18 days from SF4 release and about ~30+ ELO gain!

[lucasart]

Indeed, the pace of Stockfish improvement is amazing, the development framework constructed by Gary is awesome.
Clearly no individual or two-person team can keep up with this in the long run, so this could mean the end of commercial chess engines as we currently know them. Maybe in 2 years time only Stockfish and derivates will continue to be developed.

Robert

+1

[Adam Hair]

I have to concede the point that ideas are plentiful enough and that computer resources are still more important for the pace in the development of a chess engine (even though testing of ideas is much more efficient now than in the past).

[lucasart]

I believe 'ideas' are fairly easy to come by, for ex:
bonus for rook behind passed pawn
penalty for undefended piece
etc.

The difficult part is the testing...

You hit the nail on the head. As I already mentioned in another post, testing is a SERIOUS bottleneck for us, ideas are not.

In a way this is sad for me - I feel that I can compete with anyone with ideas, but I cannot compete with testing horsepower which is really where it's at. It seems like chess has been locked forever in a horsepower race. It used to be whoever has the biggest machine wins the tournament, now it's whoever has the most testing resources. I am considering switching over to something that will reward creativity a lot more.

Try Go! Last time I looked, the best humans still spanked computers in Go.

[Don]

I believe 'ideas' are fairly easy to come by, for ex:
bonus for rook behind passed pawn
penalty for undefended piece
etc.

The difficult part is the testing...

You hit the nail on the head. As I already mentioned in another post, testing is a SERIOUS bottleneck for us, ideas are not.

In a way this is sad for me - I feel that I can compete with anyone with ideas, but I cannot compete with testing horsepower which is really where it's at. It seems like chess has been locked forever in a horsepower race. It used to be whoever has the biggest machine wins the tournament, now it's whoever has the most testing resources. I am considering switching over to something that will reward creativity a lot more.

Try Go! Last time I looked, the best humans still spanked computers in Go.

I think Go is open for much more creativity and probably more lucrative. I had a palm program I wrote for GO and marketed and it did amazingly well so there is probably a lot more money in Go too. Chess programs are a dime a dozen - in one rating list I saw there are several THOUSAND chess programs although some are different versions of the same program.

Don

[Henk]

If that is the case. Chess programming should be no more than a hobby. Problem is that it often takes too much time from me. Feel like an addict sometimes.

[Don]

If that is the case. Chess programming should be no more than a hobby. Problem is that it often takes too much time from me. Feel like an addict sometimes.

It's difficult to make a living from computer chess and if you are in computer chess either as a hobby or to make a living you have to do it out of love of the game.

If you look on the web there are people who are bigger losers than us though - for example there is a contest to make the loudest noise with vehicle sound systems - they call it db drag racing, they devote most of their time and money to building these sound system - not even to play music - just to make noise:

http://www.cnn.com/2003/TECH/ptech/07/0 ... ereo.kill/

You have to be thinking, "get a life" or "this guy has too much time on his hands." I know that there are people who think that about us!

There are things far more important to me than computer chess, but I am hooked too and I do love computer chess. I devote part time to it now but I could not justify even that if I could not at least supplement my income with it.

[JuLieN]

If that is the case. Chess programming should be no more than a hobby. Problem is that it often takes too much time from me. Feel like an addict sometimes.

It's difficult to make a living from computer chess and if you are in computer chess either as a hobby or to make a living you have to do it out of love of the game.

If you look on the web there are people who are bigger losers than us though - for example there is a contest to make the loudest noise with vehicle sound systems - they call it db drag racing, they devote most of their time and money to building these sound system - not even to play music - just to make noise:

http://www.cnn.com/2003/TECH/ptech/07/0 ... ereo.kill/

You have to be thinking, "get a life" or "this guy has too much time on his hands." I know that there are people who think that about us!

There are things far more important to me than computer chess, but I am hooked too and I do love computer chess. I devote part time to it now but I could not justify even that if I could not at least supplement my income with it.

And let's not forget that having their engines compete is the closest nerds have to an excuse for pretending to do sport.

[Mithu]

There is a strange trend seen by commercial authors. There are two types of commercial authors - the ones who are silent and leave everyone guessing and the ones who are outspoken and share whatever they know or experienced (though the veracity of it is a debatable issue ).

