Not logged inRybka Chess Community Forum
Up Topic The Rybka Lounge / Computer Chess / Why is an source engine so superior to Komodo and Houdini?
- - By rocket (****) Date 2019-07-04 14:58
The code is out in the public, so why on earth isn't it at least tied with the best commercial programs? Isn't it a bit embarassing?

CCRL
Against Houdini 6 40/40
+11−0=39

Against Komodo 11.3.1 40/40
+13−2=33
Parent - - By Labyrinth (*****) Date 2019-07-04 20:27
Don't know why it would be embarrassing. An entire community of people are working on Stockfish, coding and testing. Houdini is a solo project, and Komodo has a very small team with really only one actual programmer.
Parent - - By rocket (****) Date 2019-07-05 08:41
Because everything about the program is out the in the public?

If your rival is open to you with all his tricks, and yet you can't beat him a single time, I'd say it's embarassing.
Parent - By Labyrinth (*****) Date 2019-07-05 16:29

>If your rival is open to you with all his tricks, and yet you can't beat him a single time, I'd say it's embarrassing.


I think that you just don't know what you're talking about, at least not enough to justify throwing around the term 'embarrassing'.

The 'tricks' here are heuristics and code optimizations that work with sf's codebase, adapting them to other programs isn't trivial, and even if you do, due to the holistic nature of chess programs it might not even help.

It's also against the law to take code from an open source program and use it in a commercial, closed source program depending on the license, and I believe stockfish uses such a license.

I recommend learning some code as it will give you a lot more respect for the kind of work these guys do.
Parent - - By user923005 (****) Date 2019-07-05 17:58
The Stockfish team has enormous resources.
There is an army of programmers, testing new ideas.
There is a cavernous mass of compute power pounding away around the clock.
The commercial engines don't even have 10% of those resources.
A better question is, how do the commercial engines, with such limited resources, stay competitive?

Sure, they could copy SF and even legally sell it.  But why would you pay for something you can get free?
And any copied stuff would still have the same problems as the original setup: SF has the biggest pile of resources.
There is a similar effort for LC0.  That is why LC0 also advances rapidly.

But the commercial systems offer things that make them worthwhile.
Secondary opinions on long term analysis.
Better tactical analysis.

In a similar way, other open source engines are not in the same strength bracket but are also extremely useful.
A good example is Sting.  It is the only engine that can solve some studies in a reasonable period of time.

The commercial systems have nothing to be ashamed of.  In fact, both Houdini and Komodo have been the strongest engines in the world at some particular points in time.
To compete with free labor and free compute power and still do well is very remarkable.

So why not simply celebrate them all.
To write an ultra-simple chess program that can even legally play chess is quite an achievement.
To write a world class engine is something that a small handful of gifted people can accomplish.
Nothing to belittle, so far as I can see.
Parent - - By rocket (****) Date 2019-07-06 20:10
Stockfish tactical is weaker than Houdini tactical?
Parent - By user923005 (****) Date 2019-07-06 21:03
At problem solving, Houdini Tactical is best for a large collection of problem types.
Sting also has areas of excellence.
if you give engines enough time, it can ameliorate tactical weaknesses to a degree.
Parent - By user923005 (****) Date 2019-07-05 18:02
The Stockfish team has enormous resources.
There is an army of programmers, testing new ideas.
There is a cavernous mass of compute power pounding away around the clock.
The commercial engines don't even have 10% of those resources.
A better question is, how do the commercial engines, with such limited resources, stay competitive?

Sure, they could copy SF and even legally sell it.  But why would you pay for something you can get free?
And any copied stuff would still have the same problems as the original setup: SF has the biggest pile of resources.
There is a similar effort for LC0.  That is why LC0 also advances rapidly.

But the commercial systems offer things that make them worthwhile.
Secondary opinions on long term analysis.
Better tactical analysis.

In a similar way, other open source engines are not in the same strength bracket but are also extremely useful.
A good example is Sting.  It is the only engine that can solve some studies in a reasonable period of time.

The commercial systems have nothing to be ashamed of.  In fact, both Houdini and Komodo have been the strongest engines in the world at some particular points in time.
To compete with free labor and free compute power and still do well is very remarkable.

So why not simply celebrate them all.
To write an ultra-simple chess program that can even legally play chess is quite an achievement.
To write a world class engine is something that a small handful of gifted people can accomplish.
Nothing to belittle, so far as I can see.
Parent - - By gsgs (***) Date 2019-07-05 05:04 Edited 2019-07-05 05:19
you can't just copy itand sell it due to the licence

-----------edit------------------------------

well, I was wrong. You can add some stuff , recompile it rename it and sell it.
But you have to publish the source. I wonder, why apparently noone is already
doing this. If they don't want to publish the source, they could obfuscate it
and add some user- and time-dependent keywords to make copying harder.
Or even licence it somehow, while still publishing the source , I don't know

> Stockfish is free, and distributed under the GNU General Public License version 3 (GPL v3).
> Essentially, this means that you are free to do almost exactly what you want with the program,
> including distributing it among your friends, making it available for download from your web site,
> selling it (either by itself or as part of some bigger software package), or using it as the starting
> point for a software project of your own.
>
> The only real limitation is that whenever you distribute Stockfish in some way, you must always
> include the full source code, or a pointer to where the source code can be found. If you make
> any changes to the source code, these changes must also be made available under the GPL.

So, what's the strongest "public domain" chess-engine without licence ?
Parent - By user923005 (****) Date 2019-07-05 06:47
Ivanhoe
Parent - By Uly (Gold) Date 2019-07-05 08:14
This is like asking "why aren't all free engines as strong as Stockfish?" Getting an engine up there is really hard, it doesn't matter if it's commercial or open source. People are having to reinvent the wheel. People have succeeded in making engines stronger than Stockfish by starting from a Stockfish base, so others still consider them "Stockfish", and being at the top requires continue development, so those projects eventually get caught up, and surpassed, by Stockfish dev, so some people just add their changes to the latest source code.

Why don't you go and create an engine from scratch that is as strong as Stockfish? The code is out there in the public, asking you to do it isn't different than asking Komodo's* or Houdini's** developers to do it, that they're commercial isn't different.

* ) I think Larry Kaufman answered this at some point. The reason Komodo isn't stronger than Stockfish despite the source being open is that they have tried all those things from Stockfish that make it so strong in Komodo, but it makes it play weaker, so they only kept what made it play best. Every program is different, and actually, that's a reason new chess programmers keep popping up, because there's no "right way to do things", and not even Stockfish has them right (if it did, Stockfish from 2020 wouldn't be much stronger than what we have now)

** ) I know Houdini wasn't started from scratch. But can you get some already existing open source code engine and make it stronger than Stockfish like that? No, no, I'm not mocking, go ahead and do it, but until then, commercial engines aren't at least tied for best for the same reasons nobody else is tied for best with Stockfish.
Up Topic The Rybka Lounge / Computer Chess / Why is an source engine so superior to Komodo and Houdini?

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill