Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

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

Moderators: hgm, Rebel, chrisw

bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by bob »

kranium wrote:
bob wrote: Norman, give it a rest. IP*/Robo* are _clearly_ clones of something, produced by decompiling something.
Bob...

"If" (and that's a big 'if') Ippolit is the result of 'decompiling' (as you seem to somehow know as fact):

Decompiling is not 'wrong' or 'illegal' in any way...
it is a completely legitimate form of discovery...(according to the US Supreme Court)
Do some reading first. You can't decompile (reverse-engineer) and then sell that result as your own. That is a clear and simple copyright/patent infringement. You are certainly allowed to look inside to see how it works, and to even develop an interface that will talk to the thing you reverse-engineered. But you can't sell the reverse-engineered thing legally. You can't distribute it for free. It belongs to the original author.

You can look at it, and publish the ideas used inside. But not the actual code.

It's no different than taking something apart to see how it works...
(this is the epitome of human curiosity and the basis of enormous amounts of technological progress...)
Correct. But you can't take something apart, then manufacture an identical copy, and sell that as your own. You can't even sell it period.
It's discovering 'ideas'...it is not 'copying'!
You have a strange definition of "discovering". Discovering does not mean copying something verbatim...
The source code has to be recreated, and doing this is a major major task that requires a seriously skilled engineer.
(and you know that!)

Same bullshit argument that keeps coming up. That is "translation" (going from machine language back to C) that is the same as the process of compiling (going from C back to assembly language). That is not original creativity. It is copying.
bob wrote:
hgm wrote:It is often difficult to spot clones. It is easy, however, to spot clowns... :lol: :lol: :lol:
I thought it was "spot on" myself. Nicely done, HGM. :)
bob wrote: There are plenty of original programs. Just because you are incapable of writing something original doesn't mean others can't and haven't...
?
The fact that you feel the need to jump on the 'personal insult' bandwagon because I post something you disagree with is really sad...(pitiful).
Not _nearly_ as pitiful as believing that copying by translation is original work.

Writing what conforms to your strict academic definition of an 'original' program does not interest me...
Of course it doesn't. And that's irrelevant. Most don't have your severely warped sense of ethics and legality.
i.e. it is not the 'end all' and goal of chess programming...
my interest is working to improve existing code, pushing the envelope of chess playing programs...and I have successfully done that.
Where, exactly?


I (and many others) am not at all interested in suscribing to you narrow-minded and vain idea that the only worthwhile goal is to create something absolutely new, original, like a Hemmingway novel.
Sorry that's utter nonsense.
To those with no creativity, I won't disagree. But thankfully there _are_ many authors that do create new and unique literature that I enjoy reading. My favorite is Matt Reilly (scarecrow, area 7, Ice Station, Contest, etc.) Others are Clive Cussler, Andy McDermott, Patrick Robinson. And I have NEVER, not one time, found any example where one of them copied the storyline, much less the actual text, of one of the others, and tried to foist it off as their new work. Of course they do seem to have a strong ethical sense.


Besides, there are hundreds of absolutely mediocre ('crappy') and boring 'original' programs..who the hell needs another?

And to suggest I'm not capable is simply a low blow...
another of the many insults I've endured from the self-absorbed, ego-centric, puritanical, and contemptuous Talkchess establishment.
Sorry. Sometimes the truth hurts.

The work I've done on RobboLito, Igorrit, Fire, etc... speaks for itself.
It's unfortunate that you have no clue about (or refuse to recognize) the quality and quantity of work I've done on these (and other) now state-of-the-art programs
(all of which make Crafty look like a patzer BTW)
Certainly they are stronger. But Crafty was written by me. Robo was _copied_ from someone else. I'll take the original road myself. I find a sense of satisfaction in doing something original, rather than copying what someone else did and then claiming it as my own work. But you won't get that point, I am sure...


It seems clear that (as your career winds down), your only goal is to discredit as many programs and authors as possible in order to inflate and cement Crafty's legacy.

Norm
Nope. There are several that are now better than me at doing this. Doesn't detract from my enjoyment of doing it however. The ICGA asked us to investigate Rybka. We did. The rest will be exposed once they make a final ruling. Robo and friends will _never_ be welcome in any event, which is certainly a good thing. Other programs (Junior, Shredder, Stockfish, etc are going to continue to be welcomed in any event unless one of those is discovered to be a derivative work.

But you won't ever get that point...
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by bob »

tpetzke wrote:I am a programmer of one of those absolutely mediocre ('crappy') and boring 'original' programs and it exists because I wanted to prove to myself (not the world) that I'm able to create one.

