World Chess Computer Champion?

[hgm]

The rule says: "must be mentioned in the registration". It does not say: "You can lie about it as long as we know that you are lying". That (admittedly quite original) interpretation seems to be entirely yours. There can be serious doubts as to whether anyone in the CSVN would share it.

Anyway, it doesn't seem to contradict my original statement that the CSVN rules are exactly as restrictive as the ICGA rules. It is just that you claim they are not going to be enforced that makes the difference.

[bob]

Well, I don't know anything about TCEC, as it is completely without interest to me.

Well, it's usually not wise to enter a discussion you don't know anything about . So far only 6 of the 46 voters consider the ICGA as the legitime body for the world-title.

That would have been quite different 2½ years ago.

Something must have happened, but what?

a "herd mentality" has decided that the ICGA is irrelevant. 90% of those "voters" are not programmers. I'd imagine if you ask women in the US, 90% would vote to "get rid of NASCAR, football, basketball, and all the other events that occupy so much prime-time TV."

I don't care what the users, the fans, the whatever think about how the ICGA event is run. I only care about what _I_ and other _programmers_ think. We are the reason the event is held in the first place, not all the spectators. If they are interested, they are free to come and watch what is done. They do NOT get to decide what is done and how. It is a "take it or leave it" deal tailored to the programmers, as opposed to the general public.

I estimate that during those 2½ years more than 20 chess programmers have raised their voice against the ICGA. You can begin with the 12 who criticized the Rybka verdict. IOW, don't say their is no problem.

I don't know if there still really is an interest among the new generation of chess programmers to have a real world championship (times have changed, the internet as main culprit), but *if* there still is an interest then the inelasticity has to change.

Why don't you raise the issue in a separate thread here? Ask exactly what their issues are/were? Most will likely address the Rybka issue and nothing else. Of those, the majority will probably mention the verdict as their primary complaint.

But raise the issue and see what you get. Forget this imaginary 20 programmer number. I'd only consider input from those that have actually ATTENDED a WCCC event at some point in time, i.e. a real participant, rather than those that have never experienced the events at all.

Democracy out, elitism in, you never cease to amaze me.

Do you ever read? "why don't you raise the issue here in a separate thread? Ask what the programmer's issues are." Does that sound "anti-democratic"? What is wrong with you?

You certainly never fail to amaze ME.

Robert Hyatt - Forget this imaginary 20 programmer number.

Robert Hyatt - I'd only consider input from those that have actually ATTENDED a WCCC event at some point in time, i.e. a real participant, rather than those that have never experienced the events at all.

There have to be a sane base for a discussion, with an exclusion attitude like yours it's unlikely it will be fruitful. But for a start check the CSVN rules.

http://www.csvn.nl/nl/nieuws-mainmenu-2 ... t-ict-2013

My take here is that if you haven't participated, you are not qualified to discuss rules that must be followed by participants. You won't see the reason for some of the rules unless you have actually "been there, done that, got the T-shirt."

I don't care what spectators want. They can make requests. If the requests don't damage the event given the perception of the participants, fine to do it. But I don't want a hundred demanding Houdini, Robolito, Firebird, Ippolit and who knows what else to be admitted.

The similarity test might be usable. But not to include or exclude. We already know it screws up on inclusions, based on the twin fish example. It seems certain enough that it can also have a false positive. My take on this, ... I'd prefer to take this up under the specific rule thread I started to keep all of this together, so more up there.

[bob]

You don't believe Robert, when he says it is fully original?

2. Any other programmer from whom a contribution is present in the program must be mentioned in the registration together with the nature of their contribution. The use of such contributions must (obviously) be legitimate.

If so, it seems that CSVN rule #2 is just as devastating for it. Do you really think Robert would ever be willing to admit there are contributions from others in Houdini?

The difference is that Houdini is never allowed to play in ICGA, in CSVN it has a chance even if Robert sticks to you know what. The origins of Houdini are public knowledge, the CSVN question is already answered.

I don't see the problem myself. It has ZERO business competing. But let's change that for a bit and say it is OK. Then what about Ippolit? Firebird? Robolito? How MANY of those do you let in? Every last modified version? You either open the door to chaos or keep it firmly locked to derivatives. With one exception practiced by the ICGA. You can enter a derivative of an open source program with permission of the original author, which means just ONE "family member" is allowed. Which do you choose for the ippolit family? Who is the original author? These need answers.

