Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Asides and commentary on RV's decompilation and reports
1 2 3 4 512 13 Previous Next  
Parent - - By Banned for Life (Gold) Date 2013-08-18 22:13
The contents are "odd" to say the least. Particularly when he had previously used a more normal entry in the crafty derivatives...

The contents are not "odd". The contents are different than Crafty's because Vas compared the two concepts and liked Fruit's better. Most other engines are doing the same now.

So he copied my TT code, then he copied Fruits.

This is nonsense. He did not copy either Crafty or Fruit code in any distributed software. Saying it thousands of times doesn't make it true.

The code does not come "from the table".  The table format is driven by the code. 

Baloney. Vas decided to use the Fruit hash table concept, but made modifications already noted by Richard, where the location affected the tests to be performed.

Weren't you the one claiming Vas had developed some super-clever hashing stuff?  Where is that, exactly?  Or did you mean Fabien?

You are even stupider than you look. You do realize that Rybka has been using 8 bytes per TT record for many years now, right? Where is it? I'll give you a hint. It's in the code that actually played in most of the ICGA events. Not knowing this after doing an investigation immediately qualifies you for moron status. Congratulations!
Parent - - By bob (Gold) Date 2013-08-18 22:33
Some simple questions to answer.  (1) why two different value fields?  A TT entry can be stored with an EXACT score, a LOWER bound, or an UPPER bound.  Not all 3.  Why 4 depth fields?  A basic depth.  A depth for the best move.  A depth for each of the two scores.

(2) If he didn't copy Fruit's tt code, why is it matching Fruit's so exactly?  Richard posted the r1b tt core.  Fruit's is there for all to see in the fruit source.  I have actually compared them.  Do you need that "side by side" to grasp?  That can be arranged.

My question concerned your claim, at EVERY turn in the investigation, that Vas had done this or that originally.  Richard has already pointed out that Vas uses Crafty's rotated bitboard indexing scheme, including the 64x64 trick that was developed at the end of 2003 / early Jan 2004 and released in Crafty 19.9.  If his code is original, he just happened to have wrote exactly what I wrote?  You have repeatedly claimed he had a "very intelligent" hash table implementation.  Yet 1.0beta is pure fruit, front to back.  This list is growing over time, because the code is there for anyone
to RE, given enough time and mental energy.  You guys asked Richard to jump in.  You will likely regret that request.  I'm not communicating with him at all unless it is here.  I'm waiting to see what he shows next myself...  Maybe it will be something that is different from Fruit, for a change?

The only modifications I saw Richard mentioned was the minor optimization where he splits store up into 4 different versions, one to store a bound, or an exact score.  That's not exactly "clean" programming, BTW.  I suspect there is more to be seen.

Feel free to now resort to your gratuitous name-calling...

BTW he did copy my code in a version of Rybka he distributed AND claimed to be his original work.  Anyone can take a disassembler and see that easy enough...  Your saying he didn't, trying to distort that "distributed" word doesn't fly.  He SENT it to Olivier to enter in a tournament.  That's distributing it.
Parent - - By Banned for Life (Gold) Date 2013-08-18 22:56
(2) If he didn't copy Fruit's tt code, why is it matching Fruit's so exactly?  Richard posted the r1b tt core.  Fruit's is there for all to see in the fruit source.  I have actually compared them.  Do you need that "side by side" to grasp?  That can be arranged.

Richard points out that Vas' TT code is customized based on it's location, unlike other implementations. It should also be noted that the TT code is very simple stuff.

My question concerned your claim, at EVERY turn in the investigation, that Vas had done this or that originally.

I never made this bogus claim. I claimed that Vas wrote his own code after studying other implementations, especially Fruit.

Richard has already pointed out that Vas uses Crafty's rotated bitboard indexing scheme, including the 64x64 trick that was developed at the end of 2003 / early Jan 2004 and released in Crafty 19.9.

Anybody can use this scheme. It's public domain. Get over it. Code is subject to copyright, indexing tables aren't. If they were, the people that came up with the indexing schemes used for trellis filtering would be billionaires. They don't even get a thank you.

The important thing is that Vas wrote his own bitboard code, and the fact that you keep babbling on about the indexing array shows that you know that he wrote his own code.

You have repeatedly claimed he had a "very intelligent" hash table implementation.  Yet 1.0beta is pure fruit, front to back.

