Not logged inRybka Chess Community Forum
Up Topic Rybka Support & Discussion / Rybka Discussion / Christmas gifts ?????????
1 2 3 Previous Next  
- - By Silvian (***) Date 2007-12-11 07:13 Edited 2007-12-11 07:24
Hi Sid S !
Hi Turbo J !
Hi unwashed massess !

I finded-on a very serious site (!)-a lot of new engines .
First of all I thinked Rybka has a lot of "childs".
But they are-all-unbelievable strong !
Test them !
Link:
http://www.exactachess.com/modules.php?name=Downloads&cid=5
Parent - By john conway (*) Date 2007-12-11 08:30
Parent - - By Zruty (*****) Date 2007-12-11 10:07
Don't take this seriously.
Silvian once again tries to confuse the calm pond, hehe, of Rybka forum :)

All these Bugga, Mutha, Clonka and (ultimately) Shitta look too, well, Shitta to me.
Parent - - By InspectorGadget (*****) Date 2007-12-11 10:25
These names sound like they just came out of hell.
Parent - - By turbojuice1122 (Gold) Date 2007-12-11 11:06
I don't know anything about Christopher Conkie, so I don't know whether or not he's trustworthy in saying that these are original engines from their authors.  For the time being, it would be best to play it on the safe side and assume these engines are either truly clones or are filled with viruses.  Naturally, if someone like Tord, Fabian, Stefan, Vas, etc. private messaged me saying that they wish to remain anonymous and that one or more of the engines are their creations, I would openly change my position.  If this is legitimate, then there are only a few programmers who can possibly qualify to be able to create an engine at the level of 2750.
Parent - - By Vempele (Silver) Date 2007-12-11 12:47 Edited 2007-12-11 13:03

> I don't know anything about Christopher Conkie, so I don't know whether or not he's trustworthy in saying that these are original engines from their authors. For the time being, it would be best to play it on the safe side and assume these engines are either truly clones or are filled with viruses.


He hates clones with a passion (as evidenced by his seemingly obsessive trolling in certain Talkchess threads). It's just a bad joke - definitely not viruses, at the very least.
Parent - By turbojuice1122 (Gold) Date 2007-12-11 13:26
Okay, that's good to know :-).
Parent - - By john conway (*) Date 2007-12-11 14:26

> He hates clones with a passion (as evidenced by his seemingly obsessive trolling in certain Talkchess threads). It's just a bad joke - definitely not viruses, at the very least.


Is it clones that he hates, or is it engines with names ending in "a", or both?
Parent - By Vempele (Silver) Date 2007-12-11 14:28

>Is it clones that he hates, or is it engines with names ending in "a", or both?


Clones and the concept of open source (especially when applied to strong engines). Sloppy.
Parent - - By Christopher Conkie (**) Date 2007-12-12 12:04
Just for you, named after you. You are famous......

http://www.exactachess.com/modules.php?name=Downloads&op=getit&lid=98

;-)

Christopher
Parent - - By Christopher Conkie (**) Date 2007-12-12 13:36
Just one more thing.....

Wanka which is 2750 is not yet released.....

It is quite possibly the strongest Wanka ever.

As to not knowing me Mr TJ1122.

I knew about and was here a long time before you were a twinkle in a fish's eye.

http://www.superchessengine.com/vasik_rajlich.htm

You know what they say about ignorance and bliss........

Christopher
Parent - By turbojuice1122 (Gold) Date 2007-12-12 13:42
Hmmm...but a total of eight posts?  I also don't keep much track of the TalkChess stuff, as I generally try to stay away from that forum, so hopefully you'll excuse me for not knowing who you are since you haven't been active on this particular forum.  Obviously my remark above wasn't meant as a stab at you, but simply an agnostic position when not much was known here.
Parent - - By RFK (Gold) Date 2007-12-12 18:19
With all due respect there buddy-I see your name plastered in the upper left of that judicious article, but ah! No questions were hither to for coming from you. I find your, rather explosive emanations on to this page as rather arrogant. I don't really give a crap who you are.
Parent - By turbojuice1122 (Gold) Date 2007-12-12 19:19
I think we've come to a general understanding, and we can all be nice now :-)
Parent - - By turbojuice1122 (Gold) Date 2007-12-12 13:38
Wow, I feel really honored :-)  Definitely something to save for my grandchildren when such time comes.

