GPL discussion, sense and nonsense

[Rebel]

Hi Roger,

Hello Ed Schröder,

First of all, an awestruck hello from a fan of yours for years!

I am talking about the DOS Rebel days (I have them still!).

I do believe that you do Dr. Hyatt an injustice.

He is not referring to being anti-commerce, he is against the commercial chess programming approach of take and take and take, strip out all you can from the open source programs, ask dozens of questions in various fora and then run and hide, sharing nothing.

A chess information black hole if you will. Nothing escapes to the surface.

You made Rebel's inner workings public some years ago. Do you know of any commercial programmer - retired or not - doing something as amazing as that?

No?

Neither do I at the moment...

Now wait a minute. Have you considered the amount of work? Not everybody has the talent and/or energy and/or time and/or the needed HTML knowledge for such a job.

My experiences on tournaments is the total opposite of your and Bob's judgement as well. Just ask questions and you will get an answer. Chess programmers are people with a passion and a mission, talking about their passion is in their genes, the commercial aspect is just a small obstacle, in their hearts they remain loyal to their passion, they talk.

To conclude, a little bit of history:

Frans Morsch gave recursive null-move, SMK gave us LMR. 2 breakthroughs in CC.

Recursive null-move good for 100 elo on nowadays hardware.

LMR good for 200-300-400 elo while profiting from Recursive null-move even more!

Bob (and everybody else) use it. His finger-pointing is to commercials is unjustified, in fact they contributed the most.

Ed

[Rebel]

I did not intend my comments to be directed at _all_ commercial programs. But we can go back to the 80's to see how "commercial interests" overwhelm "logic". The WMCCC in Germany. Each "company" could have 4 entries. Hardware restrictions (must be commercially available at the time of the tournament). Etc. Don't know if you were at the one in Travemunde, Germany (Valvo wrote about this under the topic "Terror in Travemunde"). Since the commercial division allowed one company to enter 4 copies, that led to conflicts of interest for the participants where the highest-scoring entrant from a company could get paired against another entry from the same company, and collusion would lead to the entry with the highest score winning that game. Claims of non-released hardware. Machines guarded like there were gold bars inside, to keep prying eyes and fingers from spotting something questionable.

While I agree with your sentiment that has nothing to do with the principle of commercial programmers unwilling to share, the topic at hand.

The above delineated situation was all about a financial deal between Levy and the directors of the (then) big companies producing the dedicated units. Programmers had no say in this.

In this respect also the Deep Blue - Kasparov event falls. One big commercial show. Hsu cooperated. And we are not going to accuse Hsu, now won't we?

As for the rest please read my answer to Roger. Commercial programmers are responsible for the REAL BIG breakthroughs of the last 10-15 years. If you agree to that we can close this subject.

Ed

[Chan Rasjid]

There is the general law that attempts to protect the innocent as well as the ignorant. But still people are expected to take proper care in their actions. So there is the legal notion of "caveat emptor" - let buyers beware.

Carrying over old traditional copyright laws to computer source codes have another very fundamental difficulty in that the text of the "book" of source codes also incorporates the very final product that readers of the book would produce. Copyright protects copying of the very text of a book and not the idea conveyed but the text. A book about how to make bolts and nuts enable the reader or end-user of the book to produce a final physical product. A book of "codes", if it is to benefit a reader, would help the end-user to produce a final product - but in this case the product are also source codes that may well be part of the very text of the "book". If the author writes badly, it is not so bad. But if he writes perfectly and if the same copyright laws hold, then every line of text becomes a constraint and others are allowed only to make similar, but non-optimal, products.

Applying the same copyright laws on computer codes is possibly trivial if it only involves publishing codes that infringes on copyright of some other codes. If we only copy but don't tell, everything's fine We could even legally compile our "copied" codes into a top ranking chess program to sell for money. If we sell and also tell, then again it becomes a problem.

