The Future of Hiarcs

This forum is for general discussions and questions, including Collectors Corner and anything to do with Computer chess.

Moderators: Harvey Williamson, Steve B, Watchman

Forum rules
This textbox is used to restore diagrams posted with the fen tag before the upgrade.
Post Reply
User avatar
Fernando
Admiral of the Fleet
Posts: 3059
Joined: Tue Jul 31, 2007 4:35 pm
Location: Santiago de Chile

The Future of Hiarcs

Post by Fernando »

I know this is not perhaps the best moment to talk of a relatively disappointing performance, specially if there is a commercial aspect to keep in mind, but I would like very much to hear Mark and Harvey to explain us how they see the result, how they are going to fix the issues, etc.
BTW, I still consider as I do since lot of time that the sheer aspect of strength is not anymore so interesting for human players that, in any case, are overpowered by ANYONE of the 50 best engines, commercial or amateur.
At the same time, the field or terrain for a more friendly approach doing of a chess program a real partner, teacher, friend etc is open. Fritz is more or less of that kind, BUT there are still many features that nobody has considered, so engaged has been every programmer in the strength factor alone. Yes, I Know I am and I had been a little bit obsessive with this, but there you have it, again...

Fern, still waiting Godot....
Festina Lente
Celadus
Member
Posts: 256
Joined: Sun May 06, 2012 11:29 am

Re: The Future of Hiarcs

Post by Celadus »

Fernando wrote:I know this is not perhaps the best moment to talk of a relatively disappointing performance, specially if there is a commercial aspect to keep in mind, but I would like very much to hear Mark and Harvey to explain us how they see the result, how they are going to fix the issues, etc.
BTW, I still consider as I do since lot of time that the sheer aspect of strength is not anymore so interesting for human players that, in any case, are overpowered by ANYONE of the 50 best engines, commercial or amateur.
At the same time, the field or terrain for a more friendly approach doing of a chess program a real partner, teacher, friend etc is open. Fritz is more or less of that kind, BUT there are still many features that nobody has considered, so engaged has been every programmer in the strength factor alone. Yes, I Know I am and I had been a little bit obsessive with this, but there you have it, again...

Fern, still waiting Godot....
I agree with your comments. And in my humble opinion Fritz is still the best
regarding teaching. Blunder check and full analysis are without equal.
User avatar
Harvey Williamson
Site Admin
Posts: 6079
Joined: Sun Jul 29, 2007 6:57 am
Location: Media City, UK
Contact:

Re: The Future of Hiarcs

Post by Harvey Williamson »

Fernando wrote:I know this is not perhaps the best moment to talk of a relatively disappointing performance, specially if there is a commercial aspect to keep in mind, but I would like very much to hear Mark and Harvey to explain us how they see the result, how they are going to fix the issues, etc.
BTW, I still consider as I do since lot of time that the sheer aspect of strength is not anymore so interesting for human players that, in any case, are overpowered by ANYONE of the 50 best engines, commercial or amateur.
At the same time, the field or terrain for a more friendly approach doing of a chess program a real partner, teacher, friend etc is open. Fritz is more or less of that kind, BUT there are still many features that nobody has considered, so engaged has been every programmer in the strength factor alone. Yes, I Know I am and I had been a little bit obsessive with this, but there you have it, again...

Fern, still waiting Godot....
I am not sure there is such a big issue from such a small sample of games. However this was not a good performance from Hiarcs. We did drop half points in games we should have drawn or even won.

The nice thing about a bad result is that it makes us look at all the games and see what went wrong.

Perhaps the best thing about these kind of face to face tournaments is that we actually watch the games and see where things go wrong. One thing all testers are, probably, guilty of is just looking at overall results from 100s of games without studying the games.
User avatar
Mark Uniacke
Hiarcs Author
Posts: 1458
Joined: Sun Jul 29, 2007 1:32 pm
Location: United Kingdom
Contact:

Post by Mark Uniacke »

It is easy to read too much into just a few games especially with an unfinished experimental engine in the tournament with vastly different hardware, but we did encounter some issues I was not aware of.

Our time usage was not good due to a number of factors and led to an unnecessary shortage of time in many games in the late endgame.

Our king safety algorithm has some issues.

We had a search instability issue which seemed to surface when thinking for long periods especially where positions were transposing alot, and is related to the hash table. It had a knock on effect of causing extra time usage and occasional subpar moves.

I believe these factors alone were responsible for us throwing a number of half points.