So with this being the case, doesn't this give me the legal right to know who is the engine's author? :-)
Parent - - By Christopher Conkie (**) Date 2007-12-12 13:40
No...they wish to remain anonymous.

This is how cloners treat programmers. Now the programmers do it back to those thieves.

Its really simple.

Good chess.

Christopher
Parent - - By turbojuice1122 (Gold) Date 2007-12-12 13:43
As I said above, I do hope these are actual programmers and not just more cloners in disguise.  It's an interesting joke.
Parent - - By Christopher Conkie (**) Date 2007-12-12 14:11 Edited 2007-12-12 14:29
They are all real programmers. You might play with their official engines under their well known and original names. The engines are all their own work. It was their idea.

The boot is well and truly on the other foot now. CCRL test Strelka which is a clone.

They won't test what is made by an original author although. Why? It's no different to "Yuri Osipov". Still made up names.....

Only difference is that these ones are really made by the authors of the source and not some bile ridden hack or disassembly.

I can assure you that you have never played with these engines before. Each author changed the way that the engine played before renaming them, some quite radically. They will not tell anyone that they made them. It is their plan and I think it is a wonderful plan.

Ask yourself one simple question.....

If Strelka were 1900 elo, do you think these testers would be so partizan in keeping it in their list?

This says alot about elo hungry testers who don't give a damn about the intellectual property of programmers.

I do not post here often as Vas knows. What I can say is that but for a matter of days, you could well have been writing here in a forum made by ME. I was quite busy at that time however. I still am, with both Arena and Exactachess.

The Team Rasputin project is a pastime......

:-)

Christopher
Parent - - By turbojuice1122 (Gold) Date 2007-12-12 14:24
There are two good explanations as to why CCRL doesn't test these new engines: (a) their names often tend to be either offensive or vulgar, and (b) the authors remain anonymous.  It is said that Yuri Osipov could be found directly if one wanted to do so; I don't know the truth of this, but the same could be said of some legitimate engine authors, too--on the contrary, do you have proof or very substantial evidence that there is no such person named Yuri Osipov who formed the [clone] engine we now call Strelka?  Now looking at their testing of Strelka: I don't think that Strelka should be included on any rating lists, as I've said before that this gives it an endorsement of legitimacy, something that it shouldn't have.  Strelka is in almost the same league as Rocket, and if you use Rocket on Playchess, you get banned--the only difference is that Strelka is a clone of a free engine; similarly, there aren't moral qualms if one downloads and uses Rocket if he has already bought the commercial Rybka from which it's derived.

I think that CCRL can take Strelka off of the rating lists and still take statistics from testing for future purposes, some of which may end up being legal purposes.  If they keep it on the rating list, then they should have no problem testing the Rocket engines.
Parent - - By Christopher Conkie (**) Date 2007-12-12 14:37 Edited 2007-12-12 15:16
BliksKottel.....

http://www.oellermann.com/blikskottel

You have some South Africans in here.

Do you know what BliksKottel means?

BliksKottel = Shithead (or something like that)

OK?

Since when do you get an attack of politeness syndrome?

Being anonymous is irrelevant. Each clone we come across normally has a faceless coward behind it.

Anyway.....what's in a name?

I bet if Tossa is 2900 elo it would make a difference to testers.......and I bet you that it could be entered into the next WCCC.

I have a mum and dad, I dont need Nanny Banks to tell me or respected programmers about his oh so saintly and christian lifestyle. If he were such a do gooder, Strelka would be gone from CCRL pages. The thing is CCRL is run by a Russian. The same russian who is a mod in the forum where there released Strelka. The same Strelka that is hosted on a russian site alongside other luminaries such as Kaissa and Patriot.

Am I getting through?

Soon there will be a......(funnily enough).....

Vulga
Wanka
Tossa
Pisska
Wixa
Dipsticka
Basta
Spanka
Lolka
Osipovka
Ringa
CCRLka
Banka
Speightka
and even a Kaka (wait a minute that's football.....no?)

:-)

See if we careka.....

I bet you know the answer alreadyka.

Chriska
Parent - - By turbojuice1122 (Gold) Date 2007-12-12 15:28
BliksKottel.....
http://www.oellermann.com/blikskottel
You have some South Africans in here.
Do you know what BliksKottel means?
BliksKottel = Shithead (or something like that)
OK?


Okay...what relevancy does this have?

Since when do you get an attack of politeness syndrome?

I am always polite to people who are polite to me, and I am often polite to people who aren't--for example, yourself thus far, in spite of the fact that I have done nothing to you.  However, I can't be pushed forever, as can be seen in replies to a couple of other people on the forum.  The percentage to whom this applies on this forum is in the single digits.

You also seem to be acting hostile, thinking that you are speaking to an opponent.  As you should have seen, I know quite well that Strelka is a clone or the equivalent thereof (I think these discussions about what defines a clone are extremely stupid and equivalent in our case to discussing what the definition of "is" is), as I have been one of the most strongly vocal opponents of it on this forum.  However, I'm also looking at the bigger picture: the issue with Strelka could turn into a legal battle, in which case the CCRL and CEGT data (yes, CEGT data--it is tested by some people, but removed from the rating list) could become quite valuable.

As for the names of these engines, assuming that they are actually from legitimate programmers, then I would support testing them.  However, since these lists are meant for public consumption, it would be stupid to put them on the rating lists under their current names, whether they offend me or not.
Parent - - By Christopher Conkie (**) Date 2007-12-12 15:49 Edited 2007-12-12 15:59
The relevance is that the name is rude. You just dont know it because it is in a different language.

No difference then.....

It's very much a matter of choice. The programmers (all of whom write in the CCC Programmers forum) leave that choice to the testers. What they will miss out on could be quite significant if the testers don't test them. The naming is abhorent deliberatly but they are not clones of any kind.

You can't write an engine completely from scratch and have it called a clone on the basis of it's name.

BTW..what is wrong with the name Massaka or Clonka?

Clones themselves (like Strelka) are abhorent so in turn, nothing should really have changed for those testers.

I am only the programmers messenger......it is their plan.....

I think they have made their point quite well.

If you are on the side of original authors work then you will embrace their efforts to eradicate plagiarism.

You can have fun doing it as well....what more could you ask for?

:-)

Christopher
Parent - - By turbojuice1122 (Gold) Date 2007-12-12 16:22
I think we'll have enough people testing "on the side" such that if someone comes out with something rated at 2900, we'll know relatively quickly.  I won't be doing it in the meantime, though, since I only have one computer and it's going to be tied up with a few other things for the next few months.  (One is testing with CEGT--as I recently started testing Shredder 8 for the list, I would also like to test Fritz 7 and Deep Fritz 10 1 CPU; perhaps after that, I'll see how people react if I start testing one of these...)

I really don't consider Strelka "abhorent" because it sounds like Yuri (assuming that's his name) truly has disassembled Rybka 1.0 beta, something that was believed by most to be virtually impossible; since he has done this with a free engine, I don't see it as morally abhorent; it would be a different story completely if it was done with, say, Rybka 1.2 or Rybka 2.3.2a.  However, I think it's stupid to include it in rating lists, because that gives the implication that it is legitimately a new engine; I also think it's bad for people to give the practice actual legitimacy and pretend that something else has been done.

It would be quite amusing if Sucka 11.22 ended up as 2800 elo, in spite of the fact that it seems to be intended as an insult to me (which seems strange, as I don't see how I've offended any of those programmers...)
Parent - - By Christopher Conkie (**) Date 2007-12-12 16:29
No insult to you. The name was based on a Bulgarian who seems to think that disassembling commercial engines is ok.

It was the version number that was based on you.

;-)

Christopher
Parent - - By turbojuice1122 (Gold) Date 2007-12-12 16:50
Ahh, I see; thanks for the clarification :-)

As for the person in question, I am unsure of what are his true thoughts--for all we know, he may believe Strelka to be a clone, too, but simply enjoys arguing about it.

A possible downside to flooding the community with all of these enginekas is that it might encourage cloners to put their crap out there as well, thinking that it will be easy to mix in with the rest of these.

One thing I don't understand is the reason for the anonymity of the authors--it would seem that they would more effectively put across their points by coming forth more directly, and their "little engines" would also be more likely to be tested, especially if it was someone like Anthony, Fabian or Stefan, who are quite well-known; it seems logical that one of them may have that 2750 engine, as I think they're the only people capable of writing weaker engines than their own who also put out UCI or Winboard engines rated 2750 (of course Vas could also do it, but I imagine he's quite busy now).
Parent - - By revengeska (**) Date 2007-12-14 09:48
Don't worry turbo, there aren't any decent, sane programmers on board with his line of reasoning.  This is a guy that believes the GPL is the root of all evil.  You won't see Vas, Stefan, Anthony, or anyone agree with him.  Tord himself has released his program as open source!  Good programmers know that there is little value in reinventing the wheel and doing something that has been done a million times in the past.  Christopher wants a dream world where writing chess engines is some sort of sporting contest, even at the expense of hindering average engine strength(does anyone really believe Rybka would be as strong as it is without Fruit or Crafty having been open source?).  You don't have to take code from a program directly to benefit from someone else's work, the ideas alone are enough.  Richard Stallman would have a fit.

