Are chessprogrammers jealaous ?

[Don]

I don't think any Chess programmers really care much about this. It is the non-programming, non-testing,non-tournament-organizing basically no-good-for-anything people that are incessantly making trouble about this issue. Come to think of it, they are actually behaving a lot like they are eaten by jealousy. Not sure what they could be jealous about, though. If they were cloners, they might envie the recogntion authors of original engines get for their hard work, and their 'own' programs merely are shrugged off as clones. But they don't manage even that...

What annoys here is your arrogance and pointing finger, but you must not forget that you lulled yourself into a false sense of competence.

HGM is extremely competent. I think you are the one with the inflated sense of self-worth and his status in computer chess outshines yours by far.

[Hood]

I doubt anyone can explain the entire program, from parallel search to endpoint evaluation... and know what is done where... and how... and most importantly, why...

It is nice to know all but... is it necessary ?

The big IT systems consist of many screens, calculations programs, databases files. No one knows all about. Sometimes no one knows how it is working but it is being used and is being useful.
Such a system is developed further, even.

Rgds Hood

If you don't understand the program, exactly how are you going to produce any significant improvements? You might tweak a piece here and there to improve speed. But the occasional major change encompasses the entire code, so that if you have no idea how things work together, it will be impossible to understand the potential interactions. SMP search is a classic example...

We are not talking about "using" a chess program, but are talking about modifying the code to improve the skill level.

Lets return to the cars.
One can improve engine, one can improve aerodynamic etc.

The same in the programming, we are working on programm modules not on a whole programm.
It is possible to improve each module independently and get overall progress. If a proper interface between modules is defined one do not have care about all .

It is of course finding a local maximum in every module and theoreticly the man aware of all interactions can find a global maximum of the function but theoretictly. Practicly the other way is quicker.

Rgds
Hood

[Hood]

yes, but humans do not have to behave with such antypathy

[jefk]

[quote="bob"]

jef >PS how's it in the Stockfish case ? Isnt that one
built one Glaurung or so ?[/quote]

The problem is complex. If Ivanhoe had come directly from Fruit, there would be no problem with Ivanhoe regarding the GPL assuming they release all source changes.
[end quote]

ok thx for the reply; yes i see the problem, and i like to go
beyond arguments about programmer competence, but enter
into software legality, its still about intellectual property rights,
whether its software, or cars. Ok reverse engineering,
like the Chinese make a Rolex, is illegal, especially if
you call it Rolex, ofcourse. But now its seems a hacker
vs hacker case, as these socalled decembrist claim - with
some evidence- they simply made something based on Fruit,
as Rybka is based on Fruit as well (Nb if R1 contains many
Fruit 'modules', its common sense to assume R2.3.4
also contain Fruit stuph isnt it). So ok, if the Ivanhoe do'nt
make Ivanhoe opensource, no big deal, i don't care.

Point is, despite all these comparisons with cars, modules
and whatever, is that there are indications that a commercial
product (Rybka) has used elements of an open source program
(Fruit) without admitting it. Understandable, as re-write
would be lots of work, mostly thinking and then writing,
and otherwise with another opensource release bigbucks
wouldnt be likely (in theory possible, to sell the compiled
version for money, but then you have to make the
compilation really hard i guess, if you release the
sources as well; harder then even uh Crafty i would say

So.. main point is not (only) whether V. the little fish author
is lying or not, and whether programmers are jealous or not,
point is i suspect the Fruit author Fabien is entitled to
part of the Rybka profits, i suspect. Yes ofcourse he then
would have to sue V i suppose.

Example, as some guys seem to be fond of this:
if an airplane is using a Rolls Royce engine, well
then they pay for it, isnt it (and yes, they even keep
calling it a Rolls engine). Now if a Chinese company
would steal the drawings from Rolls Royce and
rebuild the same engine, use it in a supermodern
airplane, as technology has to improve, isnt it, and
then sell it to others as the new Airbus, without paying
royalties to Rolls, what would you think of that ? (well
imho it would be industrial espionage and fraud).

Sure, in this mechanical world there is no 'open source' movement,
but wait a minute, there is science. If you make some drawings in
university for a new airplane, publish it in your Phd paper,
and subsequently on internet, there is no big deal if you don't
sell it as a new airbus and publish your references.

Other example: just like the open source guys simply are
giving away their knowledge apparently for the kick of
it, or recognition, or being proud of improving technology
etc, someone like Bill Gates is giving away billions
to continents like Africa. Now assume a guy there,
lets say the Idi Amin guy, or Mubarak or so grabs a few
millions of the whole, and then claims its his own. hmm,
corrupt you would say ? Well basically its the same.

So the discussion seems to be going in this black&white
direction, you either have legal or an illegal engine.
And the legal engines can be open source (stockfish)
or secret/original (Hiarcs, Shredder, Zappa probably etc)

Most of the 'illegal' engines (ippo, invanhoe, most likely houdini
are *not* commercial, so who cares you would say,
its a hacker world on the internet, it like buying a rolex
in thailand for a few bucks, (almost) nobody cares.
The open source engines usually also are not commercial
(but could be, in compiled way) but thats another point again.

But.. if an 'illegal' engine (and there are indications R is
at least partly illegal) then they certainly shouldnt be
commercial, or at least sold and advertised as original
stuff, ie for big bucks like eg via Chessbase.