Now what are the issues that we are kept in the dark ?

1. The actual progress or even a roadmap of targets achieved.
2. The schedule of work or time that is being given on development.
3. Whether they ARE looking at the development or if they have stalled it altogether.
4. The improvements that are made. There is no mention of what improvements or code changes or even deletions that took place though whatever is made known is more for marketing.

This silence appears to be more pronounced once the independant commercial authors are 'roped' in by the big bosses who drive a commercial juggernaut.

I think all the above facts may have been seen by all of us in the case of many commercial authors. And one more marketing thing is the silent build up to a release. All of a sudden a guy extolls the virtues of X-engine and you can slowly see a group of new posters suddenly asking a lot of questions and building up the crescendo so to say! Well so much for reading a detective novel...

It is for this reason I have respect for Don and others like him compared to the silent ones who seem to have veiled agendas... atleast so it seems.

I for one would like to buy from Don or Mark (HCE) than from Stefen or Shay for this very reason!
And its the same even for a GUI!

[Don]

There is a strange trend seen by commercial authors. There are two types of commercial authors - the ones who are silent and leave everyone guessing and the ones who are outspoken and share whatever they know or experienced (though the veracity of it is a debatable issue ).

Now what are the issues that we are kept in the dark ?

1. The actual progress or even a roadmap of targets achieved.
2. The schedule of work or time that is being given on development.
3. Whether they ARE looking at the development or if they have stalled it altogether.
4. The improvements that are made. There is no mention of what improvements or code changes or even deletions that took place though whatever is made known is more for marketing.

This silence appears to be more pronounced once the independant commercial authors are 'roped' in by the big bosses who drive a commercial juggernaut.

I think all the above facts may have been seen by all of us in the case of many commercial authors. And one more marketing thing is the silent build up to a release. All of a sudden a guy extolls the virtues of X-engine and you can slowly see a group of new posters suddenly asking a lot of questions and building up the crescendo so to say! Well so much for reading a detective novel...

It is for this reason I have respect for Don and others like him compared to the silent ones who seem to have veiled agendas... atleast so it seems.

I for one would like to buy from Don or Mark (HCE) than from Stefen or Shay for this very reason!
And its the same even for a GUI!

You don't know how much saying that means to me and I appreciate it. I sometimes feel that I get kicked in the gut for my contribution - people often question my motives, my honesty, my character and in some ways I have made myself a target.

[M ANSARI]

It is interesting that so much progress can still be made in evaluation. I guess evaluation can be the great equalizer when it comes to hardware. The latest version of Houdini or Stockfish would easily be able to beat Deep Blue with its old software and massive hardware (even by today's standards) with one core. But evaluation improvements will eventually reach a wall, but this wall seems still a couple of hundred ELO's away. I think this shows that Vas had it right with Rybka, and the "secret" to his massive progress with Rybka was in the system he used to test millions of games very quickly. This was improved on with Houdini, and it seems like it is taken to another level with Stockfish.

Eventually I think progress will be much faster if an engine is separated into modules as this would allow progress to be made in areas where chess engines are still weak. Endgames are still a weakness, and if you take away the dramatic hardware progress, you would find that software wise ... not much progress was made there. An endgame module that recognizes when an endgame starts will be able to understand how to avoid fortresses if it has an advantage (or look for it if it has a disadvantage) will understand which rook endgames are still winning even a pawn up, will understand when exchanging to an opposite bishop ending still allows a win etc... This module would make use of super fast daughter cards that allow for on the fly MC calculation, and would have fast access to EGTB's or bitbases without affecting the core engine in speed.

So there is still a lot of progress to be made even if the gains in evaluation reach an end. Already engine evaluation is at a level that a human cannot understand, looking at some of the Stockfish vs. Houdini LTC games, I cannot help but be amazed at how little the moves make sense. I can follow a game by the top super GM's and understand most of the moves, but with the latest engines I can really say that I have no clue on earth of what is going on, and most likely even the top GM's would also be clueless. But in endgames I can still see when these top engines make a silly mistake or exchange into a no win endgame when they have an advantage.