Last Rybka Thread for now...

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.
Post Reply
User avatar
kranium
Posts: 1922
Joined: Thu May 29, 2008 8:43 am

Re: Last Rybka Thread for now...

Post by kranium » Sat Aug 30, 2008 7:24 pm

Graham Banks wrote:
Damir wrote:Instead of all this fuss with Rybka beiing a Fruit clone, or having violated the

GPL of free source, ''how come nobody checks on Fritz 11 or Naum 3 for that

matter''.,

Since both of these 2 programs havecopied Rybka's style of play, why not

decompile them and see what is inside?

I am sure you will find plenty of interesting stuff there.
Because they have tunnel vision. They're only interested in discrediting Rybka it would seem.
none of us have any interest in simply discrediting rybka...
rybka 1.0 was disassembled in early May, 2008 by Rick Fadden in an effort to prove Strelka was a Rybka clone....we had nothing to do with it.

that was how it began...rybka was not chosen at random, or because we don't like Vas, or because it's the srongest, or because of envy.

the concerns began there...with the post by Rick.
this info has been presented, and the link to the work has been posted many, many times during this 'discussion', but i guess ignored.

we did not do the disassembly, Rick Fadden did...a loyal Rybka customer, and probably without a 'creative' effort to make it similar to Fruit.

our mistake was to become concerned by it and present it to the chess community.
Last edited by kranium on Sat Aug 30, 2008 7:27 pm, edited 2 times in total.

bob
Posts: 20923
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: ChrisW and Uri: "the GPL is invalid"

Post by bob » Sat Aug 30, 2008 7:25 pm

Uri Blass wrote:
bob wrote:
chrisw wrote:
tiger wrote:
chrisw wrote:
tiger wrote:
chrisw wrote:
tiger wrote:
chrisw wrote:
Alexander Schmidt wrote:
chrisw wrote:7. Explain why your case should not be thrown right out of court with demand for costs.
I don't plan to go to court with it. I think you missunderstood my first post, everyone shall decide for himself if it is enough or not. I will not convince you, and you will not convince me.
Fair enough, I understand. But.

If you talk GPL licence, you talk legal. If you talk legal then it's fair to step ahead into a possible legal process. I argue that the thought experiment of stepping into legal process is going to be enough to make anyone recoil from it. Hence the GPL licence talk is weakened to zero as the GPL, is basically unenforcable even if the breaches you argue have taken place.


Finally you say it!

Wow! How many days of arguing using absurd reasoning against the provided evidence?

Wouldn't have been simpler to state it right from the start?

"Zach, Christophe, Norman, we don't care about any evidence, the GPL is invalid anyway."

As far as I can see, there are two programmers arguing against the evidence: you and Uri. And both of you have stated now that they think the GPL is invalid.
// Christophe
Christophe, the reason I don't usually bother responding to your posts is because I would not be responding to factual material or factual arguments but to your distorted and fantased interpretation of what I am saying. Commonly known as the straw man.

When you stop with the pre-assumptive statements "absurb reasoning", "we don't care about any evidence" and so on, I might be bothered to make a sensible communication, but I don't feel obliged to talk with or in propaganda mode. That's for posturing purposes only, imo.

As to the GPL. Hardly an impressive document, imo. Made by well motivated amateurs. Unfortunately they forget there's a real world out there.


So I'm correct if I say that you do not care about the GPL because it is invalid, in your opinion.

I think it's a very important point that could have been said a thousands of posts ago.



// Christophe
No, again you try to put words into my mouth. Again trying to set up straw man.

I think the GPL is not enforcable in cases where heavy modification of the GPL protected code has been made. I think it is enforcable for someone who changed 10% or 5% and tried calling it his own. But a massive rewrite, no way. Not that I believe this happened in the case in question anyway.

I also think that any person assigning his code into GPL licence knows perfectly well that he is only really potentially stopping the kind of usage that turned Crafty in Drafty or whatever they called it back whenever. Ie fiddling with text strings, calling it your own name and so on. The licensor knows perfectly well that competitors and other persons are going to be majorly hacking his stuff and pinching things, either lifting code or lifting knowledge. My guess is that Fabien understood that from the start, which is why he apparently shows little concern now. Did he use GPL because it was the best choice out of several or did he use it because it was all there was?


I'm not trying to put words in your mouth, I'm trying to understand your thoughts by rephrasing them in simple terms. So you can correct me if I am wrong:

- The GPL cannot be enforced if I start from a GPL-protected source and change enough of it.

- Fabien knew from the start that derivated work including starting from his source and not only taking ideas would appear, so if the license explicitely prevents this, then it is not faithful to his real intentions.



// Christophe
I don't want to second guess Fabien's intentions. Fact remains that GPL is off the shelf and unmodified and without alternatives. It is reasonable to assume it is used because it is all there is rather than because it does exactly what is wanted.

Are you suggesting Vas took Fabien's entire source and modified it bit by bit? That would be somewhat outrageous given the lack of evidence.
I have never seen such utterly nonsensical ramblings and justifications.

The GPL is ambiguous and can't be enforced. (it can)

The GPL doesn't apply, because it was all that was available and really wasn't what Fabien wanted to use, but he used it in the absense of nothing better. (nobody has to use the GPL, you _choose_ to use it, you are not _forced_ to use it.)

The GPL hinders progress (exactly the opposite of what it actually does).

The GPL only applies to exact syntax, not to lines that are similar in syntax and identical in semantics.

The GPL doesn't apply to single lines here and there.

The list goes on and on. And it is all complete nonsense.
The way that you use the GPL clearly can cause people to quit programming because they are going to be afraid of witch hunt against them.
How? If they want to start with an open source program, they just have to be open-source themselves. If they want to be commercial, they have to start from scratch. If they can't write code from scratch, then just maybe they ought not tackle computer chess first?


The evidence that I saw does not convince me that rybka broke the GPL.

If one programmer undersands the code of another programmer and later write his code then it is logical to find more similiarities even if copy and paste is not done.

Uri
We are not looking at _similarities_. We are looking at "semantical equivalence" which is a huge difference. We all have move generators. Many of us might use bitboards. A smaller set of those might call their generator "GenerateMoves". But it is unlikely that someone has done what I have done (GenerateCaptures, GenerateNonCaptures, GenerateChecks, GenerateCheckEvasions) and then when you look at each of those things are done in exactly the same way (it would not matter whether you generate knight or bishop movs first but if you generate them in exactly the same way it is an issue). And by the time you get down to source-code level, there will not be many semantically-equivalent pieces of code. Regardless of what _anybody_ claims here, it just doesn't happen. There are just too many different ways to express the same algorithm in terms of source code.

bob
Posts: 20923
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Last Rybka Thread for now...

Post by bob » Sat Aug 30, 2008 7:29 pm

Graham Banks wrote:
Damir wrote:Instead of all this fuss with Rybka beiing a Fruit clone, or having violated the

GPL of free source, ''how come nobody checks on Fritz 11 or Naum 3 for that

matter''.,

Since both of these 2 programs havecopied Rybka's style of play, why not

decompile them and see what is inside?

I am sure you will find plenty of interesting stuff there.
Because they have tunnel vision. They're only interested in discrediting Rybka it would seem.
Right. Lets ask two people to spend time disassembling and studing a dozen programs. And demand answers tomorrow. It would take a year or more doing those in parallel. Why not one at a time?

User avatar
Zach Wegner
Posts: 1922
Joined: Wed Mar 08, 2006 11:51 pm
Location: Earth
Contact:

Re: Last Rybka Thread for now...

Post by Zach Wegner » Sat Aug 30, 2008 7:34 pm

Albert Silver wrote:The worst part, IMHO, is that none of what has been shown is remotely related to Rybka's playing ability. I am unqualified to comment on the code itself, but what has been made clear, is that all that is being discussed is code about how the engine sends moves to the interface, and how it gets them back.
No, it's not clear. In fact the main concern I have in the UCI code that was posted is the time control code, which is most definitely related to playing strength.

I have posted elsewhere about similarities in the board representation, the root search function, and the hash table structure. All of these are directly related to playing strength.

Albert Silver
Posts: 2969
Joined: Wed Mar 08, 2006 8:57 pm
Location: Rio de Janeiro, Brazil

Re: Last Rybka Thread for now...

Post by Albert Silver » Sat Aug 30, 2008 7:39 pm

Zach Wegner wrote:
Albert Silver wrote:The worst part, IMHO, is that none of what has been shown is remotely related to Rybka's playing ability. I am unqualified to comment on the code itself, but what has been made clear, is that all that is being discussed is code about how the engine sends moves to the interface, and how it gets them back.
No, it's not clear. In fact the main concern I have in the UCI code that was posted is the time control code, which is most definitely related to playing strength.