Vas has been using 8-bytes per hash record for many years now. This was used in at least most (maybe all) of the ICGA events. You are the moron who overlooked this. Let me remind you that Rybka 1.0 Beta never competed in the WCCC.

You guys asked Richard to jump in.  You will likely regret that request.  I'm not communicating with him at all unless it is here.  I'm waiting to see what he shows next myself...  Maybe it will be something that is different from Fruit, for a change?

Nobody will regret Richard's assistance. Unlike you, Richard is both competent and honest.

The only modifications I saw Richard mentioned was the minor optimization where he splits store up into 4 different versions, one to store a bound, or an exact score.  That's not exactly "clean" programming, BTW.  I suspect there is more to be seen.

This isn't new. Anyone who looked at Strelka 2 back in 2008 noted that Rybka 1.0 Beta used a Fruit-like TT implementation. This was changed to a more compact representation (8-bytes per record) shortly thereafter.

BTW he did copy my code in a version of Rybka he distributed AND claimed to be his original work.

Wrong. He did not distribute Rybka prior to Rybka 1.0 Beta.

He SENT it to Olivier to enter in a tournament.  That's distributing it.

Wrong. Once again, you are showing you are ignorant of the meaning of distribute. Sending to Olivier for his basement tournament doesn't meet the bar. Not even close.
Parent - - By cipri (**) Date 2013-08-18 23:08
He SENT it to Olivier to enter in a tournament.  That's distributing it.

Wrong. Once again, you are showing you are ignorant of the meaning of distribute. Sending to Olivier for his basement tournament doesn't meet the bar. Not even close.


I guess, here the dictionary can help.

For the ones interested in solving such problems, the great mathematician wrote the book series: Mathematics and Plausible Reasoning
It helps you solving such every day tasks.

I wished, all these discussions would be moved out of this thread, and just the posts related to the ones of richard. He took the time to contribute to this problem, and now we are spamming this thread.
Parent - - By Banned for Life (Gold) Date 2013-08-18 23:36
I guess, here the dictionary can help.

Not unless it's a legal dictionary, but even that won't help because the Copyright Act doesn't define distribution. Instead distribution has been defined by case law. See for example: http://blog.internetcases.com/2008/04/30/distribution-under-the-copyright-act-requires-more-than-merely-making-copies-available/ Of course this is for the US...

I wished, all these discussions would be moved out of this thread, and just the posts related to the ones of richard. He took the time to contribute to this problem, and now we are spamming this thread.

That would be a good thing to do, but it will last only as long as Bob can control his urge to stick his thumb on the scale.
Parent - - By nebulus (****) Date 2013-08-19 13:11
It's rather simple... you put something online and no one downloads it - there's no distribution, but if someone does - there's a distribution.

Here you have a case of someone actually receiving a copy. A modified copy of a program has been transferred from one person to another, ergo it's a distribution.
Parent - - By Ugh (*****) Date 2013-08-19 13:39
Transfer is a necessary but not sufficient condition for distribution, it also depends on the reason for the transfer and the relationship between the transferrer and transferee.

Imagine a software house where programmers are coding a product. At various stages they transfer the "product" to someone in quality control for testing. These are transfers, but they are not distributions.

It is not a distribution to transfer a product to a buying department of a retailer or distributor for them to assess whether or not to stock the product.

It is not a distribution to transfer a product to a magazine for purposes of creating and printing a public review.

Clearly, in software, distribution is a phase of bringing a product to market. There is production, marketing, promotion, distribution, sale, etc etc etc. And these terms need to be considered and defined in this context.

The question is whether transferring a copy to a private tester, whether or not that tester makes his results public or not, is "distribution" under this analysis. I would say not. If anything at all, it would come under "marketing" or "promotion". And since the thing was never brought to market anyway, and was not intended for market, it is all a bit moot.

Notwithstanding all that, it was in breach of the Hyatt licence if it was a development from Crafty. But it was not a "distribution".
Parent - - By nebulus (****) Date 2013-08-19 14:05
I agree that this depends on the relationship between the transferrer and transferee. Nowdays, "internal" departments may not necessarily be the same company. If it's not the same legal entity, chances are this is a distribution and not merely a transfer or internal distribution. Also worth mentioning a quote from GPL FAQ: "However, when the organization transfers copies to other organizations or individuals, that is distribution. In particular, providing copies to contractors for use off-site is distribution."

> Notwithstanding all that, it was in breach of the Hyatt licence if it was a development from Crafty. But it was not a "distribution".


I don't agree with this. It was a one count of distribution.
Parent - - By Ugh (*****) Date 2013-08-19 14:39
Hyatt's was not a GPL licence and it neither uses nor defines the word "distribution".

Therefore, in this case, one should stick with defining the breach for what it was, the "you shall not enter into tournament bla bla". The introduction of a new, undefined term, "distribution", done by Hyatt serves only to confuse and obfuscate, and is clearly an attempt to widen the supposed infraction by association with term normally used for something else (like Chessbase selling Houdinis, for example).

It's a deliberate creative use of language to implant a fantasy infraction into the minds of readers. Then use the same word, "distribution", in it's normal marketing context, spinning a kind of dishonesty web around the target. You didn't work out how Hyatt operates yet? He has done the same thing with "semantic equivalence" and "copy". These are now used interchangeably by many people in these threads who didn't notice that they are not equal. Hyatt has used them as synonyms now, so many times, that even his opponents are using them the same way. Goebbelesque groupthink.
Parent - - By nebulus (****) Date 2013-08-19 15:16
I'm perfectly aware that Crafty doesn't use GPL. I was merely pointing out that it was worth mentioning that FSF, who has a record of dealing with lawyers and defending the GPL in court, has a seemingly different opinion on the matter.
Parent - - By Ugh (*****) Date 2013-08-19 15:29
Ah, ok. What the FSF are actually doing is putting down a marker to define or make clear how they, the FSF, intend that the word, distribution, be defined in relation to their contract.

Any contract can define a word, or clarify use of the word within that contract in any way it wants. Often it is done for clarity and to remove ambiguity (courts hate ambiguity), or also it is done to deliberately widen or change the definition to the advantage of one side. Even to slip a meaning past the other side's lawyers if they are not sharp enough.

But what we can't do, is pull a word, defined in a contract, out of the contract and apply that definition elsewhere.

If there is no definition in the applicable contract, then the two sides are down to arguing out the meaning. And I do think, know almost, that the definition I've given for this word, in the case where it was not defined here, is correct.

Basically:

http://en.wikipedia.org/wiki/Distribution_(business)

http://en.wikipedia.org/wiki/Marketing_mix
Parent - - By nebulus (****) Date 2013-08-19 16:00
First of all, it's license, not a contract. It may and may not be considered as such, this obviously depends on local law. Secondly, it isn't defined in the license, so I didn't take anything out of the context.

While the FAQ is merely a guideline, I would assume they (FSF) know better, since they have been dealing with the legal issues regarding the definition of "distribution". BTW, the reason why they introduced "convey" in GPL3 was because they found out that in some jurisdictions where the term "distribution" was actually defined, the legal definition (in those jurisdictions) didn't necessary match with the intended meaning.
Parent - - By Ugh (*****) Date 2013-08-19 16:44
I didn't say you did do anything. Apologies if my writing style led you to think that.
Parent - By nebulus (****) Date 2013-08-19 23:42
No worries. I only intended to emphasize this, not to critique, argue or otherwise. No apology was necessary.
Parent - - By bob (Gold) Date 2013-08-19 18:18
Can we stop with the manufacturing stuff?

Vas and Olivier had no employee/employer relationship.  Olivier ran a tournament he organized and controlled, Chess Wars.  Vas was a soon-to-be commercial engine author that simply entered his program in an event that required original programs only.  No way you can wrangle that into "not distributing it."  Play with words all you want.  But the receiver had ZERO ralationship with the sender.  The sender violated MY copyright / EULA.  And he violated the Chess Wars rules.

It doesn't get any simpler than that.  Olivier is not a "private tester".  He ran public events and solicited participants.  You are trying to somehow imagine that this was an "in house" deal where it might be acceptable.  It was not.

It was most definitely a distribution.  There is no "minimum quantity" required to call something a distribution.  My IRA will eventually "distribute" my IRA accumulations to me and only to me.  It might be a one-time distribution, a monthly distribution, or anything I choose.
Parent - - By Banned for Life (Gold) Date 2013-08-19 19:52
It was most definitely a distribution.  There is no "minimum quantity" required to call something a distribution.  My IRA will eventually "distribute" my IRA accumulations to me and only to me.  It might be a one-time distribution, a monthly distribution, or anything I choose.