I also tend to avoid the talkchess forums these days, people blow this cloning "problem" all out of proportion.  The only people that care are amateur programmers and fans who listen to their rants.  I mean, really, when questioned about the idea of Strelka being a clone of Rybka, Vas's response was basically: "Who cares?".  Unless it's dipping into the profits of the programmer directly by cloning the latest version and releasing it, most programmers really don't care if people take older versions and have fun with them.  If it's the latest versions of proprietary programs that are being cloned illegally, I can see a problem.  If it's a weaker version of a program or a program who's author has abandoned it, who cares?  I just don't consider it a big problem.

Then there's the idea of open source programs like Toga being clones(in this case, of Fruit).  Just imagine how the open source community would react if someone came in and preached that Linux distributions infringed on Linus Torvalds' intellectual property and that they were all clones of Linus' original release.  Only in computer chess do you ever hear things so ridiculous.

I've given up on trying to explain this to people on the talkchess forums.  I don't have the time to deal with morons.
Parent - - By turbojuice1122 (Gold) Date 2007-12-14 13:50
The problem is the intellectual property issue, and Vas particularly lately has had far more than just a "who cares" attitude on the issue: there are many original ideas in Rybka that are the reason behind its unparalleled strength, and are the reason why it is a profitable engine.  Thus, with "Yuri Osipov" freely providing Rybka's code in the form of Strelka to any other programmer is quite a problem from the free market standpoint.  It was also a problem at the time that CCRL was supporting this by openly testing it and calling it a different engine.  That has changed: CCRL now calls it a Rybka engine, and it is basically only tested for statistical purposes, just like at CEGT (which doesn't include it on the rating lists at all).

Also, I can guarantee that if one of these engines is released and has an elo of 2750 or anything close, you're dealing with someone like Tord, Naumov, Fabian, Anthony, etc., i.e. someone well-known (or else he is losing a source of income, which doesn't make sense).
Parent - - By revengeska (**) Date 2007-12-14 19:25
Vas has said in a prior post that he doesn't care if someone takes his Rybka 1.0 beta and has fun with it.  I haven't heard of Yuri Osipov freely providing code, but I've been out of the scene for a while.  I tend to go with the "innocent until proven guilty" standpoint; that is, if the author claims it's a unique engine, there's really no harm in testing it until proven otherwise.  A clone is never going to top the rating lists(at least not by more than a couple points), and the more popular it is the quicker it'll be discovered.  Take the engine out of the lists and the testing with all the other engines will still be legit in determining engine strength, the results wouldn't have been skewed any.  I agree that freely providing code(machine code?) to programmers in this case would be a problem that'd need to be dealt with, but most of the fuss came before that time.

That's quite a guarantee, to say that no other programmer on the planet has the ability to write a 2750 elo engine.  I know you're thinking from the experience standpoint, but with code like Toga, Glaurung, Fruit, etc out there for people to look at, it wouldn't take a good programmer too long to get up to speed with the latest techniques and build a 2750 elo engine.  The days of Fritz being "mysteriously" so strong are over.  Depending on which elo group you're comparing to(the ones that humans are compared to or a computer rating list?), there's probably at least a dozen other engines that are stronger, at least half are so far ahead that it's not even worth trying to sell the engine.
Parent - - By BB (****) Date 2007-12-14 20:16
if the author claims it's a unique engine, there's really no harm in testing it until proven otherwise.