I have posted elsewhere about similarities in the board representation, the root search function, and the hash table structure. All of these are directly related to playing strength.
If you think you found something, is there any particular reason you do not ask Vas himself or post in his forum, as invited? It would seem to be eminently more logical to simply ask the man himself, than ask 200 people who did NOT write Rybka, or ask someone to forward your questions, no? Registration at the Rybka forums is as painless as here. :roll:

Albert
"Tactics are the bricks and sticks that make up a game, but positional play is the architectural blueprint."

User avatar
Graham Banks
Posts: 35088
Joined: Sun Feb 26, 2006 9:52 am
Location: Auckland, NZ

Re: Last Rybka Thread for now...

Post by Graham Banks » Sat Aug 30, 2008 7:40 pm

bob wrote:
Graham Banks wrote:
Damir wrote:Instead of all this fuss with Rybka beiing a Fruit clone, or having violated the

GPL of free source, ''how come nobody checks on Fritz 11 or Naum 3 for that

matter''.,

Since both of these 2 programs havecopied Rybka's style of play, why not

decompile them and see what is inside?

I am sure you will find plenty of interesting stuff there.
Because they have tunnel vision. They're only interested in discrediting Rybka it would seem.
Right. Lets ask two people to spend time disassembling and studing a dozen programs. And demand answers tomorrow. It would take a year or more doing those in parallel. Why not one at a time?
Let's see if it happens. I doubt it will.
gbanksnz at gmail.com

Damir
Posts: 2473
Joined: Mon Feb 11, 2008 2:53 pm
Location: Denmark
Full name: Damir Desevac

Re: Last Rybka Thread for now...

Post by Damir » Sat Aug 30, 2008 7:41 pm

Is anyone here interested from copying a style of certain program?

Cause that's exactly what is going to happen if Rybka 1.0 beta source is

released to the public?

Some authors already did that with Strelka code when it was publicly

available.

The code contained some of the Rybka and Fruit sources.

Some of them managed to improve their programs, some of them did not,

and now they want another gift, so they can do the same to Rybka ??

Is this what chess programming is all about, to copy other people's work?

User avatar
Zach Wegner
Posts: 1922
Joined: Wed Mar 08, 2006 11:51 pm
Location: Earth
Contact:

Re: Last Rybka Thread for now...

Post by Zach Wegner » Sat Aug 30, 2008 7:54 pm

Albert Silver wrote:If you think you found something, is there any particular reason you do not ask Vas himself or post in his forum, as invited? It would seem to be eminently more logical to simply ask the man himself, than ask 200 people who did NOT write Rybka, or ask someone to forward your questions, no? Registration at the Rybka forums is as painless as here. :roll:

Albert
This is precisely what happened: http://www.talkchess.com/forum/viewtopic.php?t=23249

CW said he sent it a couple days ago, but no response as of yet. And for the record, I am registered at the Rybka forum, and I (in addition to several other people) have voiced some concerns. No response, except one "Rybka is 100% original".

bob
Posts: 20923
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Last Rybka Thread for now...

Post by bob » Sat Aug 30, 2008 8:03 pm

Damir wrote:Is anyone here interested from copying a style of certain program?

Cause that's exactly what is going to happen if Rybka 1.0 beta source is

released to the public?

Some authors already did that with Strelka code when it was publicly

available.

The code contained some of the Rybka and Fruit sources.

Some of them managed to improve their programs, some of them did not,

and now they want another gift, so they can do the same to Rybka ??

Is this what chess programming is all about, to copy other people's work?
:)

That is the issue that has been raised with respect to Rybka itself. (Was it copied from Fruit). So you have to be careful how you phrase your questions or you make a point you really didn't want to make. :)

Damir
Posts: 2473
Joined: Mon Feb 11, 2008 2:53 pm
Location: Denmark
Full name: Damir Desevac

Re: Last Rybka Thread for now...

Post by Damir » Sat Aug 30, 2008 8:15 pm

Yes, but Rybka 1.0 beta was 100 elo stronger than commercial version of Fruit when it was released.
Surely some of the ideas were unlike to be found in Fruit which make the Rybka 1.0 beta no Fruit like.

Post Reply