Oh, I almost forgot about this one:
...
First public Critter on CCRL is dated april/2009
First public Doch on CCRL is dated nov/2009
AFAIK, Ippolit was released may/2009, but due to heavy censorship it was not very known until october/2009
> Did you write your main program before or after the release of Robbolito?
...
> Also, the early Doch predated the Ippo release.
First public Critter on CCRL is dated april/2009
First public Doch on CCRL is dated nov/2009
AFAIK, Ippolit was released may/2009, but due to heavy censorship it was not very known until october/2009
> First public Doch on CCRL is dated nov/2009
> AFAIK, Ippolit was released may/2009,
Ummh!
I've always wondered if those who run around pointing figures with a censors eye on others, are just trying to avoid suspicion and stay under the radar themselves.
The original Ippolit was incomprehensible. Only when it was released as Robbolito was it useful to most other programmers, and as I recall it was still written in Albanian/Italian and so we still didn't learn much from it until the English version (I think it was FireBird) came out. I think that was well after both Critter and Doch were out. Anyway, you have convinced me that although you use many Ippo formulas, I was mistaken to imply that you used Ippo as a guideline for Critter in general. My apologies.
Hey Larry,
A suggestion and a request over at talkchess:
http://talkchess.com/forum/viewtopic.php?t=41884
A suggestion and a request over at talkchess:
http://talkchess.com/forum/viewtopic.php?t=41884
> I recall it was still written in Albanian/Italian
It was written in C.
Thank God. I thought I had totally missed out on a new computer language (Albanian/Italian)!
Of course I meant the comments.
> Anyway, you have convinced me that although you use many Ippo formulas, I was mistaken to imply that you used Ippo as a guideline for Critter in general. My apologies.
Apology accepted, I will try to shut my mouth from now on...
Larry - while Vida wrote his own program using Ippo as a "template"
Richard - On the other hand Komodo(Doch) came out as a 3000 engine out of the blue... Now who used something as a "template"?
A hint is given in the Doch 09.980 README file:
Also, much credit goes to the authors of open source chess programs.
Many of the ideas and techniques for doch have been borrowed from
these wonderful works of art.
Richard - On the other hand Komodo(Doch) came out as a 3000 engine out of the blue... Now who used something as a "template"?
A hint is given in the Doch 09.980 README file:
Also, much credit goes to the authors of open source chess programs.
Many of the ideas and techniques for doch have been borrowed from
these wonderful works of art.
![[gb]](/mwf/v12/flags/gb.png "United Kingdom")
Larry,
Can you explain why your early commentary regarding improvements being made to Komodo {started many months ago I think - and every now and then we had a dusting of comments like we are gaining several elo per week!} and the results appearing now in IPON just don't stack up? What do you believe is the reason for this?
Can you explain why your early commentary regarding improvements being made to Komodo {started many months ago I think - and every now and then we had a dusting of comments like we are gaining several elo per week!} and the results appearing now in IPON just don't stack up? What do you believe is the reason for this?
Testing isn't an exact science, and when the dust settles, Larry could even be right, give or take a few Elo. Also remember that his estimates don't always come up short. Rybka 3 beat his initial rating estimates by a good margin...
Early CCRL results were very favorable, as well as some private test results. I still think 20 is about what the average gain on the lists will be.
> Komodo 4 was supposed to be released today (SP version), apparently there was some delay with getting it compiled suitably. Maybe it will be a Christmas surprise for everyone! It's no longer free, but pretty inexpensive ($19.95).
>
> The MP version should follow in January. Price $39.95, but if you buy the SP now, the MP will be half price so the SP now is still effectively free for anyone planning to buy the MP.
I think it is a fine idea to charge for your work. I may even buy the SP version. (probably will)
> I would expect Rybka 5 will surpass Houdini. Perhaps all that's needed to do this is to duplicate the changes Houdart made to create Houdini 1.5 from Ivanhoe, which have already been published by Critter author Richard Vida. If Ivanhoe is really pretty much a "clone" of Rybka, that ought to work. Anyway I'm sure Vas has plenty of his own improvements.
>
> We need one really good idea for Komodo to catch Houdini, I think. If Rybka is first, Komodo second, and Houdini third in a few months, that will be fine with me!
