POLL - We are the Computer Chess Community. We control ourselves.

http://talkchess.com/forum3/viewtopic.php?f=2&t=76650

Read the post of Andrew first.

I need to think about this.

I agree it is unethical to sell a free product, by renaming the product. To mislead the unknowing consumer.

But I have a few questions.

If the Fire engine for example is a Stockfish derivative. Is Fire an illegal engine? Or just another branch of Stockfish. As it is not being sold for profit.

And is there a way to know for sure? My concern is who decides.

"This community runs on the court of public opinion, public discourse, and public debate."

Because this is not what I have observed.

"3. Yes, the community need a new rating list from scratch with only orginal works."

Correct me if I am wrong, but Stockfish itself is not an original work.

Admin wrote:

http://talkchess.com/forum3/viewtopic.php?f=2&t=76650

Read the post of Andrew first.

Joan d’Arc

Admin wrote:

http://talkchess.com/forum3/viewtopic.php?f=2&t=76650

Read the post of Andrew first.

I read Andrew's post, and I feel that I was supposed to vote for option 2 or 3 - but the way it's worded, I felt I had to vote for option 1.

Someone could branch an existing code base, make some important changes, and re-offer it back under the same license. This would be a positive contribution, but I feel that this would make it a "clone", a word which seems to be being used in a derogatory way here.

Why put your source code under a license that allow people to use it and sell it.

Now they are trying to circumvent their own license by putting pressure on the rating sites to not use them by publicly announcing their grievances all over the web. Also they say they want to inform the public not to buy Chessbase products, I call this unethical.

I see that my opinion is a minority one. I never liked the Rybka 3 (read Ippolit and friends) derivatives on rating lists. Nowadays I have seen CCRL examples having Sugar, Shashchess listed and Stockfish not, because they were 5 elo (or so) better, that's the world upside down. Now we have FF1 (Lc0), FF2 (Stockfish), they might be legal but they remain what they are, clones. I can't take the top rankings of rating lists serious any longer.

Admin wrote:

I see that my opinion is a minority one. I never liked the Rybka 3 (read Ippolit and friends) derivatives on rating lists. Nowadays I have seen CCRL examples having Sugar, Shashchess listed and Stockfish not, because they were 5 elo (or so) better, that's the world upside down. Now we have FF1 (Lc0), FF2 (Stockfish), they might be legal but they remain what they are, clones. I can't take the top rankings of rating lists serious any longer.

I do not disagree with you. And I do not like it. But I can not trust this community to be fair to all engines, and programmers. Everyone only wants to be fair. When it is convenient!

Is another of the periodic events where computer chess generally tears itself apart. Look back at prior episodes and there’s one feature in common (you can see it now in the A Grant talkchess post), one person, it usually appears to be one each time, claiming for himself all moral high values, which he sure will endlessly, self-promotingly, tell you all about, trying to engineer a posse, with him as leader of course, against some other who it is claimed has transgressed all morality, and, if all plays out according to the usual plan, everybody else takes sides, a giant fight ensues, objective apparently the triumph of good and the destruction of the bad. Good versus evil bla bla. Actually, in my opinion, it’s about narcissistic and psychopathic power, a giant ego-power trip, that invariably backfires. Remember Hyatt. More recently HGM.

Chris Whittington wrote:

Is another of the periodic events where computer chess generally tears itself apart. Look back at prior episodes and there’s one feature in common (you can see it now in the A Grant talkchess post), one person, it usually appears to be one each time, claiming for himself all moral high values, which he sure will endlessly, self-promotingly, tell you all about, trying  to engineer a posse, with him as leader of course, against some other who it is claimed has transgressed all morality, and, if all plays out according to the usual plan, everybody else takes sides, a giant fight ensues, objective apparently the triumph of good and the destruction of the bad. Good versus evil bla bla. Actually, in my opinion, it’s about narcissistic and psychopathic power, a giant ego-power trip, that invariably backfires. Remember Hyatt. More recently HGM.

We are not ignorant of his devices!

But we know for years and years that chessbase is working this way.
They have one main GUI they almost never changed. Ugly kind of gui.
Not standard. And they put bugs back in with each new edition.

Really good GUIs look different.
And they are used different.
Take Shredder, rebel GUIs and even chess system tal as DOS engines were more standard then any chessbase gui. Arena GUI.

We know how they do fake news when advertising their products.
And we know how they come with big hardware to the tournaments , often with more then one entry,
To get a title.

I was never a fan of rating lists, no matter if ssdf or ccrl or cegt.
Also never a Chessbase fan.
In the beginning wuellenweber wanted a free game collection club and did not want to make profit with it.

Then came Friedel.
And suddenly Chessbase and the free club of game collectors changed into a commercial product.