I think there should be an equivalent caveat regarding releasing source codes that basically should be private. I don't know the term to coin. (I never got to understand what they were talking about ... free beer ... free as ... you are offered a free beer and you can hold on to the glass but not to drink it's content ... ?)

Rasjid

[Peter C]

Hi Roger,

Hello Ed Schröder,

First of all, an awestruck hello from a fan of yours for years!

I am talking about the DOS Rebel days (I have them still!).

I do believe that you do Dr. Hyatt an injustice.

He is not referring to being anti-commerce, he is against the commercial chess programming approach of take and take and take, strip out all you can from the open source programs, ask dozens of questions in various fora and then run and hide, sharing nothing.

A chess information black hole if you will. Nothing escapes to the surface.

You made Rebel's inner workings public some years ago. Do you know of any commercial programmer - retired or not - doing something as amazing as that?

No?

Neither do I at the moment...

Now wait a minute. Have you considered the amount of work? Not everybody has the talent and/or energy and/or time and/or the needed HTML knowledge for such a job.

My experiences on tournaments is the total opposite of your and Bob's judgement as well. Just ask questions and you will get an answer. Chess programmers are people with a passion and a mission, talking about their passion is in their genes, the commercial aspect is just a small obstacle, in their hearts they remain loyal to their passion, they talk.

To conclude, a little bit of history:

Frans Morsch gave recursive null-move, SMK gave us LMR. 2 breakthroughs in CC.

Recursive null-move good for 100 elo on nowadays hardware.

LMR good for 200-300-400 elo while profiting from Recursive null-move even more!

Bob (and everybody else) use it. His finger-pointing is to commercials is unjustified, in fact they contributed the most.

Ed

Dr. Robert Hyatt gave us rotated bitboards, good for about 100-200 elo on modern hardware.

Fabien <insert French last name that escapes my mind right now here> gave us Fruit, without which almost every program out there would be 200 elo weaker plus there would be no Rybka.

Tord Romstad contributed tons of stuff to computer chess.

Daniel Dshawl (is that how you spell it?) gave us bitbases (IIRC his were the first), good for about 20-40 elo.

The Ippo guys gave Ippolit. 'nuff said.

Miguel Ballicora gave us Gaviota Tablebases, with support for on-the-fly bitbase construction, good for about 10-30 elo on a hard disk and 50ish elo on a SSD.

BTW, are you sure Fabien didn't invent LMR? Or am I just misinformed?

Peter

[Richard Vida]

BTW, are you sure Fabien didn't invent LMR? Or am I just misinformed?
Peter

AFAIK it was Stefan Meyer Kahlen who was pioneering LMR.

[Rebel]

Dr. Robert Hyatt gave us rotated bitboards, good for about 100-200 elo on modern hardware.

Fabien <insert French last name that escapes my mind right now here> gave us Fruit, without which almost every program out there would be 200 elo weaker plus there would be no Rybka.

Tord Romstad contributed tons of stuff to computer chess.

Daniel Dshawl (is that how you spell it?) gave us bitbases (IIRC his were the first), good for about 20-40 elo.

The Ippo guys gave Ippolit. 'nuff said.

Miguel Ballicora gave us Gaviota Tablebases, with support for on-the-fly bitbase construction, good for about 10-30 elo on a hard disk and 50ish elo on a SSD.

Peter

I am not arguing the contributions of non-commercials nor highlighting the contributions of commercials. I am responding to Bob's accusation that commercials do not share.

Ed

[hyatt]

Hi Roger,

Hello Ed Schröder,

First of all, an awestruck hello from a fan of yours for years!

I am talking about the DOS Rebel days (I have them still!).

I do believe that you do Dr. Hyatt an injustice.

He is not referring to being anti-commerce, he is against the commercial chess programming approach of take and take and take, strip out all you can from the open source programs, ask dozens of questions in various fora and then run and hide, sharing nothing.

A chess information black hole if you will. Nothing escapes to the surface.

You made Rebel's inner workings public some years ago. Do you know of any commercial programmer - retired or not - doing something as amazing as that?

No?