But man you gotta get over this weird Houdini obsession and your rather petty trash talking of other engines. Just be happy and content with the fact that you have been a big part of the development of two of the strongest engines around. That IS something to be proud of.
I don't have any criticism of Houdini or Critter as engines, I just point out that they are both very similar to Ivanhoe, especially Houdini. I have very high regard for Critter author Richard Vida, but not for Houdini's author Robert Houdart because he refused to admit that Houdini 1 was a virtual copy of Ivanhoe.
> <span class="htt">In Response to</span> lkaufman
I don't have any criticism of Houdini or Critter as engines, I just point out that they are both very similar to Ivanhoe, especially Houdini. I have very high regard for Critter author Richard Vida, but not for Houdini's author Robert Houdart because he refused to admit that Houdini 1 was a virtual copy of Ivanhoe.
Your posts are rather veiled in the way they are presented. Take the high road and just boast about Komodo's evaluation, difference in move choice etc. etc.
Make me want your engine because of what it provides, don't try to make me NOT want other engines. (especially the potshots you always seem to take at Critter regardless of the compliments you just gave Richard Vida)
I'm not criticizing Critter or even Houdini. I think it's only fair to warn people though that if they already have the free engines Ivanhoe or even Critter, if they want to buy a new engine that gives completely unrelated output they should choose Komodo or Rybka over Houdini. Houdini is highly correlated to Ivanhoe and somewhat correlated to Critter. Since Critter is free of course I would not discourage anyone from using it; I'm just advising people to choose one of the group (Ivanhoe, Critter, Houdini) and one or more from the others (Rybka 4, Komodo 4, Stockfish 2.1.1) if they use engines for analysis and want uncorrelated engine outputs. I don't mind having Komodo compete with Ippo derivatives, but they should be treated as one group. I agree that Critter is a special case, not one of the Ippo programs but too similar to be considered unrelated.
Basically what I'm saying is that if someone already has Houdini and Critter, he may say "Why do I need a third engine?" My answer is that the two he already has have a lot in common, there is value in having a totally unrelated opinion. Sort of like, if my doctor says I need an operation, and I get a second opinion from another doctor in the same group who agrees with him, I might still want a fully independent opinion.
In my new opening book, I used both Houdini and Critter when I needed live analysis (Komodo MP not yet being ready), reserving Komodo for IDeA analysis where MP is worthless if you have many positions to analyze. But after a while I rarely bothered to use both Houdini and Critter anymore, because they agreed so often. They are both great, but similar.
Basically what I'm saying is that if someone already has Houdini and Critter, he may say "Why do I need a third engine?" My answer is that the two he already has have a lot in common, there is value in having a totally unrelated opinion. Sort of like, if my doctor says I need an operation, and I get a second opinion from another doctor in the same group who agrees with him, I might still want a fully independent opinion.
In my new opening book, I used both Houdini and Critter when I needed live analysis (Komodo MP not yet being ready), reserving Komodo for IDeA analysis where MP is worthless if you have many positions to analyze. But after a while I rarely bothered to use both Houdini and Critter anymore, because they agreed so often. They are both great, but similar.
![[mx]](/mwf/v12/flags/mx.png "Mexico")
> I'm just advising people to choose one of the group (Ivanhoe, Critter, Houdini) and one or more from the others (Rybka 4, Komodo 4, Stockfish 2.1.1)
What matters the most is the moves chosen by the engines. No offense intended, but I find silly to suggest that it's bad for the user to use Stockfish and Komodo together, or Critter and Houdini together, when what is bad is using Houdini and Rybka together as THAT will cause the most move redundancy in analysis. You grouping of engines doesn't seem to have any meaning in practice.
> I find silly to suggest that it's bad for the user to use Stockfish and Komodo together
Okay, I misread, but still, you're suggesting that Houdini and Critter are in the same group, while Rybka and Houdini aren't, when in practice, for having original move choices, it's the other way around (Houdini and Rybka in same group, Critter playing original moves).
I don't have any strong opinion on whether the difference between Rybka 4 and Houdini is greater than the difference between Critter and Houdini. You may be right. I regard those three programs as "cousins". What is clear is that Houdini is much closer to Ivanhoe than any of the above three are to each other, and that all three of the above are much closer to each other than any of them are to other programs like Stockfish, Komodo, Shredder, Fritz, etc.
