Nostalgia

[zullil]

Besides, this whole competition spirit is pathetic. When someone takes something from DiscoCheck, I am proud to have contributed. I'm not envious, and feel they are getting an illegal competitive advantage by stealing my "intellectual property" (and don't even get me started on the concept of "intellectual property" it's such propaganda from the software patent and copyright lobbies)

Would you feel this way if DiscoCheck were a commercial product, upon which you depended---at least in part---for income?

That would be a pretty stupid business model. No one is going to buy a 2750 elo engine, when they can get several 3100+ elo ones for free.

There is no money to be made by writing a strong engine. Perhaps it was the case 10 years ago, but not anymore. You cannot live from the sales of your chess engine today.

If you want to make money with chess software, perhaps you should write a GUI instead. And ideally for MacOSX, or IOS, as users of these systems don't have much of a choice, and are used to the concept of proprietary and commercial software.

Denying my hypothetical is not quite the same as answering my question. My point is that some individuals do choose (bad business model or not) to sell their engines, and obviously these individuals have a reason---which has nothing to do with whether or not the engines are original---for not revealing their code.

That said, I'm a fan of open-source programs, and I'm glad you have chosen to make DiscoCheck open to all.

[chetday]

I think one has to make a distinction between the programmers' point of view and end users' point of view.

Other than programming several all text adventure games in a variation of Basic for an Apple IIe BBS system (GBBS) in 1986, I'm not a programmer.

But I am an interested end user of chess engines, and, for me, the ideal engine would be one that played attacking chess with gambits and surprises and wins and losses. Someone posted the other day that the ideal engine for them would be one that played Morphy-like games, and that comment hit me like a tire iron between the ears because that's what I'd like to have, too.

Don't get me wrong, I also enjoy the other engines I'm playing with and against -- both free and commercial. I bought Houdini 3 because it's the strongest out there and I bought Komodo 5 to support Don because I'm learning a lot from reading his posts about engine construction as well as the engineering and philosophy that goes behind them.

Calling me a chess patzer is generous, so engine ELO really doesn't matter a hoot to me personally. But I can certainly say that it would were I trying to program an engine since ELO seems to be the brass ring to goal set for.

As for going back to the old days, I think of Jay Gatsby in F. Scott Fitzgerald's wonderful novel, who, when told he can't repeat the past, replies, "Can't repeat the past. Of course you can."

And then the rest of the novel moves inexorably to prove that no, sorry, you can't repeat the past... just as you can never really go home again.

[lucasart]

Nostalgia

I yearn to have those ancient days back when chess engine authors used to write their source code themselves. You braced yourself, launched an editor, created a blank file and personally wrote "#include <stdio.h>" in the upper left corner.
How can we go back there ?

Matthias.

Go back for what ?
Just to reinvent the F****** wheel from scratch ?
It's just a waste of time.
It is easier, faster and better to take one of top free open source engines and add some ELO to it, or to take more of top engines and mix them together into a single one, exactly like Vas and Robert did, they just forgot to release the source code

And life goes on

Have you ever written your own chess engine ?

If not, then why do you talk about what you don't know with so much assurance ?

[carldaman]

That would be a pretty stupid business model. No one is going to buy a 2750 elo engine, when they can get several 3100+ elo ones for free.

There is no money to be made by writing a strong engine. Perhaps it was the case 10 years ago, but not anymore. You cannot live from the sales of your chess engine today.

If you want to make money with chess software, perhaps you should write a GUI instead. And ideally for MacOSX, or IOS, as users of these systems don't have much of a choice, and are used to the concept of proprietary and commercial software.

I don't necessarily agree with the first point. If the play is exciting and imaginative and preferably the price not too high, I'd buy and (have recently bought) 2750-rated engines.

This is where much of the vision is lacking. Take your GUI suggestion, add great analysis features, graphics, etc. and combine it with an elegant playing style from your engine and you will have an attractive package.

I remember in the old days, it looked like CS-Tal was going in this direction, but all of sudden development ceased for good. All there was needed were additional GUI improvements. The engine wasn't the strongest but played very interesting chess, especially for those days. Now we have much stronger engines, but the GUIs are lagging behind.

We recently had a discussion about engines with unique playing styles. The problem here lies with a good portion of the chess software consumers, (most players are not that strong, either, which doesn't help) who want the strongest possible engine. Why they would be so obsessed with that I don't fully understand, since most may benefit more from something that would actually help them play better chess (i.e. find better plans, ideas etc), and a weaker but 'smarter' engine could be better suited for that.