Such results while disappointing do act as a strong motivator so the impact on the engine will be positive.
Best wishes,
Mark

https://www.hiarcs.com
User avatar
Fernando
Admiral of the Fleet
Posts: 3059
Joined: Tue Jul 31, 2007 4:35 pm
Location: Santiago de Chile

Post by Fernando »

I hope, Harvey and Mark, that my post will NOT be taken by you as a funeral discourse or something like that. No doubt you will extract some lessons about it and Hiarcs will come again, even stronger. Nevertheless and no matter if we see on this result a great or a little issue, my point is, I believe, still valid. Some extra effort even beyond the explorer GUI could and should be done to make of H. a truly amiable person full of features to this date not delivered by any commercial program to date. .

Always yours
Fern
Festina Lente
LoveChessAH
Member
Posts: 17
Joined: Mon Aug 05, 2013 3:48 pm

Post by LoveChessAH »

I agree to what Fernondo wrote - also understood the message from Harvey - of course the message from Mark.

I whish for good improvement to next Explorer and good performance for next HIARCS.
I think also it is not easy for HIARCS Team - don´t how many people are working therefore and special tester they have.

On the other side i am surprise how much chess player are using more than 16 cores - also a point to think about for future criteria - improve chess knowledge or optimizing chess speed?

by the way - good luck to HIARCS Team.
User avatar
mclane
Senior Member
Posts: 1600
Joined: Sun Jul 29, 2007 9:04 am
Location: Luenen, germany, US of europe
Contact:

Post by mclane »

I do trust in mark.

Computerchess has become very difficult in our days.

The fruit revolution, the Komodo/stockfish fight, the situation that stockfish is open source and ANY competing programmer can cannibalize ideas out of it makes it difficult to continue your OWN way of doing things. It somehow destroys creativity . I hope mark can cope with the many difficult problems.

Nobody has an interest to see all programs be clones of fruit, clones of stockfish or rybka.

Computerchess can IMO only continue if the computer programs remain individual entities.

It makes no sense to see many similar clones invade the scene.
Fans would not like it.

We have a similar situation in car business.

If all cars are same, it makes no sense that they all exist. Only if the cars concentrate on certain special areas their existence makes sense.
What seems like a fairy tale today may be reality tomorrow.
Here we have a fairy tale of the day after tomorrow....
User avatar
Fernando
Admiral of the Fleet
Posts: 3059
Joined: Tue Jul 31, 2007 4:35 pm
Location: Santiago de Chile

Post by Fernando »

mclane wrote:I do trust in mark.

Computerchess has become very difficult in our days.

The fruit revolution, the Komodo/stockfish fight, the situation that stockfish is open source and ANY competing programmer can cannibalize ideas out of it makes it difficult to continue your OWN way of doing things. It somehow destroys creativity . I hope mark can cope with the many difficult problems.

Nobody has an interest to see all programs be clones of fruit, clones of stockfish or rybka.

Computerchess can IMO only continue if the computer programs remain individual entities.

It makes no sense to see many similar clones invade the scene.
Fans would not like it.

We have a similar situation in car business.

If all cars are same, it makes no sense that they all exist. Only if the cars concentrate on certain special areas their existence makes sense.

No doubt. And precisely making AN EXTRA effort in the frills, features, etc is the way you get that special thing you ask.
Due to the development of chess AI every chess engine is more or less similar because the search techniques are kind of a commodity, still leaving room to personal variation, but less and less, so by definition the area to explore is the other one, the cocoon of the engine..
Hiarcs explorer is a great GUI, BUT, let me be frank, is more like something for a pro than for me.
In terms of sheer fun, which is what I believe most of us search for, an old and bizarre gui like that of Chess System tal I is far more funny and exciting that Hiarcs explorer. I am NOT interested, not being a professional chess player but just an expert level amateur, in studying databases and things like that. I want to be ENTERTAINED and for that purpose I play chess. Guess than 90% of chess player are like me. In fact, 100% in this site of dedicated units collectors and players are like that or they would not be collector of such weak engines surrounded by wood or plastic. .

F
Festina Lente
User avatar
IA
Member
Posts: 433
Joined: Wed Sep 10, 2008 1:12 pm
Location: Spain

Post by IA »

I really liked when I was in Hiarcs ChessBase and you could buy separately the UCI engine without Interface Explorer.

I'm disappointed not to have Hiarcs on the Android platform.... :cry:


All these things together with the great competition there have now gotten Hiarcs has been relegated as the main engine of chess.

I hope Hiarcs lift the flight the next few years and I've always been a fan of this fantastic engine, but is not currently in any of the platforms that I like.

Anyway keep me Hiarcs admiration for the Team and for the good times I spent with the engine Hiarcs (Palm and “PC UCI and ChessBase”) and wish them luck.