Differences on the origins of the engines are meaningless, what matters is how they differ in their move choices and evaluations. In this regard Critter is completely original and I don't think it's being treated fairly.
I think that at the same depth (not the same time) Critter and Houdini will choose the same move far more often than say Critter and Stockfish or Critter and Komodo. But this is only my subjective opinion, I haven't done a study to prove so.
I've used these engines in my analysis of correspondence chess, and I've seen a general lack of originality from Houdini but it's not specific with Critter. This only includes positions with several playable moves in where the redundancy of engines goes down, if there's a clear best move and Critter and Houdini agree on it early on, I don't count it as redundancy.
> if they want to buy a new engine that gives completely unrelated output they should choose Komodo or Rybka...
How can you come on here after Don Dailey signed on to two petitions against Vas, and now you are then telling people that they should consider choosing among, what!
>( Rybka 4, Komodo 4, Stockfish 2.1.1)
And then what, when it becomes convenient down the line, Don goes and signs on to a few more petitions to keep Vas from competing?
And your calling into question Critter, Houdini, Ippolit is supposed to make you one of the good guys?
You're just playing both ends against the middle.
The pivotal issue here isn't the dipping into Ippolit ideas- everyone is dipping into that pink elephant's shit. That seems to turn out to be more the rule than the exception.
But personally I think Don needs to relinquish his obsession with the Rybka/Fruit issue. It seem quite silly and dishonest his going along with the banishment of Vas for life, and continuing to head hunt him on other websites that consider inviting Vas to compete.
But now for the sake of making a buck Don now decides NOT to group Vas' chess engine with that of "clones" but with his own and Stockfish?
I sincerely hope that Komodo is all that it can be. But more importantly, maybe it is time to stop playing dirty politics and just let the chess engines speak for themselves by how well they perform.
The Ippolit engines have never been able to get beyond where they are now, and not because they are shunned.
Houdini is where it is in the top spot. Beyond the reach of the Ippolit engines and anything else that is out there, stop playing stupid and blind.
No matter what you think of Robert Houdart - he did what Ippolit was unable to do, what Vas wouldn't do, what Don Dailey has yet to prove he can do and you know the rest.
Don and I don't agree on everything. When it comes to issues like the Fruit/Rybka matter, there is no particular reason we need to agree; we have each given our opinions on the matter independently, and they are different. Don would probably not agree with everything I say here. These are opinions, not matters of fact that we need to agree upon. Our views on religion are also very different, but that does not interfere with our work or our friendship.
As for Houdini, I consider that it IS one of the Ippos, just the best of the lot. I would be willing to bet a huge amount of money that Don and I could take the Ivanhoe code base and modify it with everything we know, and end up with a stronger program than Houdini within a couple months. But he would not enjoy doing this, and it would take an awful lot of money to persuade him to do so, even though there is nothing illegal about it. So we'll continue to improve our own program rather than someone else's.
As for Houdini, I consider that it IS one of the Ippos, just the best of the lot. I would be willing to bet a huge amount of money that Don and I could take the Ivanhoe code base and modify it with everything we know, and end up with a stronger program than Houdini within a couple months. But he would not enjoy doing this, and it would take an awful lot of money to persuade him to do so, even though there is nothing illegal about it. So we'll continue to improve our own program rather than someone else's.
> Don and I don't agree on everything. When it comes to issues like the Fruit/Rybka matter, there is no particular reason we need to agree;
The above may very well be true, and I can appreciate that. Some might even see it as legitimizing Rybka . But the timing is very poor.
In point of fact, it is Don who is new to the market and it is he who is selling his chess engine.
And Rybka still looms large against the horizon.
How would it look if Don Dailey did what you are doing in linking his chess engine, Komodo, to Rybka as a selling point, after all that he has said and done?
It just looks disingenuous when you come and do it for him and he gets away with indirectly sticking to Vas one more time.
Houdini is another issue.
I'm not aware that I was linking Komodo to Rybka in any way. I'm just saying they each provide an independent point of view.
