Goodbye Talkchess

[LucenaTheLucid]

Technical expertise is hardly required to recognize that someone has put up a link to dubious, controversial, or unvetted software - whatever you want to call it. A policy of disallowing the tacit approval or promotion of such programs (via download links) until a reasonable determination is made doesn't seem dangerous enough to merit all this hand flapping; it certainly doesn't rise to the level of personal health care and executions.


Pardon my interjection...but some have questioned the Rybka 1 beta program origins...are links to it disallowed? It has been disassembled and compared to the Fruit source code some of which it shares almost exactly...

[mariaclara]

hmmmmm...............

last I heard Talkchess forum is owned by ICD.

Benstoker,

were you kicked ou t by its owner

No. You are quite wrong. The question is not reserved exclusively for the mods to decide. Each and every individual is free to draw that line, as you call it. If one wishes to participate in a forum free of censorship line-drawing debates and secret society sub-forums, he now has an alternative.

Vaya con Dios.


-Sam-

I suppose this means I am about to be censored and banned from this forum. In case this is my last post before execution, live long and prosper.

[Graham Banks]

hmmmmm...............

last I heard Talkchess forum is owned by ICD.

Benstoker,

were you kicked ou t by its owner

He hasn't been banned. He was just being sarcastic.

[Peter Berger]

I don't understand this forum software well . I want to reply in this thread, but not to a special post, so sorry if I did it wrong.

I want to remind you guys of a chess engine that was highly controversial in its days for a little while. Ruffian by Per-Ola Valfridsson.

Has anyone ever *really* talked to this man *at all* ? Dunno - it preceded Fruit but it arrived in a similar way as Fruit did ( but without source) - one day it had just arrived, extremely strong in its days. The author had no interest to participate in *any* kind of public discussions - his contribution was his engine, and that was it.

Bob Hyatt might not want to remember but as his engine was the strongest open source engine these days, he accused it of being a Crafty clone for a while and was very verbal about it. Vincent Diepeveen thought Ruffian must have been a secret sting done by a well-known engine author as no one could have done such a thing - I remember a post where he hinted the most likely suspect was SMK or the author of Fritz .

For all I know it turned out to be original work that went on to be well-respected. The author never changed his policy, did never participate in an ICGA tournament or public discussions, and after a short period of time vanished again, once and for all.

Fabien's entrance to the "scene" was similar, as was Vas's. They arrived and started right at the top.

A guy I know personally and that everyone liked is Dieter Buerssner from Yace fame. His engine wasn't even a chessprogram when he started, just a probably rather sophisticated alpha-beta engine dealing with an uncommon African game. Already his first release of its chess entity was about Crafty strength.

Haven't the first appearances of Shredder and Junior been similar? They arrived , and they were at the top right away.

So what is so different about Ippolit? It arrived with some strange suspicious published decompiled code it seems - and the authors behaved in a very strange way. But that is was so strong means nothing. There is nothing new about new programs coming out of nowhere, attracting suspicion.

The public ( not the programmers who might have a clear and meaningful opinion on it, I have no opion on this myself) have deciced that its strength was the main problem ( as most couldn't make an educated guess about its origins any other way). But this isn't new at all, or is it?

[Uri Blass]

The cases of Ruffian and fruit are clearly different.

Ruffian started as strong free program but not stronger than shredder that was at the top at that time.

Fruit started as fruit1.0 and was relatively strong but not close to be the strongest.

Uri

[Dann Corbit]

I don't understand this forum software well . I want to reply in this thread, but not to a special post, so sorry if I did it wrong.

I want to remind you guys of a chess engine that was highly controversial in its days for a little while. Ruffian by Per-Ola Valfridsson.

Has anyone ever *really* talked to this man *at all* ? Dunno - it preceded Fruit but it arrived in a similar way as Fruit did ( but without source) - one day it had just arrived, extremely strong in its days. The author had no interest to participate in *any* kind of public discussions - his contribution was his engine, and that was it.

Bob Hyatt might not want to remember but as his engine was the strongest open source engine these days, he accused it of being a Crafty clone for a while and was very verbal about it. Vincent Diepeveen thought Ruffian must have been a secret sting done by a well-known engine author as no one could have done such a thing - I remember a post where he hinted the most likely suspect was SMK or the author of Fritz .

For all I know it turned out to be original work that went on to be well-respected. The author never changed his policy, did never participate in an ICGA tournament or public discussions, and after a short period of time vanished again, once and for all.

Fabien's entrance to the "scene" was similar, as was Vas's. They arrived and started right at the top.

Fabian's first iteration of Fruit that I saw was not particularly strong. Elo was a little over 2000. He obviously made rapid progress, though.

A guy I know personally and that everyone liked is Dieter Buerssner from Yace fame. His engine wasn't even a chessprogram when he started, just a probably rather sophisticated alpha-beta engine dealing with an uncommon African game. Already his first release of its chess entity was about Crafty strength.

His eval function used floating point at one time. You won't see many like that.

Haven't the first appearances of Shredder and Junior been similar? They arrived , and they were at the top right away.

So what is so different about Ippolit? It arrived with some strange suspicious published decompiled code it seems - and the authors behaved in a very strange way. But that is was so strong means nothing. There is nothing new about new programs coming out of nowhere, attracting suspicion.

Others have really been caught with their pants down though. Voyager, for instance, was full of Crafty's string table constants.
Bionic Impakt {which I relabeled as 'Belive It Or Not It's Crafty'}, Eugen also spring to mind.

The public ( not the programmers who might have a clear and meaningful opinion on it, I have no opion on this myself) have deciced that its strength was the main problem ( as most couldn't make an educated guess about its origins any other way). But this isn't new at all, or is it?

[Dann Corbit]

The cases of Ruffian and fruit are clearly different.

Ruffian started as strong free program but not stronger than shredder that was at the top at that time.

Fruit started as fruit1.0 and was relatively strong but not close to be the strongest.

Uri

My first version was 0.98a:
protocol.cpp ( 115): send("id name Fruit 0.98a");
I think that version 1.0 was the first version that worked properly as a Winboard engine on Windows.

[Dann Corbit]

The cases of Ruffian and fruit are clearly different.

Ruffian started as strong free program but not stronger than shredder that was at the top at that time.

Fruit started as fruit1.0 and was relatively strong but not close to be the strongest.

Uri

My first version was 0.98a:
protocol.cpp ( 115): send("id name Fruit 0.98a");
I think that version 1.0 was the first version that worked properly as a Winboard engine on Windows.

I was wrong twice. Version 0.98a is a fully functional UCI engine and fruit was never a Winboard engine.

Here is something from down memory lane:

Code: Select all

C:\pgn\winboard-engines\fruit_src>fruit_src
uci
id name Fruit 0.98a
id author Fabien Letouzey
option name Clear Hash type button
option name Hash type spin default 4 min 1 max 1024
option name Log type check default false
option name LogFile type string default fruit.log
option name OwnBook type check default false
option name Ponder type check default false
uciok
ucinewgame
go infinite
info depth 1
info score cp 74 depth 1 nodes 1085 time 0 pv b5c6 b7c6 d4e5 b4c3 b2c3 g4f3 g2f3
info depth 1 nodes 95701 time 46
info depth 2
info score cp 74 depth 2 nodes 97024 time 46 pv b5c6 b7c6 d4e5 b4c3 b2c3 g4f3 g2f3
info depth 2 nodes 97881 time 46
info depth 3
info score cp 74 depth 3 nodes 101168 time 62 pv b5c6 b7c6 d4e5 b4c3 b2c3 g4f3 g2f3
info depth 3 nodes 118034 time 62
info depth 4
info score cp 74 depth 4 nodes 127214 time 78 pv b5c6 b7c6 d4e5 b4c3 b2c3 g4f3 g2f3
info depth 4 nodes 153355 time 78
info depth 5
info score cp 74 depth 5 nodes 170318 time 93 pv b5c6 b7c6 d4e5 b4c3 b2c3 g4f3 g2f3
info score cp 150 depth 5 nodes 197619 time 109 pv d4e5 g4f3 g2f3 d5d4 e5f6 d4c3 b5c6 b7c6
info depth 5 nodes 211500 time 109
info depth 6
info score cp 150 depth 6 nodes 261002 time 140 pv d4e5 g4f3 g2f3 d5d4 e5f6 d4c3 b5c6 b7c6
info depth 6 nodes 283779 time 156
info depth 7
info score cp 150 depth 7 nodes 372031 time 202 pv d4e5 g4f3 g2f3 d5d4 e5f6 d4c3 b5c6 b7c6
info depth 7 nodes 411815 time 234
info depth 8
info score cp 140 depth 8 nodes 563107 time 327 pv d4e5 g4f3 d1f3 d5e4 b5c6 b7c6 f3h3 d8d5 g5f6 g7f6 e5f6
info depth 8 nodes 780127 time 452
info depth 9
info score cp 148 depth 9 nodes 1125438 time 639 pv d4e5 g4f3 g2f3 d5d4 e5f6 g7f6 b5c6 b7c6 a2a3 f6g5 a3b4 d4c3 b2c3 d8d1 a1d1
info depth 9 nodes 1489884 time 858
info depth 10
info score cp 142 depth 10 nodes 2214889 time 1279 pv d4e5 g4f3 g2f3 d5d4 e5f6 g7f6 b5c6 b7c6 g5d2 d4c3 d2c3 b4c3 b2c3
info depth 10 nodes 3220600 time 1840
info depth 11
info score cp 161 depth 11 nodes 5195464 time 2979 pv d4e5 g4f3 e5f6 b4c3 b2c3 f3d1 f6g7 h8g8 g5d8 a8d8 a1d1 d5e4 d1d8 e8d8 b5c6 b7c6
info depth 11 nodes 6837841 time 3915
info depth 12
info score cp 151 depth 12 nodes 10492672 time 6037 pv d4e5 g4f3 e5f6 f3d1 f6g7 h8g8 g5d8 a8d8 a1d1 d5d4 b5c6 b7c6 a2a3 b4c3 b2c3 g8g7 c3d4 g7g2
info depth 12 nodes 15173029 time 8751
info depth 13
info score cp 169 depth 13 nodes 22683258 time 13197 pv d4e5 g4f3 g2f3 d5d4 a2a3 d4c3 d1d8 a8d8 a3b4 c3b2 a1b1 h7h6 g5h4 g7g5 e5f6 g5h4 b5c6 b7c6 b1b2
info depth 13 nodes 30845107 time 17940
info depth 14

[Albert Silver]

Fabian's first iteration of Fruit that I saw was not particularly strong. Elo was a little over 2000. He obviously made rapid progress, though.

His eval function used floating point at one time. You won't see many like that.

Hmmm..... Does that mean it would be better suited to run on a GPU? I have seen many claim that while powerful, GPUs are not well-suited for chess engines, which is why no one really tries to make an uber-engine on it.

[bob]

Fabian's first iteration of Fruit that I saw was not particularly strong. Elo was a little over 2000. He obviously made rapid progress, though.

His eval function used floating point at one time. You won't see many like that.

Hmmm..... Does that mean it would be better suited to run on a GPU? I have seen many claim that while powerful, GPUs are not well-suited for chess engines, which is why no one really tries to make an uber-engine on it.

Floating point or not is not the primary issue. Graphics devices have poor memory latency, they are designed to stream lots of data (high bandwidth) thru, because the primary function is to manipulate pixels as quickly as possible to update a graphical display. Chess engines don't step thru memory sequentially and run into ugly latency issues. Also the concept of a "cache" is foreign.