Wow! Equating a distribution from a retirement fund with a distribution of copyright material places you in the top ranks of the world's elite super morons...
Parent - - By bob (Gold) Date 2013-08-19 19:58
I suppose that is an assessment you are qualified for, since you show "moronic" behavior all the time.

You want to wrangle around and try to artificially redefine "distribution" to mean something else.  The EULA specifically forbids his behavior.  Unequivocally.  Trying to redefine the word "distribution" does not help his case one scintilla.
Parent - - By Banned for Life (Gold) Date 2013-08-19 20:44
You want to wrangle around and try to artificially redefine "distribution" to mean something else.

Wrong. Distribution was defined for copyright by the United States Court of Appeals, Fourth Circuit, 30 June 1997. Not by any EULA.

The EULA specifically forbids his behavior.  Unequivocally.  Trying to redefine the word "distribution" does not help his case one scintilla.

Your EULA isn't worth the paper it's printed on. It's totally unenforceable and there is no reason for discussing it in any context other than comic relief.
Parent - - By nebulus (****) Date 2013-08-20 03:15
The only part of the license that I believe is unenforceable is: "The author retains sole ownership and copyright on this program except for 'personal use' explained below." [ed. note: this seems to be directed towards a derivative work]

Anyhow, if the license is void Vas wouldn't have any rights to use, modify etc., etc. the program.
Parent - - By Banned for Life (Gold) Date 2013-08-20 04:10
The only part of the license that I believe is unenforceable is: "The author retains sole ownership and copyright on this program except for 'personal use' explained below." [ed. note: this seems to be directed towards a derivative work]

There are very strict requirements for EULAs here in the US and you will note the care put into these documents if you read through one from Adobe or Microsoft or Oracle before you check the box agreeing to the terms. Rest assured that none of these companies will be working with Bob to generate their next EULA. There are way too many things wrong with this to even start explaining. I have sent it along to a few friends so that they can get a good chuckle. Maybe that was the intent.

Anyhow, if the license is void Vas wouldn't have any rights to use, modify etc., etc. the program.

It doesn't work that way. Vas can do anything he wants with the program, with the caveat that Bob can enforce his copyright, and could try to enforce his EULA.
Parent - - By nebulus (****) Date 2013-08-20 05:14

> It doesn't work that way. Vas can do anything he wants with the program, with the caveat that Bob can enforce his copyright, and could try to enforce his EULA.


Not at all: 17 USC § 106.
Parent - By Banned for Life (Gold) Date 2013-08-20 05:26
As I said, Bob can enforce his copyright and could amuse people by trying to enforce his EULA.

Vas can do anything he wants with the code subject to the above. As a practical matter, anything that Vas does in his basement will be perfectly fine. We all acknowledge that Vas shouldn't have used Bob's code in Olivier's tournaments, but it's highly unlikely that Bob could collect any damages from enforcement.

In any event, the problem is solved, at least as far as Crafty is concerned. It seems highly unlikely that anybody hoping to compete in computer chess tournaments will be copying from it in the future. So Bob will be happy knowing that his precious Crafty is safe...
Parent - - By Banned for Life (Gold) Date 2013-08-19 14:09
You didn't read the ruling which deals specifically with the case of material in an archive.

https://bulk.resource.org/courts.gov/c/F3/118/118.F3d.199.96-1399.html

This did not constitute distribution.

A copyright infringement is a violation of "any of the exclusive rights of the copyright owner." 17 U.S.C. § 501(a). One of those exclusive rights is the right "to distribute copies ... of the copyrighted work to the public by sale or other transfer of ownership, or by rental, lease, or lending[.]" 17 U.S.C. § 106(3). Generally, as permitted by what is known as the first-sale doctrine, the copyright owner's right to distribute a copyrighted work does not prevent the owner of a lawful copy of the work from selling, renting, lending, or otherwise disposing of the lawful copy. 17 U.S.C. § 109(a); see Professional Real Estate Investors, Inc. v. Columbia Pictures Industries, Inc., 508 U.S. 49, 52, 113 S.Ct. 1920, 1923-24, 123 L.Ed.2d 611 (1993). For example, a library may lend an authorized copy of a book that it lawfully owns without violating the copyright laws. See H.R.Rep. No. 94-1476, § 109, at 79 (1976), reprinted in 1976 U.S.C.C.A.N. 5659, 5693 and excerpted following 17 U.S.C.A. § 109. However, distributing unlawful copies of a copyrighted work does violate the copyright owner's distribution right and, as a result, constitutes copyright infringement. In order to establish "distribution" of a copyrighted work, a party must show that an unlawful copy was disseminated "to the public." 17 U.S.C. § 106(3); see National Car Rental v. Computer Associates, 991 F.2d 426, 434 (8th Cir.1993); 2 Nimmer, § 8.11[A] at 8-137.
Parent - - By bob (Gold) Date 2013-08-19 18:20
The unlawful copy was definitely distributed to Olivier.
Parent - - By Banned for Life (Gold) Date 2013-08-19 19:48
Definitely not. Vas' copy was perfectly legal. But we already know that your reading comprehension is close to zero, so it's not surprising that you are unable to follow the rather simple discussion on copyright distribution.
Parent - - By bob (Gold) Date 2013-08-19 20:02
(1) he copied my code, verbatim.