That is just plain semantics. You are representing a programmer who has gone completely out of his way to condemn Vas to oblivion- and here you are giving a sales pitch to buy Komodo, drawing parallels no matter how slight you may perceive those similarity to be in the "Rybka Family", or comparing Komodo to "Rybka, Komodo, and Stockfish". Mean while that excuses Don Dailey to continue his witch hunting Vas down at ever opportunity when ever an organization invites him to compete. There is a name for this, it's called, Good Cop-Bad Cop!
Larry, Komodo might in fact be a great engine! No doubt. And like I stated before, I hope it will be all that it can and more. But the politics behind it sucks.
> But more importantly, maybe it is time to stop playing dirty politics and just let the chess engines speak for themselves by how well they perform.
>
Absolutely. ICGA take note.
![[ch]](/mwf/v12/flags/ch.png "Switzerland")
So that should place it close to Houdini 2 ?
Was the previous evaluation of improvements a bit to optimistic when was said that it would around level with Houdini 2 ?
Anyhow, I am very pleased to see that the MP is getting released very soon and will buy it, SP included since its free !
Thanks and Merry Xmas !
Was the previous evaluation of improvements a bit to optimistic when was said that it would around level with Houdini 2 ?
Anyhow, I am very pleased to see that the MP is getting released very soon and will buy it, SP included since its free !
Thanks and Merry Xmas !
Komodo 4 (now released) will clearly not be at the level of Houdini 1.5 or 2.0 in blitz. Whether it will be close at 40/40' or slower remains to be seen.
We need one really good idea for Komodo to catch Houdini, I think. If Rybka is first, Komodo second, and Houdini third in a few months, that will be fine with me!
Hi Larry,
Your answer seems to imply your main objective is to beat Houdini rather than to be number 1. Do you mean to imply that if Rybka 5 takes the no.1 spot you'll be content to be number 2? Why not strive to be the best? (or did you just mean you don't mind being behind Rybka in the short term?)
Hi Larry,
Your answer seems to imply your main objective is to beat Houdini rather than to be number 1. Do you mean to imply that if Rybka 5 takes the no.1 spot you'll be content to be number 2? Why not strive to be the best? (or did you just mean you don't mind being behind Rybka in the short term?)
We intend to keep on improving Komodo. I have enough respect for Vas to say that if he chooses to go all out to make Rybka as strong as possible, I don't think anyone can catch him. But if he is "holding back" because of the rental idea, then perhaps Komodo (and maybe others) will stay ahead of Rybka.
Hi Werewolf - do you intend to run a test on K4 using your tactical suite?
> I would expect Rybka 5 will surpass Houdini. Perhaps all that's needed to do this is to duplicate the changes Houdart made to create Houdini 1.5 from Ivanhoe, which have already been published by Critter author Richard Vida. If Ivanhoe is really pretty much a "clone" of Rybka, that ought to work. Anyway I'm sure Vas has plenty of his own improvements.
>
>
I'm a bit surprised nobody made any comment about this, not even Richard Vida, so maybe I missed it and all the subsequent uproar about it too in the past year. This seems quite a statement, Richard would have published all the new ideas in a commercial engine, which by their 'new' nature are copyright of Robert Houdart, and which he only could have got by Reverse Engineering.
Let's see some of what Vas himself thought about the subject of reverse engineering being illegal or not, it is a while back but maybe sheds some light on the fact that he also thought at one point Reverse Engineering of Rybka or for that matter other software, was outright illegal:
By Sesse Date 2007-07-14 22:27 This is getting muddy, and I don't think any of us are lawyers, but my understanding of it is that a reimplementation of the type you describe would only be allowed if you did not actually use any of the code in question. There's really only good way of doing that, and that is handing off the documentation you wrote to someone else who then did the reimplementation without looking at the original code at all.
Then again, there is a point here in that reverse-engineering a piece of code (ie. trying to make sense out of a disassembler dump) often is much more work than writing it in the first place, at least if there's enough of it. Depends on a lot of factors, though...
/* Steinar */ By Banned for Life Date 2007-07-14 22:44 Reusing any code that is not clearly in the public domain would certainly be out of bounds. On the other hand, its very likely that Vas has looked at the source code from Crafty, Fruit, Toga, Glaurung etc. without risk of introducing code elements of these programs into Rybka, and since someone reverse engineering Rybka would see only a badly garbled version of the source, I'm not sure there would be too much risk of inadvertently reusing the code (assuming it was rewritten cleanly in some high-level language).