[bob]

If we look at engines that are in the top 20, then Hannibal, Protector, Spike, and Spark. And there are plenty of engines just outside the top 20 that could put in a good showing in a Swiss system tournament. Of course, my definition of a strong engine differs from that of others.

Well, Spark will be a tough call. Did you know its programmer lives in Leiden? Yet he doesn't conider it worth his time to participate there.

I don't know what is stopping Hannibal or Protector. We could ask their respective programmers. The default suspect is of course that it is too costly to attend the event.

On that subject, I think Bob's information if off. The total activities span only 7 days (including opening, conference and everything). But thhis time has to be divided between WCCC, WCSC and blitz tournament. Now in Yokohama it is true that they did this the stupid way, playing WCCC in the morning/afternoon, and the other in the afternoon/evening, so that both spanned 6 or 7 days. But because the two tournaments had exactly the same participants, no one objected or cared.

On the previous edition they followed my proposal to schedule the events after one another: 3 days of WCCC, blitz, 3 days of WCSC. Then participants only interested in the WCCC would only have to attend for 3 days. That still leaves time for eight 4-hour rounds (9:00-13:00, 13:30-17:30, 18:00-22:00) with lunch and pizza breaks, if you discount monday-morning for opening ceremony. Of course you could also squeeze in eleven 3-hour rounds: (9:00-12:00, 12:30-15:30, 15:30-18:30, 19:00-22:00).

All this is negociable, and based on past experience the ICGA seems to be pretty open-minded about it.

You are not thinking like a competitor. Let's say round 1 is on January 1, 2015.

It is typically a near 24-hour trip with airline connections and such. Do you want to arrive right before the round starts, hoping the hardware you shipped in works perfectly? No. Probably want to arrive on Friday at the latest, giving you Saturday to fix problems (since Sunday is generally a tough day to do any business. So now for our one week event, we just added 3 days on the front, travel most of Friday, get checked in, use saturday to test and setup, and then fix whatever might need fixing. Let's suppose the thing is exactly 7 days long with the usual closing banquet / awards presentation at the end. So you finish up Sunday night, stay up late to pack equipment and get it shipped back home, and then travel all day Monday. Tough to go to work on Tuesday. So add 2 days to the end. How long is that, in reality, now?

I've actually done these events. That was the nice thing about the ACM events. For the longest they were four rounds, later 5. We did two rounds Sunday, then one each monday, tuesday and Wednesday. 3 workdays gone.

We could leave Saturday, arrive Saturday, set up saturday, and be ready to go Sunday since most anywhere in the US to anywhere in the US is a half-day deal at worst. We have non-stop flights from Birmingham to most every major US city, for example.

Now, you compress the events as you suggested. But there is a conference there, and many participants present papers. The conference will likely NOT be scheduled just to coincide with the WCCC. How do any participants attend sessions and present papers? The conference is an integral part of the competition.

These things add up quicker than you might think

[hgm]

You either open the door to chaos or keep it firmly locked to derivatives. ...

That is not entirely true. If the original authors of an open-source program is not available (in theory there could be hundreds...) to decide which derivative is should be allowed to enter, you could let those that are aspirng to enter play a qualifier. Nothing chaotic about that.

The permission of the original author is already implied by the open-source license. Why would it be so important that he can actually be traced, to give it again? What if he died?

What if Komodo would want to enter the WCCC. Would the ICGA still require Don Dailey's signature on the registration form? Would they ask for a copy of his will, to see if it contains an explicit clause that spells out "I grant the beneficiaries of my estate to enter Komodo into the WCCC", and would it be barred if that cannot be produced?

[hgm]

You are not thinking like a competitor. Let's say round 1 is on January 1, 2015.

You seem to have missed the fact that I was there, in Yokohama 2013. And in Kanazawa 2010. And in Pamplona 2009. (And in Tilburg in 2011, but that hardly counts, as I drove there the morning the tourney started. And in Amsterdam 2007, where I just took the bycicle.)

It is typically a near 24-hour trip with airline connections and such.

That seems a gross exaggeration. To Yokohama was only an 11-hour (direct) flight. But of course I don't live out in the bush, but near a major international airport. But many people do.