And its mediocre because it takes a lot of time to understand the ideas and the theory behind all that stuff that makes at the end a chess engine and then find your own implementation for it. But that's the interesting part for me. I don't even use downloaded endgame tables because I found it much more interesting to understand the theory behind them and program an algorithm that calculates them from scratch.

Of course I also want my engine to be strong and if it wins a match it makes me proud because every bit in it was from me. As soon as I started to include code from others the engine might win only because of it and it would yield no gain for me what so ever.

So it makes me proud to be able to create an engine, others might be proud if they prove they are able to clone or modify one.

As soon as it comes to tournaments I'm not troubled with clones, because where my engine plays clones are not an issue. but I can understand the annoyance of those authors that have developed their now strong own engine over years and are now faced with modifications of the same program over and over again. It pollutes the engine ranking table somehow.

Thomas...
Well said. Another one that "gets it." Creating a new program is a lot of effort and time. Copying one takes no effort or time.
kranium
Posts: 2129
Joined: Thu May 29, 2008 10:43 am

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by kranium »

bob wrote:
Same bullshit argument that keeps coming up. That is "translation" (going from machine language back to C) that is the same as the process of compiling (going from C back to assembly language). That is not original creativity. It is copying.
If IppoLit was directly copied from Rybka3, as you and some have alleged...
then why was almost immediately 50+ or more points stronger?

Also:
You stated that there have been presented many positions from which RobboLito and Houdini produce identical output...
and that this proves Houdini is a 'clone' of RobboLito.

If Robbo is directly copied from Rybka, then why don't Robbo and Rybka produce the same identical output in the same positions?
By your logic, this should serve as proof that Robbo is not 'copied' from Rybka?
bob wrote: Certainly they are stronger. But Crafty was written by me. Robo was _copied_ from someone else. I'll take the original road myself. I find a sense of satisfaction in doing something original, rather than copying what someone else did and then claiming it as my own work. But you won't get that point, I am sure...
I do not 'claim' Robbo or Fire as 'my own work', and never have.
Perhaps it is you that should do some reading...
The source credit is stated many times all over my web site and also abundantly in the ample documentation included...
I don't use my real name either...I do it for the love and interest of working on trying to improve a high level chess program.
bob wrote: Well said. Another one that "gets it." Creating a new program is a lot of effort and time. Copying one takes no effort or time.
So, working on a 'derivative', (a valid fork of an open-source program) has no value?
kranium wrote: i.e. it is not the 'end all' and goal of chess programming...
my interest is working to improve existing code, pushing the envelope of chess playing programs...and I have successfully done that.
bob wrote: Where, exactly?
have you been living under a rock?
http://www.chesslogik.com
there is ample documentation available
the public 'beta' development took place om immortal chess forum, summer/fall/winter 2009 and spring 2010
Sentinel and I developed strong, stable, and feature-rich windows versions of Ippo, Robbo, Igorrit, etc.
newest and best Robbo version is:
RobboLito 0.085g3

Fire is also presented there..the source code is avaialble...
you'll seen many hundreds of improvements and beneficial changes...
kranium
Posts: 2129
Joined: Thu May 29, 2008 10:43 am

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by kranium »

Graham Banks wrote:
kranium wrote:Besides, there are hundreds of absolutely mediocre ('crappy') and boring 'original' programs..who the hell needs another?
That's a real slap in the face to any programmer who has actually made the effort to write a chess engine from scratch. :roll:
No...it's my opinion.

It is simply my personal choice 'not' to spend my time on a mediocre program...
I prefer to channel my 'effort' elsewhere,
others are entitled to do what they want with their time. I have no issue with that.

But believe me Graham, I get the drift.
The CCC environment leaves no doubt as to the message:
Led by Bob, the Talkchess 'culture' puts values on 100% originality, only...
i.e. the only valid program is written from scratch.

(I'd like to see all these programmers of the referred to 'mediocre' engines prove 100% originality...)

And anything else is trash..., and anyone working on a derivative is a fraud, a simple 'cloner'...

Bob said it himself:
hyatt wrote:Clones. Derivatives. Doesn't matter. Either is just as bad (and unacceptable) as the other.
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by bob »

kranium wrote:
bob wrote:
Same bullshit argument that keeps coming up. That is "translation" (going from machine language back to C) that is the same as the process of compiling (going from C back to assembly language). That is not original creativity. It is copying.
If IppoLit was directly copied from Rybka3, as you and some have alleged...
then why was almost immediately 50+ or more points stronger?
It wasn't. It crashed too often. Too many reverse-engineering (asm to C) errors causes this, almost certainly.

Also:
You stated that there have been presented many positions from which RobboLito and Houdini produce identical output...
and that this proves Houdini is a 'clone' of RobboLito.