Whether reverse engineering a piece of code makes sense depends on its value. I would certainly agree that anyone smart enough to reverse engineer a chess engine, would be smarter to focus on bigger fish. As Anand would put it, a person cloning a chess engine is stealing from the poor.
Alan Experts agree! Topic Is Rybka a derivative of Fruit? By Vasik Rajlich Date 2007-07-16 10:42
Practically speaking, we can say the following:
1) There will never be patents on chess algorithms.
2) Individuals can (and many will) disassemble Rybka to understand what she is doing.
3) The fact that #2 is illegal is useful - it will prevent a bigger company like Chessbase from assigning this task to one of their employees. In other words, the main engine developer will have to do it himself and lose lots of precious time.
4) The Strelka situation is unique because the author stays behind an anonymous shield and may be willing to share his discoveries.
Vas
By George Tsavdaris Date 2007-07-16 10:53 >3) The fact that #2 is illegal is useful
I'm not sure that it is illegal by itself.
I mean it becomes illegal when you try to apply what you have learnt from the disassembling....
Also I'm not sure even if the second one(apply what you...) is illegal in many countries..... There must be another thread somewhere in the archives, but I could not find it at the moment,where Vas stated that publishing reverse engineered code is certainly illegal. So maybe just publishing the ideas is not illegal but I'm not sure the distinction is always that clear. Is this some form of if Larry did not do the Reverse Engineering himself, he is free to use the gained knowledge? My understanding was that, to my limited knowledge, Richard has only published an open source Pascal version of his own engine Critter, and even Yuri Osipov's reverse engineered code of Houdini, Strelka 5, was then only shared with the author of the changes Robert Houdart. That is as far as I know, from statements made on the computer chess forums. So I still find this quite a statement from Larry!
Merry Christmas,
Eelco
Larry has been making a lot of bull crap and sensationalized statements lately just in the name of garnering sales for Komodo 4 which may or may not have caught Larry's great white whale named Houdini. Larry is slowly turning me off of the entire Komodo enterprise. He keeps taking digs at everybody including Richard Vida.
I may or may not buy Komodo 4. But Larry needs to cool it down!
I may or may not buy Komodo 4. But Larry needs to cool it down!
I don't feel that way myself, it is okay for Larry and Don to be competitive and wanting to achieve the top and I think Larry has always been open about what he is doing, that is not different now from the time he worked for Rybka. He is not going to give much specifics about actual new code they made, but then he has said that would not be his department anyway because Don is the actual programmer of Komodo. I was just surprised that Larry stated Houdini's codechanges to Robbolito had been published as I have not come across anything like that, other than Robert Houdart saying that he thought Richard Vida's version was a better effort than Yuri Osipov's, after he got the code from Strelka 5 for review from Yuri. Nothing about code, or ideas, from Houdini 1.5 being published. It would not be illegal for anyone to use that kind of information, but that is one more reason that making Reverse Engineered code public is probably something that is not legal in most parts of the world, and publishing the ideas behind it a bit on the level of Wikileaks, in my opinion. At least it shows, the level of protection offered by relatively high efforts needed for the deciphering of binaries/executable code is really much less than even some professional programmers were aware of until recently, including Vas.
Regards, Eelco
Regards, Eelco
Richard Vida posted a summary of the Houdini 1.5 new ideas on Talkchess. He explained that he was able to do this in just a week by simply comparing Houdini 1.5 to Ivanhoe (or another Ippo) and looking at the differences. If Houdini was an original work this would not have been possible, but because it was just a modified Ippo the changes were easy for someone with Richard's skills to see. I even sent a copy of the posting to Vas, as I felt that since Houdini was based on a program that used Rybka ideas, it was certainly only fair that Rybka should have the option to use Houdini ideas!
> Richard Vida posted a summary of the Houdini 1.5 new ideas on Talkchess. He explained that he was able to do this in just a week by simply comparing Houdini 1.5 to Ivanhoe (or another Ippo) and looking at the differences. If Houdini was an original work this would not have been possible, but because it was just a modified Ippo the changes were easy for someone with Richard's skills to see. I even sent a copy of the posting to Vas, as I felt that since Houdini was based on a program that used Rybka ideas, it was certainly only fair that Rybka should have the option to use Houdini ideas!
I was referring to your lumping of Critter into the "ippo" family even though it is original code. A very politician tyep of thing to say.
Hi Larry
is expected a version of Komodo 4 (sp or mp) for the iphone/ipad (so iOs operating system?
and if it is expected,when will be released,and what version will be?(sp or mp?)
im really interested,and im sure there is a good market to sell Komodo at iTunes
thanks in advance for your answer
is expected a version of Komodo 4 (sp or mp) for the iphone/ipad (so iOs operating system?
and if it is expected,when will be released,and what version will be?(sp or mp?)
im really interested,and im sure there is a good market to sell Komodo at iTunes
thanks in advance for your answer
That might happen eventually, but we have too many issues related to the pc program to even talk about that now. Perhaps it would be a good thing commercially, and we are not averse to making more money. But neither Don or I is interested in the play of Komodo on substandard hardware, whether that be 32 bit old computers, phones, or anything else. What I'm saying is that our current work on Komodo is a labor of love, which we hope will incidentally produce some income, but porting the program to other platforms would be a purely business decision.
Hi Larry
Thanks for your honest answer,and your approach developing Komodo,not just to make money like other developers
Just another question:what are the improvements made to the Version 4 of Komodo from Komodo 3?
thanks
Thanks for your honest answer,and your approach developing Komodo,not just to make money like other developers
Just another question:what are the improvements made to the Version 4 of Komodo from Komodo 3?
thanks
All the weights in the eval function were re-tuned, that is the biggest difference. There were also some search improvements. Also, the time control play in sudden death was made into an option, and the default was changed rather drastically from Komodo 3. Increment play was also made configurable, but in this case the default was not changed.
thanks.
What is the expected ELO improvement from 3 version to 4 version of Komodo?(i mean SP version)
I took a look at IPON page,but tournament for Komodo 4 isnt finished yet
What is the expected ELO improvement from 3 version to 4 version of Komodo?(i mean SP version)
I took a look at IPON page,but tournament for Komodo 4 isnt finished yet
I wrote here before the release that I expect "in the ballpark of 20" elo. IPON now shows +21 elo. We have one private tester who is so far showing +52 elo, but after just 80 games. So let's say 20-25 looks like a good guess right now.
sorry,i never read that post where you wrote it
well,20 ELO points is a good gained points at that level
regards
Salva
well,20 ELO points is a good gained points at that level
regards
Salva
![[hr]](/mwf/v12/flags/hr.png "Croatia")
Sorry Larry, my bad, no such implication was intented. I did mention also Critter, and "many others", and didn't specify further just because I was too lazy. Sorry once again.
Critter is also in the Ippo family, although the code is original, not "cloned". But almost all of the ideas and formulas of Ivanhoe can be found in Critter, although in some cases they have been improved upon. Basically, all the strong programs are in one of two "families". One family contains Rybka, Ippo, Fire, Ivanhoe, Houdini, and Critter (with Critter being relatively the most original of the lot), while the other includes Stockfish and Komodo. Perhaps in the future we will adopt more Rybka/Ippo ideas into Komodo thus blurring the family distinction, but so far that has largely been a failure, for reasons I have yet to understand.
> (with Critter being relatively the most original of the lot)
And it shows! Considering its move choices and evaluations are very different from anything else.
Other strong engines (for analysis) that provide unique moves or evaluations:
Zappa Mexico II
Naum 4.2
> And it shows! Considering its move choices and evaluations are very different from anything else.
>
> Other strong engines (for analysis) that provide unique moves or evaluations:
>
> Zappa Mexico II
> Naum 4.2
It is interesting JUST how reliable Zappa Mexico II really is, even for such a (relatively) old engine. As I focus more and more on post game analysis of my OTB or real time internet games, I am almost amazed how important Zappa is in that process. (It also seems to be really good at endgames and I only have the 3-4 man tablebases loaded)
In many ways it comes down to trust. There is something I really trust about Zappa.
Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill