One thing more about Don's utility tool,
As far as I know sim tool has been released in 2011, right ?
I mean maybe Don Dailey's view about 60%, was on slower single processors (e.g around 2010 years harwares)
Because my sim test results are done on my latest fast Intel machines
And I noticed some MP engines used maximum cores...
In other words,
The hardware speed can be another factor too about why some clones, derivatives to be less than 60 % similarity
But however,
I am quite happy with current SCCT rule (55% +100 Elo)
-No any double standards ...
-More original works are allowed (no any clones; derivatives)
-No any injustices over the rest original chess engine creators
-Only the talented authors are welcome to participate
etc...
And exception that I recommend the chess engines to be tested as far possible with strong openings
Also I recommend we should test as far possible the engines which are based om more own ideas !!
Otherwise,
-If we will not follow strictly the rues: we will see much more clones / derivatives in the next years !!
And
-In case of using not so strong openings: again we will see strange results as before !!
Hopes this helps too!
Best,
Sedat
Similarity tests
Moderator: Ras
-
Sedat Canbaz
- Posts: 3018
- Joined: Thu Mar 09, 2006 11:58 am
- Location: Antalya/Turkey
-
Frank Quisinsky
- Posts: 7045
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: Similarity tests
Hi Sedat,
all is OK but +100 Elo and a clone / derivate can participant.
Please thinking about it in a quiet minute, perhaps with a good Pizza!
Do you know what programmers do in creating a derivative engine?
Possible that this group of People searching in a second, third ... sources for find out +100 Elo improvement. I like so much the messages from the Critter programmer to such topics with proofs he gave us.
If a person used one work by an other you can be sure that the same person used a lot of works by others too. Perhaps not all ... but I am sure the most.
Have a look in Smarthink.
+150 Elo since Sergei is working on his engine again.
For a good programmer 150 Elo is each time with hard work possible, complete with own ideas or in additional improved good known ideas.
But all in all ... a dervative engine have strengths from the main work and the style is the same.
Furthermore, nothing to read to ... I am using the search from engine x or other things from engine y. Sources are closed an in different cases people try to sell this stuff. Most of them what we have today is Computer chess history, develops by a lot of programmers in the past. To use this work and closed the soruces for a commercial projects isn't right.
Never, never, never we should support it!!!
The result will be a community full of clone and derivative fans and the development of interesting Computer chess things will be stoped for ever.
That is a red Card to each honest programmer, able to develops new and interesting own things in fair competition / contest.
Best
Frank
PS:
To improve IPP sources with 150 Elo isn't a wonder today!
Most are thinking wow ... but again ... isn't a wonder!
all is OK but +100 Elo and a clone / derivate can participant.
Please thinking about it in a quiet minute, perhaps with a good Pizza!
Do you know what programmers do in creating a derivative engine?
Possible that this group of People searching in a second, third ... sources for find out +100 Elo improvement. I like so much the messages from the Critter programmer to such topics with proofs he gave us.
If a person used one work by an other you can be sure that the same person used a lot of works by others too. Perhaps not all ... but I am sure the most.
Have a look in Smarthink.
+150 Elo since Sergei is working on his engine again.
For a good programmer 150 Elo is each time with hard work possible, complete with own ideas or in additional improved good known ideas.
But all in all ... a dervative engine have strengths from the main work and the style is the same.
Furthermore, nothing to read to ... I am using the search from engine x or other things from engine y. Sources are closed an in different cases people try to sell this stuff. Most of them what we have today is Computer chess history, develops by a lot of programmers in the past. To use this work and closed the soruces for a commercial projects isn't right.
Never, never, never we should support it!!!
The result will be a community full of clone and derivative fans and the development of interesting Computer chess things will be stoped for ever.
That is a red Card to each honest programmer, able to develops new and interesting own things in fair competition / contest.
Best
Frank
PS:
To improve IPP sources with 150 Elo isn't a wonder today!
Most are thinking wow ... but again ... isn't a wonder!
-
Sedat Canbaz
- Posts: 3018
- Joined: Thu Mar 09, 2006 11:58 am
- Location: Antalya/Turkey
Re: Similarity tests
Hello Frank
Pizza ?? I will be too fat )) what about sea fish? it's much better ))
Yes...I agree with you we never should support clones, derivatives engines
That's true...it's something like we give RED cart to all honest programmers
And I don't want be look like as amateur tester, e.g if I will allow all 20 strongest engines
Let's say: It does not matter clone, derivative or original...
Then almost all my SCCT top 20 participants will be Rybkanians !!
Sorry...really I am very sorry that I can't test those Rybka engines with different settings...!!!
About my requesting + 100 Elo,
So far, exception Houdini (which was deritive of Rybka) no any other derivative engine is improved to play +100 Elo
But now Houdini is not derivative it in my eyes...
Because any engine which +100 Elo stronger should be counted as original work!
Note also that,
It is not so easy as it looks...to be more than 55% similarity and to play + 100 stronger
If it's so easy + 100 Elo,
Why we can't see any clones, derivatives to be stronger than Komodo or Stockfish or Rybka ???
Btw, for example...if tomorow will appear X stronger engine to be clone of Komodo or Stockfish
Then do you know what will be ? How many people will use it ???
Then forget 100 Elo, even 20 Elo to be stronger, and look how many testers will test it ))
As you know,
If we have difficulties to explain to all testers that they are testing Rybka with different settings
Then I can not imagine...if X clone engine to be stronger than Stockfish or Komodo...
In other words, we can not avoid and stop testing top chess engine, because people love the top things
In shortly what I am trying to explain to all chess friends (maybe my efforts will be useful to avoid...),
If will test IPPO (Rybka)...At least we should test those ones, which are at least +100 Elo over Rybka
Otherwise (in case of 30-50 Elo) its something like we test same engine with different settings
Best,
Sedat
Pizza ?? I will be too fat )) what about sea fish? it's much better ))
Yes...I agree with you we never should support clones, derivatives engines
That's true...it's something like we give RED cart to all honest programmers
And I don't want be look like as amateur tester, e.g if I will allow all 20 strongest engines
Let's say: It does not matter clone, derivative or original...
Then almost all my SCCT top 20 participants will be Rybkanians !!
Sorry...really I am very sorry that I can't test those Rybka engines with different settings...!!!
About my requesting + 100 Elo,
So far, exception Houdini (which was deritive of Rybka) no any other derivative engine is improved to play +100 Elo
But now Houdini is not derivative it in my eyes...
Because any engine which +100 Elo stronger should be counted as original work!
Note also that,
It is not so easy as it looks...to be more than 55% similarity and to play + 100 stronger
If it's so easy + 100 Elo,
Why we can't see any clones, derivatives to be stronger than Komodo or Stockfish or Rybka ???
Btw, for example...if tomorow will appear X stronger engine to be clone of Komodo or Stockfish
Then do you know what will be ? How many people will use it ???
Then forget 100 Elo, even 20 Elo to be stronger, and look how many testers will test it ))
As you know,
If we have difficulties to explain to all testers that they are testing Rybka with different settings
Then I can not imagine...if X clone engine to be stronger than Stockfish or Komodo...
In other words, we can not avoid and stop testing top chess engine, because people love the top things
In shortly what I am trying to explain to all chess friends (maybe my efforts will be useful to avoid...),
If will test IPPO (Rybka)...At least we should test those ones, which are at least +100 Elo over Rybka
Otherwise (in case of 30-50 Elo) its something like we test same engine with different settings
Best,
Sedat
-
Sedat Canbaz
- Posts: 3018
- Joined: Thu Mar 09, 2006 11:58 am
- Location: Antalya/Turkey
Re: Similarity tests
Dear Frank,
Once more I'd like to point out,
Do you remember my racing pigeon story ?
This pigeon (Houdini) came back to home from distance 1000 km
Sorry...I can't kill this white pigeon !!
In other words,
Houdini is another piece of diamond (as Stockfish,Rybka,Komodo)
Why I like a lot the mentioned top engines...because those engine are based mainly on own ideas !
Otherwise Houdini would not be on top for 2 years, btw I have a saying about the current issue:
-No one can be on top by the work of others !
Best,
Sedat
Once more I'd like to point out,
Do you remember my racing pigeon story ?
This pigeon (Houdini) came back to home from distance 1000 km
Sorry...I can't kill this white pigeon !!
In other words,
Houdini is another piece of diamond (as Stockfish,Rybka,Komodo)
Why I like a lot the mentioned top engines...because those engine are based mainly on own ideas !
Otherwise Houdini would not be on top for 2 years, btw I have a saying about the current issue:
-No one can be on top by the work of others !
Best,
Sedat
-
Frank Quisinsky
- Posts: 7045
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: Similarity tests
Hi Sedat,
I know, 10kg lesser and I am OK too.
Reason are Joeys Pizza and my sister (cooks fantastic).
Sea fish is allways good!
Rybkomanians are Ippolotonians ...
In medieval times ... you can be sure ... Rybkomanians / Ippolotonians are fighting hand in hand and are the best friends.
But the good / other site ... the Gladiatdevelopmanians will win each time and put the fear of God into.
Yes, that are settings on an lower basic if we compare it with TheKing! With TheKing settings I had a lot of more fun
55% ...
In much cases the work a programmer do in a dervative engine is very good. If a programmer do such a good work it's more interesting to see a completly own work. Most of such programmers have a lot of good knowledge to do that.
Testers celebrate each 2 Elo more in derivative engines. I don't know why!? I think testers try to find out more participants for Stockfish and Komodo in playing eng-eng games. Can't see an other reason. Same testers celebrate for years engines like Shredder (number 1). Today I read what for a bad engine Shredder is. Typical is at first ... if a Person give own Money for a program ... program is canonise for ever!
We can't make the World better and different people are very defiant.
Grown men
More Information about the Topic clone / derivative and more defiant People wrote that he like that stuff. I am a bull too! That is allways the same with official fora ... much people just search a bit fun. Unfortunately, we lost the honest programmers or other important people. And in 10 years the fun searching People will go to, if no honest Person are available. Fun ended.
I don't like Rybka. Where are Vas during the times of ipp discusses? Where are the Information about Rybka in using material of others? Where are the support in questions of problems by users?
Nothing!
In my opinion, not really an important engine!
Best
Frank
I know, 10kg lesser and I am OK too.
Reason are Joeys Pizza and my sister (cooks fantastic).
Sea fish is allways good!
Rybkomanians are Ippolotonians ...
In medieval times ... you can be sure ... Rybkomanians / Ippolotonians are fighting hand in hand and are the best friends.
But the good / other site ... the Gladiatdevelopmanians will win each time and put the fear of God into.
Yes, that are settings on an lower basic if we compare it with TheKing! With TheKing settings I had a lot of more fun
55% ...
In much cases the work a programmer do in a dervative engine is very good. If a programmer do such a good work it's more interesting to see a completly own work. Most of such programmers have a lot of good knowledge to do that.
Testers celebrate each 2 Elo more in derivative engines. I don't know why!? I think testers try to find out more participants for Stockfish and Komodo in playing eng-eng games. Can't see an other reason. Same testers celebrate for years engines like Shredder (number 1). Today I read what for a bad engine Shredder is. Typical is at first ... if a Person give own Money for a program ... program is canonise for ever!
We can't make the World better and different people are very defiant.
Grown men
More Information about the Topic clone / derivative and more defiant People wrote that he like that stuff. I am a bull too! That is allways the same with official fora ... much people just search a bit fun. Unfortunately, we lost the honest programmers or other important people. And in 10 years the fun searching People will go to, if no honest Person are available. Fun ended.
I don't like Rybka. Where are Vas during the times of ipp discusses? Where are the Information about Rybka in using material of others? Where are the support in questions of problems by users?
Nothing!
In my opinion, not really an important engine!
Best
Frank
-
Frank Quisinsky
- Posts: 7045
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: Similarity tests
Houdini isn't a Diamond!
Under water he can't shine.
So you can see ... it's a clone!
Give me the proof that Houdini has own ideas!
You can give me only the Information ... 150 Elo more as Robbolite.
That isn't a proof!
I am sure that Robert Houdart is very talented but so far I see only ... very talented in copy others because I can't see any proof for own things!
Give it to me!
I will see it!
Best
Frank
The Ruffian programmer sent parts of his code to an other programmer (before Leiden torunament Ruffian won). Vincent had the idea in his brain ... Ruffian is a copy of Shredder. Two programmers checked the sources of Ruffian ... no clone! It's easy today to proof own works, very easy!
Under water he can't shine.
So you can see ... it's a clone!
Give me the proof that Houdini has own ideas!
You can give me only the Information ... 150 Elo more as Robbolite.
That isn't a proof!
I am sure that Robert Houdart is very talented but so far I see only ... very talented in copy others because I can't see any proof for own things!
Give it to me!
I will see it!
Best
Frank
The Ruffian programmer sent parts of his code to an other programmer (before Leiden torunament Ruffian won). Vincent had the idea in his brain ... Ruffian is a copy of Shredder. Two programmers checked the sources of Ruffian ... no clone! It's easy today to proof own works, very easy!
-
Sedat Canbaz
- Posts: 3018
- Joined: Thu Mar 09, 2006 11:58 am
- Location: Antalya/Turkey
Re: Similarity tests
Hello Frank,
What a pity that Vas is already retired...and I miss him a lot !
You know Vas is Czech and I worked with all tourist around the world...
And to be honest,
I found Czech people to be as one of the most polite and warm people
I think this is the main reason about Vas is retired...
I mean he could not to stay any longer ...after those undeserving injustices over him!
Where are the support in questions of problems by users?
You need support ? this is the reason about why you killed Rybka ??
What kind support do you need please tell me ?? ))
Nowadays there is no support,
But at least Rybka gave a lot of fruits, where many many Gladiatorians benefited from him
Note also that Rybka is Shark fish, but female one...))!!
BTW, I noticed one small little thing in your FCT1:
You are testing Fire,Critter, be sure of this:
- it's a Rybka, but with different settings, so you are still testing Rybka !!
In other words,
Rybka is still not dead and it will live in our hearts forever !!
Best,
Sedat
What a pity that Vas is already retired...and I miss him a lot !
You know Vas is Czech and I worked with all tourist around the world...
And to be honest,
I found Czech people to be as one of the most polite and warm people
I think this is the main reason about Vas is retired...
I mean he could not to stay any longer ...after those undeserving injustices over him!
Where are the support in questions of problems by users?
You need support ? this is the reason about why you killed Rybka ??
What kind support do you need please tell me ?? ))
Nowadays there is no support,
But at least Rybka gave a lot of fruits, where many many Gladiatorians benefited from him
Note also that Rybka is Shark fish, but female one...))!!
BTW, I noticed one small little thing in your FCT1:
You are testing Fire,Critter, be sure of this:
- it's a Rybka, but with different settings, so you are still testing Rybka !!
In other words,
Rybka is still not dead and it will live in our hearts forever !!
Best,
Sedat
-
Henk
- Posts: 7251
- Joined: Mon May 27, 2013 10:31 am
Re: Similarity tests
Just ban all clones and derivatives. ELO is not most important. Competition should be fair.
-
Sedat Canbaz
- Posts: 3018
- Joined: Thu Mar 09, 2006 11:58 am
- Location: Antalya/Turkey
Re: Similarity tests
Ok...our rating lists (SCCT+FCT) are the most clean ones !Henk wrote:Just ban all clones and derivatives. ELO is not most important. Competition should be fair.
It's true that I test Houdini, where Frank tests Critter and Fire
But after all,
If you check other ratings and tournaments, off off you will be shocked...
-
Frank Quisinsky
- Posts: 7045
- Joined: Wed Nov 18, 2009 7:16 pm
- Location: Gutweiler, Germany
- Full name: Frank Quisinsky
Re: Similarity tests
Hi Sedat,
hm, but what exactly you missed?
That a person the work from Fabian used without to give any information about it? Honest, I missed other persons, not Vas. Around 100 other persons or more ... more as Vas.
I need support?
No, no ...
Customers of Rybka in the past.
We discuss for a while Critter / Fire added in FCT1.
Critter:
News 72
Fire:
Really a hard work Norman do in the past with the IPP sources. Later others used the work by Norman and closed the sources from Norman. Perhaps you forgot?
One IPP in my list and for sure I am using the work from Norman!
Rybka is Fruit ... Rybka is a setting from Fruit, sorry Sedat ...
Not more no less.
Wishfullness by yourself?
In your heart ... not in my!
I don't like such engines, sorry Sedat!
Rybka is fo me not important!
Not more important as other Fruit clones.
Best
Frank
hm, but what exactly you missed?
That a person the work from Fabian used without to give any information about it? Honest, I missed other persons, not Vas. Around 100 other persons or more ... more as Vas.
I need support?
No, no ...
Customers of Rybka in the past.
We discuss for a while Critter / Fire added in FCT1.
Critter:
News 72
Fire:
Really a hard work Norman do in the past with the IPP sources. Later others used the work by Norman and closed the sources from Norman. Perhaps you forgot?
One IPP in my list and for sure I am using the work from Norman!
Rybka is Fruit ... Rybka is a setting from Fruit, sorry Sedat ...
Not more no less.
Wishfullness by yourself?
In your heart ... not in my!
I don't like such engines, sorry Sedat!
Rybka is fo me not important!
Not more important as other Fruit clones.
Best
Frank