Regards Mark and Harvey. :wink:
gsgs
Member
Posts: 40
Joined: Thu May 26, 2016 1:54 pm
Location: hilter,germany
Contact:

Re: The Future of Hiarcs

Post by gsgs »

Harvey Williamson wrote:
Fernando wrote:I know this is not perhaps the best moment to talk of a relatively disappointing performance, specially if there is a commercial aspect to keep in mind, but I would like very much to hear Mark and Harvey to explain us how they see the result, how they are going to fix the issues, etc.
BTW, I still consider as I do since lot of time that the sheer aspect of strength is not anymore so interesting for human players that, in any case, are overpowered by ANYONE of the 50 best engines, commercial or amateur.
At the same time, the field or terrain for a more friendly approach doing of a chess program a real partner, teacher, friend etc is open. Fritz is more or less of that kind, BUT there are still many features that nobody has considered, so engaged has been every programmer in the strength factor alone. Yes, I Know I am and I had been a little bit obsessive with this, but there you have it, again...

Fern, still waiting Godot....
I am not sure there is such a big issue from such a small sample of games. However this was not a good performance from Hiarcs. We did drop half points in games we should have drawn or even won.

The nice thing about a bad result is that it makes us look at all the games and see what went wrong.

Perhaps the best thing about these kind of face to face tournaments is that we actually watch the games and see where things go wrong. One thing all testers are, probably, guilty of is just looking at overall results from 100s of games without studying the games.

you could be studying the games, 100s of them ... automatically.
Identify the positions, where other programs have a better evaluation
Carl Bicknell
Member
Posts: 252
Joined: Mon May 04, 2009 10:06 pm

Post by Carl Bicknell »

Mark Uniacke wrote:It is easy to read too much into just a few games especially with an unfinished experimental engine in the tournament with vastly different hardware, but we did encounter some issues I was not aware of.

Our time usage was not good due to a number of factors and led to an unnecessary shortage of time in many games in the late endgame.

Our king safety algorithm has some issues.

We had a search instability issue which seemed to surface when thinking for long periods especially where positions were transposing alot, and is related to the hash table. It had a knock on effect of causing extra time usage and occasional subpar moves.

I believe these factors alone were responsible for us throwing a number of half points.

Such results while disappointing do act as a strong motivator so the impact on the engine will be positive.
The time bug was quite painful to watch, but overall this looks very promising. You've identified several areas to be improved which will add to the already impressive results (we hope) over version 14.

I have high hopes for HIARCS 15. (and CE2 - how's that going?)
User avatar
vonderlasa
Member
Posts: 71
Joined: Sun Mar 13, 2011 3:14 am

Post by vonderlasa »

Any chance for an iPhone upgrade?
User avatar
Harvey Williamson
Site Admin
Posts: 6079
Joined: Sun Jul 29, 2007 6:57 am
Location: Media City, UK
Contact:

Post by Harvey Williamson »

vonderlasa wrote:Any chance for an iPhone upgrade?
Yes, but not sure when it yet.
gsgs
Member
Posts: 40
Joined: Thu May 26, 2016 1:54 pm
Location: hilter,germany
Contact:

Re: The Future of Hiarcs

Post by gsgs »

gsgs wrote:
Harvey Williamson wrote:
Fernando wrote:I know this is not perhaps the best moment to talk of a relatively disappointing performance, specially if there is a commercial aspect to keep in mind, but I would like very much to hear Mark and Harvey to explain us how they see the result, how they are going to fix the issues, etc.
BTW, I still consider as I do since lot of time that the sheer aspect of strength is not anymore so interesting for human players that, in any case, are overpowered by ANYONE of the 50 best engines, commercial or amateur.
At the same time, the field or terrain for a more friendly approach doing of a chess program a real partner, teacher, friend etc is open. Fritz is more or less of that kind, BUT there are still many features that nobody has considered, so engaged has been every programmer in the strength factor alone. Yes, I Know I am and I had been a little bit obsessive with this, but there you have it, again...

Fern, still waiting Godot....
I am not sure there is such a big issue from such a small sample of games. However this was not a good performance from Hiarcs. We did drop half points in games we should have drawn or even won.

The nice thing about a bad result is that it makes us look at all the games and see what went wrong.

Perhaps the best thing about these kind of face to face tournaments is that we actually watch the games and see where things go wrong. One thing all testers are, probably, guilty of is just looking at overall results from 100s of games without studying the games.

you could be studying the games, 100s of them ... automatically.
Identify the positions, where other programs have a better evaluation

I found this:
http://tcec.chessdom.com/live.php

there you can study computer games with the evaluation charts
over the moves of the 2 involved engines.
It would be interesting to see the evaluation charts of other
(non-playing) engines too, so to compare.

Has it been done ?
Post Reply