Neither do I at the moment...

Now wait a minute. Have you considered the amount of work? Not everybody has the talent and/or energy and/or time and/or the needed HTML knowledge for such a job.

My experiences on tournaments is the total opposite of your and Bob's judgement as well. Just ask questions and you will get an answer. Chess programmers are people with a passion and a mission, talking about their passion is in their genes, the commercial aspect is just a small obstacle, in their hearts they remain loyal to their passion, they talk.

To conclude, a little bit of history:

Frans Morsch gave recursive null-move, SMK gave us LMR. 2 breakthroughs in CC.

Recursive null-move good for 100 elo on nowadays hardware.

LMR good for 200-300-400 elo while profiting from Recursive null-move even more!

Null-move is about +80 in Crafty. Measured carefully on the cluster. LMR is right at +100 today. By itself with no null-move. Again measured carefully on our cluster. One year ago it was +80, but some recent changes boosted this a bit. But even more importantly, if you already have null-move, LMR is only about +50 more. It is nowhere near +200 or beyond...

Bob (and everybody else) use it. His finger-pointing is to commercials is unjustified, in fact they contributed the most.

Ed

How do you figure that? Don Beal wrote the first paper on null-move. Fruit was the first known instance of LMR (known as history pruning). So I am not exactly sure how you say they contributed the most. To wit:

iterated search? Slate
Transposition table? Greenblatt
Exhaustive search actually works? Slate
Null-move? Beal
LMR? Fruit was the first I knew of it, although you can find discussions with Bruce Moreland and myself (with a few others) back in 1997 that discussed the idea at length, but we never fooled around with which moves to reduce and which to not reduce.
Endgame databases? Thompson
Large opening book? Thompson
Bitboards? Slate
Singular extensions? Hsu/Campbell

Etc. I hardly think that "commercials contributed the most" is anywhere within a light-year of reality.

[hyatt]

I did not intend my comments to be directed at _all_ commercial programs. But we can go back to the 80's to see how "commercial interests" overwhelm "logic". The WMCCC in Germany. Each "company" could have 4 entries. Hardware restrictions (must be commercially available at the time of the tournament). Etc. Don't know if you were at the one in Travemunde, Germany (Valvo wrote about this under the topic "Terror in Travemunde"). Since the commercial division allowed one company to enter 4 copies, that led to conflicts of interest for the participants where the highest-scoring entrant from a company could get paired against another entry from the same company, and collusion would lead to the entry with the highest score winning that game. Claims of non-released hardware. Machines guarded like there were gold bars inside, to keep prying eyes and fingers from spotting something questionable.

While I agree with your sentiment that has nothing to do with the principle of commercial programmers unwilling to share, the topic at hand.

The above delineated situation was all about a financial deal between Levy and the directors of the (then) big companies producing the dedicated units. Programmers had no say in this.

There were programmers _present_. They agreed to the "thrown games". I am talking about the tournament behaviour, win at all costs because the advertising is so important (back then).

In this respect also the Deep Blue - Kasparov event falls. One big commercial show. Hsu cooperated. And we are not going to accuse Hsu, now won't we?

What did Hsu do other than build the machine that beat Kasparov? Absolutely no cheating or unethical behaviour.

As for the rest please read my answer to Roger. Commercial programmers are responsible for the REAL BIG breakthroughs of the last 10-15 years. If you agree to that we can close this subject.

Ed

If if's and but's were candy and nuts, we'd all have a very merry Christmas.

No, I don't agree, see my reply to your response.

[hyatt]

BTW, are you sure Fabien didn't invent LMR? Or am I just misinformed?
Peter

AFAIK it was Stefan Meyer Kahlen who was pioneering LMR.

How would you know since he certainly did not discuss it publicly...

[hyatt]

When did Frans Morsch supposedly give us recursive null-move? It was certainly in Cray Blitz in 1989 as the version on Carey's "computer chess history" web site will show. The first reference I am aware of is "Selective search without tears" by Don Beal which laid this out nicely.