(2) he modified it, sent it to someone not connected with him.  Violated EULA.

(3) he submitted it to Olivier with no source.  Violated EULA

(4) he submitted it to be entered into an original-program only tournament.  Violated EULA.

Hence, his copy was NOT "legal" according to the EULA he was bound by.

So it is not surprising you can't follow simple reasoning and keep trying to manufacture SOME scenario where what he did was OK.  It wasn't.
Parent - By Banned for Life (Gold) Date 2013-08-19 20:47
There is no doubt that he violated the intent of your EULA. But that does not constitute a breach of copyright.

Vas has stated it was stupid to compete with Rybka 1.4 though 1.6.1, and I agree with his assessment.
Parent - - By nebulus (****) Date 2013-08-20 02:45
From the link above:
"Moreover, in this case, the plaintiffs do not even have any evidence that anyone used or looked at an infringing copy during the limitations period."

"A library's allowing on-premises public use of an unauthorized copy should probably infringe a copyright. Nonetheless, I believe that current law does not deem this sort of use an infringing "distribution," and that, in any event, there is no evidence of such use in this case." [ed. note: emphasis is mine]

In any case, I'm not sure this fuss about "distribution" is really necessary.

17 USC § 117 - Limitations on exclusive rights: Computer programs (In particular (b) Lease, Sale, or Other Transfer of Additional Copy or Adaptation):
"Adaptations so prepared may be transferred only with the authorization of the copyright owner."

Could you point to the license text where such rights were granted?
Parent - - By Banned for Life (Gold) Date 2013-08-20 04:02
§ 117 . Limitations on exclusive rights: Computer programs as its title suggests, details limitations on the rights of copyright holders. The rights of copyright holders are given in § 106.
Parent - - By nebulus (****) Date 2013-08-20 05:18
§ 117 (b) expressly prohibits the dissemination of adaptations to third parties.
Parent - - By Banned for Life (Gold) Date 2013-08-20 05:36
No, no, no. You're looking in the wrong section. § 106 gives the copyright holder's rights. Section § 117 provides exceptions to the copyright holder's rights detailed in § 106 and section B details how the first-sale doctrine applies to software. From the Fourth Circuit case above:

Generally, as permitted by what is known as the first-sale doctrine, the copyright owner's right to distribute a copyrighted work does not prevent the owner of a lawful copy of the work from selling, renting, lending, or otherwise disposing of the lawful copy.
Parent - - By nebulus (****) Date 2013-08-20 05:53
I'm not looking at the wrong section. 17 USC § 106: "Subject to sections 107 through 122, the owner of copyright under this title has the exclusive rights to do and to authorize any of the following [...]". As such, it is a subject to § 117 (b), which like I said expressly prohibits the dissemination of adaptations to third parties.
Parent - - By Banned for Life (Gold) Date 2013-08-20 06:09
You can have the last word, but your interpretation is wrong.
Parent - By nebulus (****) Date 2013-08-20 06:25
I think you're missing a point, we are talking about dissemination of adaptations. Therefore, I'm considering your interpretation as wrong. I see no point in repeating myself over and over again, and likewise, will put an end to the discussion here and now.
Parent - - By Banned for Life (Gold) Date 2013-08-18 17:43
Note that, as usual, Bob references nothing specific in his response. Just continues to claim that there is a huge volume of ICGA evidence that can't be summarized. He believes that if he repeats this bogus claim enough times, it will become true.
Parent - - By RFK (Gold) Date 2013-08-18 20:23 Edited 2013-08-18 20:26

> He believes that if he repeats this bogus claim enough times, it will become true.


In his mind he has to present a solid unshakable front. In order to do that he cannot afford to question himself.    A normal person would not be so closed off and impenetrable. He has not ever suggested the possibility no matter how remote that he might be wrong. That is not normal behavior.

[edit]

over 6,800 posts not whimper of a maybe!
Parent - By RFK (Gold) Date 2013-08-18 20:28
When he is caught in the trap of his own words -he falls back on denial.
Parent - - By Ugh (*****) Date 2013-08-18 17:35
The possibility that Vas went through the Fruit mobility code for queen, rook, bishop and knight, and, step by step, copied them, converting to bitboards is not just discredited, it is laughable. Quite apart from everything being different, the laborious complexity of what you suggest is the "copy method" is so laboriously complex and tedious; that the idea Vas didn't just code mobility direct from well established public knowledge and formulas clearly can't have occurred to you. So fixed you were in "copied" mode.

No doubt someone can URL point to the already fully documented discredit of your nonsense.

Just find me one programmer, just find one, to come here and argue it out that the mobility code was copied.

Ok, challenge for you .....
Parent - - By bob (Gold) Date 2013-08-18 17:51 Edited 2013-08-18 18:02
I will remind you, "mobility didn't sink the ship."  It was one small weight.  A BUNCH of weights took the ship down.

Let me ask you another question.

You saw Richard's post of the r1b hash store code?  Did you even THINK to look at Fruit's source.  He gave you the hint.  There are lots of quirks.  Multiple depth values and such.  All in BOTH.  hmmm...

Another coincidence I guess we are supposed to assume?  even though it doesn't look like other hash store code in other programs???
Parent - - By Ugh (*****) Date 2013-08-18 19:29
Surprisingly enough you are unable to find a programmer to come here and argue out the mobility case. Mobility is 10% plus of the ICGA technical case, bu you can't support it. Ok, readers will draw their own conclusions,
Parent - - By bob (Gold) Date 2013-08-18 21:31
Mobility is NOT "10% of the ICGA case".  Sorry,  Pure nonsense...
Parent - By Ugh (*****) Date 2013-08-19 10:56
4 parts in 38, equally weighed, looks like MORE THAN 10% to me. From that Watkins got his statistical conclusions for his report, one half of the ICGA technical evidence.

Page density on mobiliy for the Zach report, the other half of the ICGA technical evidence was around 10%

Is the % key on your pocket calculator not functioning?
Parent - By Ugh (*****) Date 2013-08-19 10:59
I take it you have no supporting programmer able to come herre and argue out the mobility part of the ICGA attack documents? Which, by the way, extrapolate to all parts of the attack documents dealing with "semantic equivalence" of code written for bitboard and mailbox data structures. As your mobility case failed, so do all the other parallel cases, which is most of both documents.
Parent - - By Banned for Life (Gold) Date 2013-08-18 17:59
I will remind you, "mobility didn't sink the ship."  It was one small weight.  A BUNCH of weights took the ship down.

This is yet another example of a complete lack of specifics after the mobility example blew up. Bob initially put mobility forward as a perfect example of copied code converted to bitboards, claiming that it was provable via semantic equivalence...
Parent - - By Ugh (*****) Date 2013-08-18 19:27
Or as the emperor Hirohito might have put it in 1945: "the mobility comparison situation has developed not necessarily to Japan's advantage". Mobility accounts for four parts in thirty eight of the Watkins attack document, more than ten percent. It forms also the same importance in the Zach document. Mobility is not "one small weight", it forms more than ten percent of the ICGA case. More than ten percent of their case, just for starters, is laughable BS.

So, let's give Hyatt another chance. Provide one, just one programmer, to come here and argue out the case the the mobility code was copied. He turned down the opportunity once, readers can draw their own conclusions.

And, from mobility, an iconic example of code between the two programs using different data structures to execute a similar known idea; from mobility, we extrapolate to the rest of the evaluation where the data structures are very different, the code very different, and the idea that a step by step copy was carried out is equally laughable.