The only questions would be: what is "unique" (as opposed to [say] "derived" as with Fruit/Toga/Blueberry, or even just UCI parameter twiddling), and what is "proven"?
Parent - - By revengeska (**) Date 2007-12-15 02:12
I would say a unique engine would be one where either the source code has changed dramatically, or the strength has changed dramatically as a result of changing the source.  UCI parameter twiddling is a feature of engines and is not changing the source at all.  Forks of open source projects like what Toga is to Fruit is unique, there's definately a difference in playing strength and source.
Parent - - By BB (****) Date 2007-12-15 02:51
either the source code has changed dramatically, or the strength has changed dramatically as a result of changing the source... Forks of open source projects like what Toga is to Fruit is unique, there's definately a difference in playing strength and source

I'm not sure I totally agree with this last statement. Move generation is essentially unchanged from Fruit to Toga. Most of the changes are not overly substantial: adding knight outposts, lazy evaluations, fiddling with king safety, additional extended history pruning, some extra endgame knowledge, etc. The rating gain from Fruit 2.1 to Toga II 1.0 was maybe 50 ELO (though both have improved by now). I would not call the change in either the source or the strength "dramatic" - as an analogy, it is about the differences/gains I would expect (say) from FOOBAR version 7 to FOOBAR version 8 if there was no major rewrite by the author (though by Toga II 1.3.4, the difference is more apparent). It just happens that the authors are different in this case. I would certainly think it quite strange if CCRL were to put both Fruit and Toga in the Pure list.
Parent - - By revengeska (**) Date 2007-12-15 19:32
Funny, because what you listed is exactly what I would qualify as dramatic.  You listed a lot of changes there, and 50 elo is a rather big deal.  Now since the newer Fruits have come out, I'd expect the differences in code to be even larger(though some stuff could've been reasonably taken from Toga and put in Fruit.  In the open source world they're called project forks, and in this case you had one person working on the original code(going in one direction), and you had another person working on the original code(going in another direction), and of course they're going to have different conclusions.  Yeah, I'd say there's enough of a difference to treat them as different engines.
Parent - - By BB (****) Date 2007-12-15 20:41
You listed a lot of changes there, and 50 elo is a rather big deal.

These changes only affected one part of the programme. The Fruit "shell" is still there, with slightly different things under the hood. More substantial changes could be, say, changing (parts of) the basic structure to bitboards, or changing the pawnhash to save info about pawn weaknesses (rather than just a score), and then score these weaknesses according to the type of material left (for instance, isolated pawns vis-a-vis major/minor pieces). Running diff on the source of Fruit 2.1 versus Toga 1.3.1 gives about 1100 new lines of code (some of which are merely cosmetic [and some of which are commented out, after being rejected I guess] - also I forgot to add multipv in my previous list of differences, as it doesn't change the strength), compared to about 15000 total lines in Fruit 2.1.

Furthermore, I don't think of 50 ELO as a big deal in the comparison here (though Fruit 2.1 vs Toga 1.3.1 is closer to 100 ELO), as my programme can easily go up or down by this amount by fiddling enough with eval terms, pruning, etc. It could just be a matter of not trying hard enough in the first place to get optimal parameters, and then later [having someone else] put in the elbow grease to get the improvements. I would say in this vein that Fabien likely did not spend much time in optimising the early Fruit versions:

> contrary to popular belief, there really are bugs in Fruit.  Even search bugs.  I just couldn't be bothered with fixing them so far.  Sorry that I can't give you more hints, for now I am using them to find clones effortlessly.


Indeed, Fruit 2.0 was the first to have a "real" evaluation function (he couldn't be bothered before that), and Fruit 2.1 was likely still rather inchoate in this area. It is not surprising that Toga could get a reasonable ELO gain by (say) reconsidering king safety in a superior manner. I don't know how much of a difference lazy evaluation should make (has anyone ever tested this), but the laziness was more Fabien's in this case it seems. Adding extended history pruning when history pruning is already there is not exactly a major change either. The added endgame knowledge included a trading bonus ( commented out) and a re-evaluation of rooks/knights according to Kaufman. [Another eval change was that the tempo counts +20/-20 in the opening and +10/-10 in the endgame in Toga].

Yeah, I'd say there's enough of a difference to treat them as different engines.

Do you consider (say) Fritz 9 and Fritz 10 to be "different engines" or the latter to be the essentially the same engine with (presumably) some improvements?
Parent - - By revengeska (**) Date 2007-12-15 22:32
You have to understand though that Toga is a fork of the project developed by someone else, alongside the development of the normal Fruit.  They take different directions.  Of course the core is going to be the same, but if Fabien admitted that he intentionally left bugs in his program to find clones(and more on this goofy reasoning later), and Toga came in and fixed them, it's all the more proof that the projects had different goals.  The flaw with your Fritz 9/Fritz 10 reasoning is that one(or a few) programmers are working on the same code to produce an upgrade.  In this case, you have two different programmers working on one piece of code separately to achieve different results.  This is the idea of a fork in a project, and inevitably as each is developed further, the code will differ more and more.

I'm assuming you were quoting Fabian there.  What I don't get was: why did he release his code as open source under the GPL if he was worried about people "cloning" his code?  The only reasonable explanation I can come up with for those comments is that he's trying to catch violators of the GPL who are using Fabien's code without re-releasing it to the public.  If this is the case, I completely understand.  Otherwise, it wouldn't make any sense at all.
Parent - By BB (****) Date 2007-12-16 19:20
<regarding GPL and cloning>If this is the case,

I think that is one consideration. Another is if someone entered a Fruit-derivative into a tournament, he and the organisers would like (at least) to know that this is the case.

This is the idea of a fork in a project, and inevitably as each is developed further, the code will differ more and more.

My claim is that Toga has not been developed sufficiently far from the fork-point for it to be considered outside the Fruit family. The codebase of interest (ignoring input/output, utility functions, etc.) is at least 75% identical (probably closer to 90%, though I realise that this statistic need not be of great import). Some people might even insist on calling it Fruit/Toga. Analogously, if (say) Amir and Shay go separate ways, and agree that each can use the current Junior code as a starting point for future endeavours, I would label any resulting programmes from this codebase as being in the Junior family (at least until a major rewrite is done - various people are rumoured to be doing Fruit/Toga rewrites, to include bitboards for 64-bit processors, MP support, etc., and these might be suitably different in the end).

I'm assuming you were quoting Fabian there.

Yes, you can always count on him for some unusual comments. :)

> Although I believe I could keep on increasing strength by adding more and more eval terms, I have little interest in doing so.  I would not learn anything in the process, unless I develop new tuning/testing techniques.  Ideally I would like to spend more time in alternative software, like my own GUI perhaps (specific to engine testing/matches).

Parent - - By turbojuice1122 (Gold) Date 2007-12-14 21:24
if the author claims it's a unique engine, there's really no harm in testing it until proven otherwise.

First, looking at previous posts in this forum, it is clear that "Yuri Osipov" is a pseudonym.  Second, looking at some posts back from July, this person apparently made the claim to a friend of having disassembled Rybka.  Third, it has DEFINITELY been proven that this is not a unique engine--there is no use trying to skirt that.

I agree that freely providing code(machine code?) to programmers in this case would be a problem that'd need to be dealt with, but most of the fuss came before that time.

There have been two really big "fusses" with Strelka.  The first was this past summer when Strelka 1.8 came out.  The second was at the time of the Belka controversy, and that fuss has primarily taken place here.  Both of those have been much larger than any other fuss among the English-language speakers.  They were also both after "Yuri Osipov" released it to programmers.

That's quite a guarantee, to say that no other programmer on the planet has the ability to write a 2750 elo engine.

We are talking about established programmers.  Christopher has made it quite clear that these are established programmers.  Also, you must keep in mind that these would not be the programmers' strongest engines.
Parent - - By revengeska (**) Date 2007-12-15 02:29
First, looking at previous posts in this forum, it is clear that "Yuri Osipov" is a pseudonym.  Second, looking at some posts back from July, this person apparently made the claim to a friend of having disassembled Rybka.  Third, it has DEFINITELY been proven that this is not a unique engine--there is no use trying to skirt that.

I was talking about engines in general, I know it's been proven that Strelka's a clone.  But people are quick to condemn without looking deep enough into the facts.

There have been two really big "fusses" with Strelka.  The first was this past summer when Strelka 1.8 came out.  The second was at the time of the Belka controversy, and that fuss has primarily taken place here.  Both of those have been much larger than any other fuss among the English-language speakers.  They were also both after "Yuri Osipov" released it to programmers.

I haven't really looked at the Belka controversy, I've been pretty much AWOL the last few months.  The Strelka controversy started though when people were noticing similarities between evaluations, and started throwing accusations without further basis.  I'm not one to condemn based solely on evaluations.  It's reason to look into the matter, not enough to convict someone.

We are talking about established programmers.  Christopher has made it quite clear that these are established programmers.  Also, you must keep in mind that these would not be the programmers' strongest engines.

Depends on your definition of an "established programmer".  Personally, I think the "anonymous" part is just a ploy to draw attention to themselves.  Until you know who the programmer is, I can't believe you'd know what their backround was or their intentions are in creating these engines.  For all we know, these could be clones of another engine themselves.
Parent - By turbojuice1122 (Gold) Date 2007-12-15 03:45
I'm not one to condemn based solely on evaluations.  It's reason to look into the matter, not enough to convict someone.

I agree--it actually wasn't until near this past summer that I "switched over" and realized that Strelka was a clone--the previous arguments before then had been based only on evaluations and behavior in illegal positions; what really started to "do it" was the trends in the evaluations instead of the evaluations themselves, especially in a bunch of interesting positions that showed the two engines share the same bugs in addition to everything else.  Anyway, if we were to go just based on evaluations, we'd really have something with Fritz 11, but I don't currently believe that anything suspicious is going on with that.
Parent - - By Christopher Conkie (**) Date 2007-12-14 19:38 Edited 2007-12-14 19:51
Are you readyka for Tossa?

2750 eloka maybe moreka

From the dark sideka...........

Tossa 0.00000001 x32 by Laika Satelliteka

http://www.exactachess.com/modules.php?name=Downloads&op=getit&lid=99

Released under the R.I.P. Licenseka (revengeka is possibleka).



LOLKA

:-)

Chriska
Parent - - By turbojuice1122 (Gold) Date 2007-12-14 21:32
I've tried to download this several times--each time, it says that I typed an invalid pass code, even though I didn't.
Parent - - By Christopher Conkie (**) Date 2007-12-14 21:46
It's case sensitive.

;-)

Christopher
Parent - - By turbojuice1122 (Gold) Date 2007-12-14 23:03
I assumed that--still didn't work
Parent - - By Banned for Life (Gold) Date 2007-12-14 23:08
Tough shitta. :-)

Alan
Parent - - By Christopher Conkie (**) Date 2007-12-14 23:19 Edited 2007-12-14 23:30
Noka...........

http://www.exactachess.com/hmmm/turbospecial/download/Tossa.zip

Don't say I'm not goodka to youka........

Woofka woofka.....

:)

Chriska

(This link will disappearka soon.......3 hourska)

Tell me when you have itka.......
Parent - - By turbojuice1122 (Gold) Date 2007-12-15 00:10
Thankska.  Very interesting choicekas for the directory namekas. :-)
Parent - - By Christopher Conkie (**) Date 2007-12-15 00:32 Edited 2007-12-15 00:44
Here to pleaseka....

The linka is gone nowka.

It's actually more like 2850 btwka.....i wonder how they do itka......

;-)

Thinka about itka......

Place your betska who will winka......

:)

Chriska
Parent - By turbojuice1122 (Gold) Date 2007-12-15 04:35
Ahhh, very cute :-).

Upon testing, I see that technically speaking, the author of Tossa is none other than Vasik Rajlich himself.  Seeing as Tossa is just Strelka, and Strelka's true author is Vas, we have the link naturally.
Parent - - By davidwhite (***) Date 2007-12-12 18:11
Turbo,     it's very possible that I missed something along the way but please tell me what the
reasoning is for you to be spending your time testing Shredder 8, Fritz 7 and Deep Fritz 10.1 on 1 CPU.
              I assume that it has something to do with ascertaining circumstances where the latest
iteration of a program performs at an inferior level to an earlier,and once highly regarded,version.

              It's too often taken as a given that an overall improved version of a program represents
an across-the-board advance in its competence.I'm not certain but I believe that's rarely,if ever,
actually true and that sometimes to achieve an overall increase in strength certain engine functions may need to be weakened.

               Please clarify these matters for me.

Thanks and regards,
David
Parent - - By RFK (Gold) Date 2007-12-12 18:58
Hi David,

Where have you been? I assume you have been reading the forum. One of the major reasons for my lack of participation in the testing of chess engines and analysing of games is my having a P4 3.2 cpu I doubt that counts for anything so I am relegated to the side lines until my boat comes in. The last time I check it was some where out there-actually, I have this painting by Clarance E Braley, pastel that is over a hundred years old, but no one has ever heard of the dude. I am a guy in the wilderness with a painting. In any event-on the brighter side-I can still write (bad-/+good) poetry and just plain madness. Go easy on Turbo-he's just a confused grad student-given to verbosity-but an okay guy.

Yours As Always,

Robert
Up Topic Rybka Support & Discussion / Rybka Discussion / Christmas gifts ?????????
1 2 3 Previous Next  

Powered by mwForum 2.27.4 © 1999-2012 Markus Wichitill