Do you want to arrive right before the round starts, hoping the hardware you shipped in works perfectly? No. Probably want to arrive on Friday at the latest, giving you Saturday to fix problems (since Sunday is generally a tough day to do any business.

That sounds a bit like paranoia. Even if it was in New York I would just fly the day before. What could go wrong? I bring my laptop, to play remote over the internet, or even apply for a loan machine with the organizers. To Japan is indeed a bit more taxing. I did actually leave on Friday for Yokohama, but mainly because I was tardy in booking a flight, so that by that time it was the only one available. Had I planned better I could have left Saturday night.

Let's suppose the thing is exactly 7 days long with the usual closing banquet / awards presentation at the end.

Well, I just explained it could (and once did) last only 3 days. Skip the banquet if you don't think it is worth the time...

So you finish up Sunday night, stay up late to pack equipment and get it shipped back home, and then travel all day Monday. Tough to go to work on Tuesday. So add 2 days to the end. How long is that, in reality, now?

Sounds a lot like ENRON accounting.

Now, you compress the events as you suggested. But there is a conference there, and many participants present papers. The conference will likely NOT be scheduled just to coincide with the WCCC.

Again, your information is not up to date. The WCCC is conducted in parallel with the conference.

How do any participants attend sessions and present papers? The conference is an integral part of the competition.

Not anymore, apparently. Only the Olympiad is scheduled to not overlap with the conference.

[bob]

You either open the door to chaos or keep it firmly locked to derivatives. ...

That is not entirely true. If the original authors of an open-source program is not available (in theory there could be hundreds...) to decide which derivative is should be allowed to enter, you could let those that are aspirng to enter play a qualifier. Nothing chaotic about that.

The permission of the original author is already implied by the open-source license. Why would it be so important that he can actually be traced, to give it again? What if he died?

What if Komodo would want to enter the WCCC. Would the ICGA still require Don Dailey's signature on the registration form? Would they ask for a copy of his will, to see if it contains an explicit clause that spells out "I grant the beneficiaries of my estate to enter Komodo into the WCCC", and would it be barred if that cannot be produced?

OK, an acceptable alternative. "family" plays a tournament to choose the representative to the ICGA. Works for me, with one exception.

Let's take Crafty. Suppose the Rybka versions that were dead copies of Crafty wanted to enter. Do I get to enter or does Rybka, assuming it is stronger?

I don't think Komodo is an issue, as the "team survived". Don can't come back from the great beyond and try to enter a second copy under his name. I think that case is not a problem. Of course, there is another one that might be. Suppose you and I work together to create a strong chess program, then we get pissed off at each other and go our separate ways. NOW who gets to enter? That might be a show-stopper.

[bob]

You are not thinking like a competitor. Let's say round 1 is on January 1, 2015.

You seem to have missed the fact that I was there, in Yokohama 2013. And in Kanazawa 2010. And in Pamplona 2009. (And in Tilburg in 2011, but that hardly counts, as I drove there the morning the tourney started. And in Amsterdam 2007, where I just took the bycicle.)

It is typically a near 24-hour trip with airline connections and such.

That seems a gross exaggeration. To Yokohama was only an 11-hour (direct) flight. But of course I don't live out in the bush, but near a major international airport. But many people do.

After flying for 12 hours, do you then do another 12 hours of work? I generally eat and go to bed. End of day 1. As I said, for overseas trips, I consider travel time to be a day. To get to London, for example, I fly from here to NYC. 2-3 hours depending on whether I choose to pay a premium for non-stop or go thru atlanta/memphis cheaper. Then a wait in NY, and a 7+ hour flight to London. And then the second day comes around after I arrive, locate baggage, get to hotel, get checked in, eat and go to bed..

Do you want to arrive right before the round starts, hoping the hardware you shipped in works perfectly? No. Probably want to arrive on Friday at the latest, giving you Saturday to fix problems (since Sunday is generally a tough day to do any business.

That sounds a bit like paranoia. Even if it was in New York I would just fly the day before. What could go wrong? I bring my laptop, to play remote over the internet, or even apply for a loan machine with the organizers. To Japan is indeed a bit more taxing. I did actually leave on Friday for Yokohama, but mainly because I was tardy in booking a flight, so that by that time it was the only one available. Had I planned better I could have left Saturday night.