Admin wrote:

I see that my opinion is a minority one. I never liked the Rybka 3 (read Ippolit and friends) derivatives on rating lists. Nowadays I have seen CCRL examples having Sugar, Shashchess listed and Stockfish not, because they were 5 elo (or so) better, that's the world upside down. Now we have FF1 (Lc0), FF2 (Stockfish), they might be legal but they remain what they are, clones. I can't take the top rankings of rating lists serious any longer.

I accept your point of view - but what about people who legitimately branch open source software, and the new branch is open source as well?

Sometimes, you just have to accept that more than one chart is needed: look at all the Billboard charts - each of the 15 main categories has multiple charts in it - link.

You seem to want a chart for "original code", but what if a code branches and both branches claim to be the original? Remember - Stockfish is a fork of Glaurung - so even Stockfish wouldn't get into your chart!  

The nice clean world in which each program was original is somewhat behind us IMO.

You can start your own list: Eric Hallsworth did that, and whether it represented good judgement by me or not, that was the list I used to respect.

My two cents.

First, I do not see why Sugar, Shashchess would be clones. I think a better word would be "forks". The code is published, credit is given, the existence of these forks is deeply in the spirit of open source and GPL. Stockfish itself started as a fork. And if they are stronger than Stockfish, so be it. SF is tested for rating lists, why not testing Shashchess? Why should being at the very top matter so much?

OTOH, I think that illegal clones should not be on the lists: the Strelkas, Houdinis, etc.

But I do believe that we should live in a free world where everybody could make some choices, at least. If the testers of CCRL want to include some engines it is OK with me. We see that CEGT testers have some bias toward commercial engines, but it is also fine. Why not? It is their right to decide for themselves.

And, let's not fool ourselves. There is no such thing like "computerchess community". There is cooperation among several authors, there are people, like me, who find some interests reading posts on CCC or here, but we are all so different, from different cultures, part of the world, that misunderstandings are a daily phenomenon.

About Andrew's post: he is really young, and, as it is usually, really ignorant both of the ways of the world, and of himself and his own motives. But that's youth.

Prometheus wrote:

 Why put your source code under a license that allow people to use it and sell it.

Now they are trying to circumvent their own license by putting pressure on the rating sites to not use them by publicly announcing their grievances all over the web. Also they say they want to inform the public not to buy Chessbase products, I call this unethical.

I agree with you that a boycott goes much too far, for me Fritz by far has the best interface, but FF1 and now FF2 don't belong on rating lists, it's supporting cloning.

BTW, welcome.

matejst wrote:

My two cents.

First, I do not see why Sugar, Shashchess would be clones. I think a better word would be "forks". The code is published, credit is given, the existence of these forks is deeply in the spirit of open source and GPL. Stockfish itself started as a fork. And if they are stronger than Stockfish, so be it. SF is tested for rating lists, why not testing Shashchess? Why should being at the very top matter so much?

Maybe I am a bit old fashioned, back in 2011/12 the community exploded with the various Ippolit / Robolito derivatives I said allow those if they make a significant contribution of 50 elo. Good examples were Toga, Fruit reloaded. I still hold this view, stubborn me.

Quote :

OTOH, I think that illegal clones should not be on the lists: the Strelkas, Houdinis, etc.

But I do believe that we should live in a free world where everybody could make some choices, at least. If the testers of CCRL want to include some engines it is OK with me. We see that CEGT testers have some bias toward commercial engines, but it is also fine. Why not? It is their right to decide for themselves.

And, let's not fool ourselves. There is no such thing like "computerchess community". There is cooperation among several authors, there are people, like me, who find some interests reading posts on CCC or here, but we are all so different, from different cultures, part of the world, that misunderstandings are a daily phenomenon.

Surely there is a computer chess community, however there is not the "we are" variation, totally agree.

Quote :

About Andrew's post: he is really young, and, as it is usually, really ignorant both of the ways of the world, and of himself and his own motives. But that's youth.

Whether his call is out of idealism, fair competition or self-serving, maybe a cocktail of those, I am with him. I come from a background where playing strength was all that matters and that topping the SSDF list was more important than winning a five game world champion title.

Other than that, there is hardly on-topic content in his thread, it's not taken seriously.

matejst wrote:

My two cents.

First, I do not see why Sugar, Shashchess would be clones. I think a better word would be "forks". The code is published, credit is given, the existence of these forks is deeply in the spirit of open source and GPL.

I'm not sure if there's a clear definition of cloning, but here's what's good and what's bad:

GOOD

Forking an open source project, crediting the original project, and offering the fork under the same license as the code it was forked from.