If Robbo is directly copied from Rybka, then why don't Robbo and Rybka produce the same identical output in the same positions?
By your logic, this should serve as proof that Robbo is not 'copied' from Rybka?
Just do some research rather than asking uninformed questions, particularly when you don't really want the answers in the first place.
bob wrote: Certainly they are stronger. But Crafty was written by me. Robo was _copied_ from someone else. I'll take the original road myself. I find a sense of satisfaction in doing something original, rather than copying what someone else did and then claiming it as my own work. But you won't get that point, I am sure...
I do not 'claim' Robbo or Fire as 'my own work', and never have.
Perhaps it is you that should do some reading...
OK, you offer the bait, I'll bite. Which one of _US_ has admitted to copying another program and then trying to claim it as their own work? Hint: "not me."


The source credit is stated many times all over my web site and also abundantly in the ample documentation included...
I don't use my real name either...I do it for the love and interest of working on trying to improve a high level chess program.
bob wrote: Well said. Another one that "gets it." Creating a new program is a lot of effort and time. Copying one takes no effort or time.
So, working on a 'derivative', (a valid fork of an open-source program) has no value?
kranium wrote: i.e. it is not the 'end all' and goal of chess programming...
my interest is working to improve existing code, pushing the envelope of chess playing programs...and I have successfully done that.
bob wrote: Where, exactly?
have you been living under a rock?
http://www.chesslogik.com
there is ample documentation available
the public 'beta' development took place om immortal chess forum, summer/fall/winter 2009 and spring 2010
Sentinel and I developed strong, stable, and feature-rich windows versions of Ippo, Robbo, Igorrit, etc.
newest and best Robbo version is:
RobboLito 0.085g3
That was my point, although it was a bit too subtle it seems. So you took a program, and modified it, and think you should get a big "attaboy" for doing this? "attaboys" are given to those that do original work. I don't "attaboy" my xerox copier each time it spits out another copy of a final exam for me...


Fire is also presented there..the source code is avaialble...
you'll seen many hundreds of improvements and beneficial changes...
I'm not even interested in doing a diff to see how many changes there were. I suspect that "many hundreds" will turn into less than a hundred, maybe much less. But I really don't care. It's a derivative/clone. Always was, always will be...
Tom Barrister
Posts: 227
Joined: Tue Oct 05, 2010 5:29 pm

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by Tom Barrister »

kranium wrote:
bob wrote: Norman, give it a rest. IP*/Robo* are _clearly_ clones of something, produced by decompiling something.
Bob...

"If" (and that's a big 'if') Ippolit is the result of 'decompiling' (as you seem to somehow know as fact):

Decompiling is not 'wrong' or 'illegal' in any way...
it is a completely legitimate form of discovery...(according to the US Supreme Court)
The Supreme Court made no such ruling.
This production is being brought to you by Rybka: "The engine made from scratch.™"
User avatar
fern
Posts: 8755
Joined: Sun Feb 26, 2006 4:07 pm

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by fern »

HAHAHAHAHAHAH
great, harvey.
Hoep Norma-n understand this is just a joke.

Fern
poisonedpawn
Posts: 76
Joined: Wed Nov 03, 2010 5:56 pm

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by poisonedpawn »

why so down on norman? i think he raises some very valid points. the 'derivitives' have never been proven to be cloned, as far as i know. on the other hand there has been recently much publicised discreditting of rybka and its author...a programe that was once the darling of this forum! if those who merely ask questions are so instantly shot down what hope of ever getting to the truth?
the geeks shall inherit the earth
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by bob »

poisonedpawn wrote:why so down on norman? i think he raises some very valid points. the 'derivitives' have never been proven to be cloned, as far as i know. on the other hand there has been recently much publicised discreditting of rybka and its author...a programe that was once the darling of this forum! if those who merely ask questions are so instantly shot down what hope of ever getting to the truth?
All that is not known for certain, is what Robo was derived from. Speculation was Rybka 3. Vas said the same.

But all of the rest is beyond doubt. Robo->ivanhoe->firebird->houdini->name any other derivative of this rag-tag gene pool you want. But Robo sits at the head and all the later programs are derivatives (to be kind) or clones (to be more accurate) (derivative is derived from another program, a clone is a near-identical copy, which is closer for at least this 'family')...
poisonedpawn
Posts: 76
Joined: Wed Nov 03, 2010 5:56 pm

Re: Houdini, Fire, IvanHoe, (and Rybka?) are 'clones'...?

Post by poisonedpawn »

i think this whole issue would be resolved if engine authors actually shared their findings with the aim of advancing computer chess, rather than just chasing a buck.
the geeks shall inherit the earth