Let's suppose the thing is exactly 7 days long with the usual closing banquet / awards presentation at the end.

Well, I just explained it could (and once did) last only 3 days. Skip the banquet if you don't think it is worth the time...

So you finish up Sunday night, stay up late to pack equipment and get it shipped back home, and then travel all day Monday. Tough to go to work on Tuesday. So add 2 days to the end. How long is that, in reality, now?

Sounds a lot like ENRON accounting.

Now, you compress the events as you suggested. But there is a conference there, and many participants present papers. The conference will likely NOT be scheduled just to coincide with the WCCC.

Again, your information is not up to date. The WCCC is conducted in parallel with the conference.

How do any participants attend sessions and present papers? The conference is an integral part of the competition.

Not anymore, apparently. Only the Olympiad is scheduled to not overlap with the conference.

I haven't been to one in a while. We did not have games overlapping the conference. For obvious reasons.

BTW if you look at past ICGA events, not ALL were 3 day tournaments. There were 9 day tournaments. There were 7 day tournaments. That it is more compressed is a good thing, but it is still a non-trivial task to take off a week from classes...

[hgm]

OK, an acceptable alternative. "family" plays a tournament to choose the representative to the ICGA. Works for me, with one exception.

Let's take Crafty. Suppose the Rybka versions that were dead copies of Crafty wanted to enter. Do I get to enter or does Rybka, assuming it is stronger?

Well, definitely the right to participate should be reserved for LEGAL derivatives. As I understand it the license of Crafty explicitly forbids people to run any derivatives they would make of it in Chess tournaments. So although it is open source, people could never claim you gave them permission to enter it in the WCCC through the license. That is different for GPL code, where it would actually be a violation of a (co-)authors to refuse permission to enter it.

I don't think Komodo is an issue, as the "team survived". Don can't come back from the great beyond and try to enter a second copy under his name. I think that case is not a problem.

I fully agree. But I mentioned it, because I see a great analogy with the case of an open-source project, of which the original author is no longer available. If people are still actively developing Toga, you could also say thet the team survived.

Of course, there is another one that might be. Suppose you and I work together to create a strong chess program, then we get pissed off at each other and go our separate ways. NOW who gets to enter? That might be a show-stopper.

Well, that can already happen under the current interpretation of rule #2. But the solution can be the same. If we cannot agree, we play a qualifier match. If the one that wins that qualifier later wins the title, we still share that title. It would not be allowed to hide the fact that we had a co-author, not even it is one we are pissed off at.

[bob]

OK, an acceptable alternative. "family" plays a tournament to choose the representative to the ICGA. Works for me, with one exception.

Let's take Crafty. Suppose the Rybka versions that were dead copies of Crafty wanted to enter. Do I get to enter or does Rybka, assuming it is stronger?

Well, definitely the right to participate should be reserved for LEGAL derivatives. As I understand it the license of Crafty explicitly forbids people to run any derivatives they would make of it in Chess tournaments. So although it is open source, people could never claim you gave them permission to enter it in the WCCC through the license. That is different for GPL code, where it would actually be a violation of a (co-)authors to refuse permission to enter it.

I don't think Komodo is an issue, as the "team survived". Don can't come back from the great beyond and try to enter a second copy under his name. I think that case is not a problem.

I fully agree. But I mentioned it, because I see a great analogy with the case of an open-source project, of which the original author is no longer available. If people are still actively developing Toga, you could also say thet the team survived.

Of course, there is another one that might be. Suppose you and I work together to create a strong chess program, then we get pissed off at each other and go our separate ways. NOW who gets to enter? That might be a show-stopper.

Well, that can already happen under the current interpretation of rule #2. But the solution can be the same. If we cannot agree, we play a qualifier match. If the one that wins that qualifier later wins the title, we still share that title. It would not be allowed to hide the fact that we had a co-author, not even it is one we are pissed off at.

The issue I was talking about is not with the crafty license, but with a perfectly "legal" derivative of Crafty (pretend the EULA does not exist and it is GPL). The ICGA maintains a no-derivative rule. So which gets first right of refusal, original or derivative? I don't want N copies of stockfish in a single event any more or less than I don't want N copies of ippolit or another open-source program.