BAD

1. Reverse engineering a program and offering it as a new program without crediting the original author

2. Making a fork of an open source project, and not upholding the open source license on the new project

3. Using some code from other software, and not declaring this in a competition whose rules require that you do so


Regarding FF2: it is clear that it exaggerates the extent to which it is both original and world beating. The only legal question is whether it has to publish the weights used in its net: given that the net is not obviously better than ones available for free, in this case it probably doesn't matter. This leaves two questions:

1. is it good value at the European price of €100?

Emphatically "no".

2. is it good value at the price on the Indian Chessbase site at about €40?

Emphatically "no".

In regard to this thread, the question is, should it appear on rating lists?

My answer is "yes": for me, public interest demands it - but as I've said earlier, I also think that, like Billboard magazine, we need multiple lists.

TheSelfImprover wrote:

In regard to this thread, the question is, should it appear on rating lists?

My answer is "yes": for me, public interest demands it - but as I've said earlier, I also think that, like Billboard magazine, we need multiple lists.

I think yes as well but it should be made clear on the list that it is a part of the Stockfish family so as not to mislead people into thinking it is an original engine.

Chessbase is like Microsoft. They exploit people and situations.

I want it opposite way.
That the programmers of computerchess get the money they deserve.
Money to mark Uniacke, stefan meyer kahlen, richard lang, etc. Etc.

Money to the people who do something. Not to a company exploiting the chances
And the market.

harvey williamson wrote:

TheSelfImprover wrote:

In regard to this thread, the question is, should it appear on rating lists?

My answer is "yes": for me, public interest demands it - but as I've said earlier, I also think that, like Billboard magazine, we need multiple lists.

I think yes as well but it should be made clear on the list that it is a part of the Stockfish family so as not to mislead people into thinking it is an original engine.

Harvey, without a poll the subject is going nowhere on Talkchess, is the option still not available?

Admin wrote:

harvey williamson wrote:

TheSelfImprover wrote:

In regard to this thread, the question is, should it appear on rating lists?

My answer is "yes": for me, public interest demands it - but as I've said earlier, I also think that, like Billboard magazine, we need multiple lists.

I think yes as well but it should be made clear on the list that it is a part of the Stockfish family so as not to mislead people into thinking it is an original engine.

Harvey, without a poll the subject is going nowhere on Talkchess, is the option still not available?

No, they still do not work.

Admin wrote:

http://talkchess.com/forum3/viewtopic.php?f=2&t=76650

Read the post of Andrew first.

D. Clones, if tested, should be grouped together with the main body of work. Not this clone shall be kept separate because it represents the Chessbase/Fritz brand. That was a crock of do-do.

matejst wrote:

My two cents.

First, I do not see why Sugar, Shashchess would be clones. I think a better word would be "forks". The code is published, credit is given, the existence of these forks is deeply in the spirit of open source and GPL. Stockfish itself started as a fork. And if they are stronger than Stockfish, so be it. SF is tested for rating lists, why not testing Shashchess? Why should being at the very top matter so much?

OTOH, I think that illegal clones should not be on the lists: the Strelkas, Houdinis, etc.

But I do believe that we should live in a free world where everybody could make some choices, at least. If the testers of CCRL want to include some engines it is OK with me. We see that CEGT testers have some bias toward commercial engines, but it is also fine. Why not? It is their right to decide for themselves.

And, let's not fool ourselves. There is no such thing like "computerchess community". There is cooperation among several authors, there are people, like me, who find some interests reading posts on CCC or here, but we are all so different, from different cultures, part of the world, that misunderstandings are a daily phenomenon.

About Andrew's post: he is really young, and, as it is usually, really ignorant both of the ways of the world, and of himself and his own motives. But that's youth.

You are no programmer, which can be seen in all of your posts and opinions.

FYI Strelka has more original code than Shash and Sugar ever had and will have.
(Shash even is completely scam with this Shashin theory thing [lol] because in reality it never existed or simply is sth else and was removed anyway)
Shash and Sugar just recycle old and long known SF or SF derivate patches and only non-programmers think they contain sth new or whatever non-programmers think)

temporarily wrote:

FYI Strelka has more original code than Shash and Sugar ever had and will have.

Which version of Strelka do you mean ?

Admin wrote:

temporarily wrote:

FYI Strelka has more original code than Shash and Sugar ever had and will have.

Which version of Strelka do you mean ?

The one he wrote, part wrote, whatever

@temporarily:

Of course I am not a programmer, I wrote it many times. But I do know that many of the patches in SugaR and Shashchess are patches tried in SF, and the authors do not hide it. But it does not change much: even if the difference between SF and ScCh or SgR is 0,2% (e.g.), it does not change their status, their legality, and their legitimacy.

I agree that the situation with Strelka is complicated, but I do think that its legal status was disputable.

Wish we were back in the 70's, 80's. No internet. No open sources. Only regular mail or phone to communicate. Everybody inventing their own wheels. At the cost of 1000 (or so) elo less. Who cares.