Then we have further problems with Hyatt's methodology. He claims "semantic equivalence" means "copied", even when the equivalence is inexact, which it always is between these programs. But you can only claim B copies A if A is unique. If A exists, is known, is public knowledge, then how can Hyatt be sure B copies one A rather than any of the other examples of A? If A exists as a formula already, how can Hyatt claim B copies it, rather than just uses a known formula? If A is obvious to any strong chess playing programmer, how can Hyatt claim the strong playing programmer even NEEDS to copy it?

It has been quite obvious to me all along, that Vas didn't copy a thing. He researched, took notes, worked out from the ideas what he could do better, added some ideas from his own strong playing and programming skills (MIT graduate and international master) and wrote the damn thing himself. Doh! Bleeding obviously!
Parent - - By Rebel (****) Date 2013-08-18 20:19
It has been quite obvious to me all along, that Vas didn't copy a thing. He researched, took notes, worked out from the ideas what he could do better, added some ideas from his own strong playing and programming skills (MIT graduate and international master) and wrote the damn thing himself. Doh! Bleeding obviously!

The thing Bob never got, never will get, or simply won't admit: 1) there are more clever guys out there than him, 2) the open sources caused an explosion of knowledge how to write a chess program without inventing 100 wheels themselves which guaranteed fast progress. Examples are authors of Stockfish, Komodo, Critter.

Richard might took a lot also from others but he wrote his own code, same for Don and Tord. Here is a snapshot of the README file from Doch, Don's return to computer chess releasing a 2900 elo program out of the blue.

Don Dailey - Also, much credit goes to the authors of open source chess programs. Many of the ideas and techniques for doch have been borrowed from these wonderful works of art.
Parent - - By cipri (**) Date 2013-08-18 21:25
Don Dailey - Also, much credit goes to the authors of open source chess programs. Many of the ideas and techniques for doch have been borrowed from these wonderful works of art.

Here I don't understand at all the word "borrowed". Because they are not giving anything back, they are just taking (if not stealing). And there is also a little suspicion that some closed source  engines, are even copying the implementation (and doing some adjustments), not only the ideas.
And something like this was also the answer of vas, saying that today, he would not start from scratch, but start with an  open source engine, and improving it.

I guess, many programmers, which haven't been writing/contributing  open source code, don't take the gpl (etc.) license very seriously. I guess, it's not a problem for some of them to copy code to have their job done.
I think, to be really able to appreciate (and respect) the open source licenses, one really needs to have been contributing to some open source projects.

In my opinion, to be sure that people are not cheating, I would not allow closed source engines anymore in computer chess competitions. (I imagine something like the competitiors sending the source code with the makefiles, and the engines being compiles right before starting the tournament).
Parent - - By Rebel (****) Date 2013-08-18 21:46
You have a wrong idea about open sources. They are uploaded by the author himself (sane by mind) in the hope others learn from it. The GPL only protects the code, not the ideas behind the code. As such Don did nothing wrong, neither Vas, neither Tord, neither Richard.
Parent - - By cipri (**) Date 2013-08-18 22:02
I understand it very good. And I even didn't say that the names enumerated by you copied. I just said, that there exist people doing this, and this is something known (that some programmers not used to open source are copying without any problems from open source projects). I was saying that i guess that some are copying more than ideas, i.e. the implementation (I was speaking in general).
And the ones who are writing open source programs, are not writing it necessary to teach somebody, I guess (at least in my case) that the program is made open source for finding new contributors so that the project can make more progress. (Another reason is, when you want to reuse gpl code, so you have open source your project too). I guess most of the open source programs, are not intended for educating somebody (and I guess especially not for the ones who are planning to reuse it in closed source projects and then compete you). Just a very very small amount of open source project are aimed for educating somebody.

I guess, for somebody not involved in OS, copying  a few lines here and there is not a big deal, but for somebody involved in OS, it can be a big crime. I suspect, that this might be also the reason, why dr. hyatt is not that well understood by some people. I know from my own experience how I was thinking and behaving when I was doing closed source programming, and how latter I changed my attitude after getting to contribute to open source.
Parent - By Rebel (****) Date 2013-08-18 23:31
Have a look here, it's a classic:

https://groups.google.com/forum/#!topic/rec.games.chess.computer/2rap15WJuFA

Bob - I believe his original request was for a program to "start with" that he could modify at will.  I suggested Crafty as but one possible starting point, because that's why it was written.
Up Topic Rybka Support & Discussion / Rybka Discussion / Asides and commentary on RV's decompilation and reports
1 2 3 4 512 13 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill