a pig in a duck's suit.
Just to confirm: does that mean you have looked in detail into the disassembled R1 binary?

In case you have, what have you seen when comparing that to the Fruit 2.1 source? Which parts of R1 did you look at: only those listed by Zach, or more (which)?

In case you haven't, and I misunderstood your remark, what else did you mean by "I have looked at the code"?

Sven
I looked at what Zach has posted. I looked at other small pieces when this first started.
So what Zach has posted is "the code" for you. Up to now I had thought you had participated in the detailled analysis itself, I was mistaken in assuming that
Please stop trying to imagine what I did. We had a _ton_ of "offline_ discussions via email. Zach, Theron, and I don't remember who else. _THAT_ is the stuff I looked it, not just what he released on his web page so far...



. You verified his publication by looking at his publication. Good. Let's assume this is science. "Looks like a duck - is a duck."
Actually, how about not "let's assume" anything? We tried to discuss this on CCC. Got way too noisy so we took it offline for quite a while. That was what convinced me something was amiss. Clear enough, now? I have said this plenty of times in the past, by the way, so it isn't exactly "headline news."


[quote

But since you know the Zach examples, which of these points would serve "best" from your viewpoint to prove that code was literally copied from Fruit 2.1 to Rybka 1.0 beta? Just one example is sufficient in the beginning.[/quote]

The first one that jumped out at me when we started the process was the code segment containing the setjmp()/longjmp() construct. You can find references to this in the past. It is an unusual way to unwind a search that is not thread-happy, and invites very subtle bugs. The usage was identical in both programs with the surrounding code. There were others.

I'll help you with the link to Zach's comparison.

Maybe Zach can provide the related disassembled code (not published yet) after you made your first choice.

As an aside, please don't get me wrong (this applies to all readers here of course): despite what someone has written in this forum recently about me, I do *not* say "Zach is a liar". That is completely false. His work regarding Fruit-R1 comparison is surely an important and impressive work. What I am attacking is *neither* the person *nor* his capabilities, both is foreign to my nature. I am addressing *only* the results in this given case, and also the way they are interpreted by you.

So let's start, play the first ball ... (I'm serious about this!)
bob wrote:As I have said, previously, _any_ copied code would be a no-no when the author says there was _no_ copied code. To disprove never, you need one example. To disprove all, you need a near-infinite number of examples. Here we are working on "none" where most are twisting it into "all".
I'll comment on this part later on.

Sven

Sven
Posts: 3902
Joined: Thu May 15, 2008 7:57 pm
Location: Berlin, Germany
Full name: Sven Schüle
Contact:

Re: Why always attacking the Champion? (Psychological questi

Post by Sven » Mon Jun 28, 2010 7:21 pm

bob wrote:
Sven Schüle wrote:
bob wrote:
Sven Schüle wrote:
bob wrote:As far as "If he might be correct" I have looked at the code. I've already verified that this is a duck. No point in trying to waste any time to attempt to convince me it is a pig in a duck's suit.
Just to confirm: does that mean you have looked in detail into the disassembled R1 binary?

In case you have, what have you seen when comparing that to the Fruit 2.1 source? Which parts of R1 did you look at: only those listed by Zach, or more (which)?

In case you haven't, and I misunderstood your remark, what else did you mean by "I have looked at the code"?
I looked at what Zach has posted. I looked at other small pieces when this first started.
So what Zach has posted is "the code" for you. Up to now I had thought you had participated in the detailled analysis itself, I was mistaken in assuming that
Please stop trying to imagine what I did. We had a _ton_ of "offline_ discussions via email. Zach, Theron, and I don't remember who else. _THAT_ is the stuff I looked it, not just what he released on his web page so far...
You can calm down again, no need to get angry. I'm not writing BS. I just show how you say things differently from time to time (see above) :-)

If that is what you looked at then you have seen more than what is currently known to the public. But as long as you insist that the proof is already there for everyone, we can only consider the "public" part, not what has been discussed privately. You could not claim "there is the proof for everyone" in one second, but then say that the important part were non-public. I know you don't, so our current base is what is published for everyone to compare. Agreed? ;-)
bob wrote:
You verified his publication by looking at his publication. Good. Let's assume this is science. "Looks like a duck - is a duck."
Actually, how about not "let's assume" anything? We tried to discuss this on CCC. Got way too noisy so we took it offline for quite a while. That was what convinced me something was amiss. Clear enough, now? I have said this plenty of times in the past, by the way, so it isn't exactly "headline news."
You know as well as I do how my sentence "Let's assume this is science." was meant. One of my points is still that I don't know how much you have really looked into the assembler code of R1. You may have written a lot around that topic but I can't recall you ever made *this* clear enough.
bob wrote:
But since you know the Zach examples, which of these points would serve "best" from your viewpoint to prove that code was literally copied from Fruit 2.1 to Rybka 1.0 beta? Just one example is sufficient in the beginning.
The first one that jumped out at me when we started the process was the code segment containing the setjmp()/longjmp() construct. You can find references to this in the past. It is an unusual way to unwind a search that is not thread-happy, and invites very subtle bugs. The usage was identical in both programs with the surrounding code. There were others.
I remember that we had some PMs about this setjmp() usage. My opinion has not changed much since then. There are not many possible ways how setjmp()/longjmp() can be used to break from an arbitrary level of recursion and return to a defined point of control by unwinding the stack. No need to explain this to you, just for everyone else: there is always exactly one "setjmp()" call that defines the point to return to, and one or more places where calling longjmp() causes that unwinding, e.g. to interrupt the search. Not many choices to implement this basic technique. And no, it is *not* unusual to do so, even if *you* claim it. I accept the note about that concept being potentially error-prone but this is what a software developer can be able to deal with. It *can* be implemented safely under certain conditions (single-threaded may be one), and when doing so it also slightly improves readability of the search code itself since you do not need "if (search is stopped)" conditions here and there. So using setjmp()/longjmp() like another program does is *not* an indication of code copying, it is reusing an idea at best.

Your turn again? I think you did not start with the "best" candidate but with the first that came into mind.

Sven

P.S.: Btw, do you know this book? Have a look at the top of page 3 of chapter one. “transformative derivative work” - didn't know that term up to now. Any change after that?

bob
Posts: 20922
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Why always attacking the Champion? (Psychological questi

Post by bob » Mon Jun 28, 2010 7:53 pm

Sven Schüle wrote:
bob wrote:
Sven Schüle wrote:
bob wrote:
Sven Schüle wrote:
bob wrote:As far as "If he might be correct" I have looked at the code. I've already verified that this is a duck. No point in trying to waste any time to attempt to convince me it is a pig in a duck's suit.
Just to confirm: does that mean you have looked in detail into the disassembled R1 binary?

In case you have, what have you seen when comparing that to the Fruit 2.1 source? Which parts of R1 did you look at: only those listed by Zach, or more (which)?

In case you haven't, and I misunderstood your remark, what else did you mean by "I have looked at the code"?
I looked at what Zach has posted. I looked at other small pieces when this first started.
So what Zach has posted is "the code" for you. Up to now I had thought you had participated in the detailled analysis itself, I was mistaken in assuming that
Please stop trying to imagine what I did. We had a _ton_ of "offline_ discussions via email. Zach, Theron, and I don't remember who else. _THAT_ is the stuff I looked it, not just what he released on his web page so far...
You can calm down again, no need to get angry. I'm not writing BS. I just show how you say things differently from time to time (see above) :-)

If that is what you looked at then you have seen more than what is currently known to the public. But as long as you insist that the proof is already there for everyone, we can only consider the "public" part, not what has been discussed privately. You could not claim "there is the proof for everyone" in one second, but then say that the important part were non-public. I know you don't, so our current base is what is published for everyone to compare. Agreed? ;-)
Not quite. You asked "what specific example caught your eye?" I replied. I looked at the rest of Zach's analysis and consider the entire kit and kaboodle quite troubling, because it shows behaviour that is not acceptable (copying code and claiming at is own-code).

The current way some try to refute this evidence is to pick some piece that has only gotten circumspect analysis, and then finding flaws with that analysis and claiming that since that is wrong, _everything_ is wrong. Some of the analysis is in-depth. Some is less precise. It is the entire body of evidence that should be looked at, because it is quite large.
bob wrote:
You verified his publication by looking at his publication. Good. Let's assume this is science. "Looks like a duck - is a duck."
Actually, how about not "let's assume" anything? We tried to discuss this on CCC. Got way too noisy so we took it offline for quite a while. That was what convinced me something was amiss. Clear enough, now? I have said this plenty of times in the past, by the way, so it isn't exactly "headline news."
You know as well as I do how my sentence "Let's assume this is science." was meant. One of my points is still that I don't know how much you have really looked into the assembler code of R1. You may have written a lot around that topic but I can't recall you ever made *this* clear enough.
bob wrote:
But since you know the Zach examples, which of these points would serve "best" from your viewpoint to prove that code was literally copied from Fruit 2.1 to Rybka 1.0 beta? Just one example is sufficient in the beginning.
The first one that jumped out at me when we started the process was the code segment containing the setjmp()/longjmp() construct. You can find references to this in the past. It is an unusual way to unwind a search that is not thread-happy, and invites very subtle bugs. The usage was identical in both programs with the surrounding code. There were others.
I remember that we had some PMs about this setjmp() usage. My opinion has not changed much since then. There are not many possible ways how setjmp()/longjmp() can be used to break from an arbitrary level of recursion and return to a defined point of control by unwinding the stack. No need to explain this to you, just for everyone else: there is always exactly one "setjmp()" call that defines the point to return to, and one or more places where calling longjmp() causes that unwinding, e.g. to interrupt the search. Not many choices to implement this basic technique. And no, it is *not* unusual to do so, even if *you* claim it. I accept the note about that concept being potentially error-prone but this is what a software developer can be able to deal with. It *can* be implemented safely under certain conditions (single-threaded may be one), and when doing so it also slightly improves readability of the search code itself since you do not need "if (search is stopped)" conditions here and there. So using setjmp()/longjmp() like another program does is *not* an indication of code copying, it is reusing an idea at best.
What if you notice the setjmp()/longjmp() stuff, and then when you look at the surrounding code, you notice, "damn, this is done here and there, this is done here and there, ..." for quite a bit. How many programmers use this in their chess engine? It is _highly_ problematic for a parallel search. It is problematic for a program that doesn't do copy/make. It is problematic, period. And you might check around to see if you can find one that uses this... The last time this came up I looked at several programs and found it only in fruit and derivatives at the time. I just checked my cluster opponents and found it only in Fruit/toga. Not in arasan, glaurung, stockfish, gnuchessx, old version of Zappa I was sent, etc. What does that say? Not a lot, taken by itself, but it is another nail one can see.

Your turn again? I think you did not start with the "best" candidate but with the first that came into mind.

Sven

P.S.: Btw, do you know this book? Have a look at the top of page 3 of chapter one. “transformative derivative work” - didn't know that term up to now. Any change after that?
Don't see where that applies here. It is not so "fundamentally transformed that it becomes a new work" when you can find duplicate blocks of code...

User avatar
Rolf
Posts: 6081
Joined: Fri Mar 10, 2006 10:14 pm
Location: Munster, Nuremberg, Princeton

Re: Why always attacking the Champion? (Psychological questi

Post by Rolf » Mon Jun 28, 2010 8:22 pm

bob wrote:
Rolf wrote:
bob wrote: As far as "If he might be correct" I have looked at the code. I've already verified that this is a duck. No point in trying to waste any time to attempt to convince me it is a pig in a duck's suit.
Two or three years ago you wrote that you had NOT looked up the code in detail, alone because of the time that would have cost. Like Sven I ask you now if you claim having looked at what Zach had shown to you? If yes, then I can skip this too because it's just not correct science.

Also just as addition, Bob, you are so busy with nice metaphors, the pigs in duck's suit, so you need the personification of what you are talking about. But these anon jerks are totally ok for you although you dont even know their nature. Perhaps they are a herd of monkeys or kangoroos under LSD! :lol:
We looked at specific code as a group. All it takes is _one_ example of copied code to settle the issue, and there are several examples. Doesn't matter how trivial you think they are, how unimportant you think they are. When someone says "I copied _no_ code" and you can show they copied _some_ code, that is enough. Yes, a group could go thru the entire binary, convert it back to C and compare to fruiit. But does it matter _how_ guilty one is? I've never seen that in a court. Just "guilty" or "innocent". That was the intent here. Nobody would want to expend the effort to compare the entire binary, when examples of copied code have already been discovered. What part of that is so hard to understand?
Yes, I dont accept your theory. Vas never said what you claimed here. Because he openly admitted that he took some public domain stuff. Dont remember that? So that we have the question what you have seen. What sort of code. Science is a bit mor differentiated than what you are presenting here. So, if I would be bob Hyatt or Theron I could now state that you have a lack of ethics because you tell the untruth about the examined code. But my best argument is still that you relied on what other had found and then presented to you.

In all psychological studies it's well known that it's always a fallacy putting three big experts into a box where they talk about their convictions in a certain problem. They will fool each other but personally they are sure that they had found the truth. This is called folie a trois. LOL

Jens and I are in a better position. We dont know each other but independantly we came to the same conclusions, well, me without all the tech details, so that our agreement is the proof that we are right. :-)

QED

P.S. Still, Bob your honest description of how you buddied together is worth our whole debate. I knew it all the time that you have been caught by delusional convictions without the scientifically necessary scepticism. Well, IMO I am in principle in such situations immune against such effects because I would only rely on myself. But I know exactly that I have no tech expertise in CC. Still, although you are totally above me I knew for sure that you three were wrong. I knew it. All three of you are too much depending on prejudices.
-Popper and Lakatos are good but I'm stuck on Leibowitz

Milos
Posts: 3987
Joined: Wed Nov 25, 2009 12:47 am

Re: Why always attacking the Champion? (Psychological questi

Post by Milos » Tue Jun 29, 2010 12:32 am

bob wrote:Don't see where that applies here. It is not so "fundamentally transformed that it becomes a new work" when you can find duplicate blocks of code...
However, it perfectly applies to Ippo/Rybka issue.
Ippo is exactly fundamentally transformed that it becomes a new work and Vas as the holder of the copyright of the work—from which such a “transformative derivative work” is derived—has absolutely no rights over it and of course he knows it. This perfectly explains his behavior.

Locked