This in turn has influenced many engine developers to seek out more strength at all cost (nothing wrong with that goal in itself), while sacrificing stylistical considerations for the most part.

If you look at a tweaked engine personality like Zappa Dissident Aggressor (=DA for short), I'm estimating that roughly 400 points had to be "lost" (!) to achieve this exciting and fascinating style of play. This wasn't even done by the Zappa programmer, but by an end user of the product, Robert Flescher, who spent countless hours at it and was then kind enough to share his findings with the community.

Was it worth "losing" 400 points off Zappa's normal strength? Sounds crazy, but I think it was, since it leaves us with a quite strong 2500-2600 engine that can find amazing ideas, not all sound, but then again so were Tal's moves often in his prime.

So, using this example, here is a case of 'less is more'. Imagine losing so much strength and ending up with an even more appealing product! This is a startling paradox, that provides some food for thought. In this regard, less strength doesn't necessarily mean a worse product, and vice versa. And, of course, one can use a full strength program and a wild personality simultaneously. It's not like strength should be abandoned altogether, and I'm not arguing for that.

I believe more imaginative vision is required in the community. I see a few who have it already. I've heard it several times -- many would-be customers would pay a reasonable price for such features (GUI + engine).

Regards,
CL

[carldaman]

I think one has to make a distinction between the programmers' point of view and end users' point of view.



But I am an interested end user of chess engines, and, for me, the ideal engine would be one that played attacking chess with gambits and surprises and wins and losses. Someone posted the other day that the ideal engine for them would be one that played Morphy-like games, and that comment hit me like a tire iron between the ears because that's what I'd like to have, too.

Don't get me wrong, I also enjoy the other engines I'm playing with and against -- both free and commercial. I bought Houdini 3 because it's the strongest out there and I bought Komodo 5 to support Don because I'm learning a lot from reading his posts about engine construction as well as the engineering and philosophy that goes behind them.

I'm interested in the same ideal type(s) of engine myself. I mean, once I've gotten the top 5 engines, and a few other very strong ones, I have to recognize that something is still missing from the big picture.

There are two types of engines I would like to see more of.

1. A wild and crazy attacking style, closely emulating a Tal or Morphy like player, with a high contempt setting and swindle capability

or

2. A knowledge-laden program geared towards strategic/positional chess, strong in the endgame, but weaker tactically and still having the contempt setting and swindle feature, as above


I would actually be interested in rather spending money on such efforts than on the next super-engine (seriously!). Think about it, we're "swimming" in strength, but we're short on style. In my other post in this thread, I gave an example where a whopping 400 rating points could be sacrificed to achieve a credible "Tal"-like personality for most players.

I'm saying this because programmers/deveopers can easily be swayed by the consumer. If we want a more appealing and useful product, we must speak up.

Regards,
CL

[Evert]

If my goal was to write a "top engine", I would indeed be wasting my time, and I should have quit a few years back. Needless to say, that is not my goal per se.

Taking a top open source engine and try to understand the source well enough to try to add some Elo points to it to me is much less fun than tinkering with my own program, and thus would be a massive waste of time.

Given the choice between trying to add 2 Elo to an already strong program, or trying to add 100 Elo to my own program I'll take the latter. I have more fun that way.

[Lavir]

Was it worth "losing" 400 points off Zappa's normal strength? Sounds crazy, but I think it was, since it leaves us with a quite strong 2500-2600 engine that can find amazing ideas, not all sound, but then again so were Tal's moves often in his prime.

So, using this example, here is a case of 'less is more'. Imagine losing so much strength and ending up with an even more appealing product! This is a startling paradox, that provides some food for thought. In this regard, less strength doesn't necessarily mean a worse product, and vice versa. And, of course, one can use a full strength program and a wild personality simultaneously. It's not like strength should be abandoned altogether, and I'm not arguing for that.

I believe more imaginative vision is required in the community. I see a few who have it already. I've heard it several times -- many would-be customers would pay a reasonable price for such features (GUI + engine).

Regards,
CL

What you say is perfectly common sense for what it concerns typical business models, when variety and choice is usually the best thing to do. The problem is that chess is not a common business market at all; the market is of too much niche to really accommodate for different models in itself and be able to sustain those.

The vast majority of users, since the "sport" is of so niche appeal, will naturally care more for the most attractive part of a sport, i.e. the performance of the "athletes" and much less on other various particulars on the same. There could indeed be other models inside, but they are representing too little people inside an already very restricted market to have some business value.

For an author to being motivated to do something as that it is just sort of "art for art's sake" more than anything else.