Now that's certainly an issue i would say, whether
the Fruit author or not would sue V. And it certainly would
be reason enough for the distributors (Chessbase and
Convekta) to reduce the price considerably i would say;
in the meantime i only can hope for the legal commercial
guys they can surpass Stockfish again, whereby i also
gave a hint to the open source compilers/ distributors
(compile & sell) although they wont care about it i suppose.

jef
PS what i would do if i would be mr V. ?
well i guess i would hire some smart & knowledgeable
programmers, you know the hacker type/Romanian
whatever, and let them work their ass off translating
the Fruit 'modules' into similar, but different stuff..
Question to Bob Hyatt:
dr Bob: do you think this would be possible in C ? my impression of this lousy language is that its possible, apparently there are lots
of different ways to represent certain algorithms, and i don't mean simply changing some parameters no i mean really change the basic coding, but
not the underlying thoughts (you cant patent these like alfa/beta, or more sophisticated Fruit ideas as they became common knowledge, isnt it).
Yes, difficult , but not impossible, in fact a nice challenge as they might try to make it more efficient coding again (then in Fruit.. )

[jefk]

a few additional comments:

[quote="jefk"]
ok thx for the reply; yes i see the problem, and i like to go
beyond arguments about programmer competence, but enter
into software legality, its still about intellectual property rights,
{snip..}
So the discussion seems to be going in this black&white
direction, you either have legal or an illegal engine.
[endquote]

nah instead of the word illegal which i used for engines
like houdini (and R1..) i prefer simply 'not-legal',
for example, i use houdini myself on Icc, at least
it less buggy then Fire, and stronger as well..
yes an engine which is not-legal, but its also free,
and Icc allows me to play with it, so no big deal;
at least Houdart isnt selling it via Chessbase..

[quote
in the meantime i only can hope for the legal commercial
guys they can surpass Stockfish again, whereby i also gave a
hint to the open source compilers/ distributors (compile & sell)
[endquote]

ok, apparenly this is being done already, as the
Bookup guy a certain 'database guy' as he calls himself
apparently sells Stockfish as part of a Masterchess
program or something like that, together with
lots of stuff by himself. Perfectly legal indeed, and
it wouldnt surprise me if he had asked permission
to the Stockfish author, and possibly even promised
him a little commission or percentage maybe..
(guess he learned from the decembrists they can
decompile stuff so in case he would sell Stockfish
as his own he would quickly be exposed.. (?) )

[quote]
etc, someone like Bill Gates is giving away billions
to continents like Africa. Now assume a guy there,
{snip} grabs a few millions of the whole,
[end quote]

not only that, but lets say in defense to some accusations
such a guy would claim it was 'free money' anyway,
so no big deal, he could 'use' it, no problem ..
(not)

ok, in case a guy would set up a foundation, in
cooperation with Gates, claims some costs & expenses,
and then use such millions, no big deal again, legal.

so, the difference is subtle, but its clear.

now in the computer chess world its not about millions,
but still, if its commercial it should be legal; dot.
Or vice versa, if an engine is 'not-legal', then
it should not be commercial; dot.
Or made legal. (dot again)

If i, knowledgeable about grids as i'm member of
the Dutch grid forum, having run the Seti stuff
for a nr of years, and done some hacking on
crafty and secretly made a grid for it on some
hundred thousand of comps, not calling it a
lousy deel-djenghis-cluster, but deep-djengis-grid,
and subsequently rent if for only thousand bucks
per hour (minimum renting time hundred hours), i
guess mr Hyatt would not be pleased.
Unless i would simply admit its based on Crafty,
then it would be an interesting advancement
in computer chess; although i doubt people would pay
so much rent for a game which is a draw anyway,
but thats another subject again..

jef

[bob]

jef >PS how's it in the Stockfish case ? Isnt that one
built one Glaurung or so ?

The problem is complex. If Ivanhoe had come directly from Fruit, there would be no problem with Ivanhoe regarding the GPL assuming they release all source changes.
[end quote]

ok thx for the reply; yes i see the problem, and i like to go
beyond arguments about programmer competence, but enter
into software legality, its still about intellectual property rights,
whether its software, or cars. Ok reverse engineering,
like the Chinese make a Rolex, is illegal, especially if
you call it Rolex, ofcourse. But now its seems a hacker
vs hacker case, as these socalled decembrist claim - with
some evidence- they simply made something based on Fruit,
as Rybka is based on Fruit as well (Nb if R1 contains many
Fruit 'modules', its common sense to assume R2.3.4
also contain Fruit stuph isnt it). So ok, if the Ivanhoe do'nt
make Ivanhoe opensource, no big deal, i don't care.

Point is, despite all these comparisons with cars, modules
and whatever, is that there are indications that a commercial
product (Rybka) has used elements of an open source program
(Fruit) without admitting it. Understandable, as re-write
would be lots of work, mostly thinking and then writing,
and otherwise with another opensource release bigbucks
wouldnt be likely (in theory possible, to sell the compiled
version for money, but then you have to make the
compilation really hard i guess, if you release the
sources as well; harder then even uh Crafty i would say

So.. main point is not (only) whether V. the little fish author
is lying or not, and whether programmers are jealous or not,
point is i suspect the Fruit author Fabien is entitled to
part of the Rybka profits, i suspect. Yes ofcourse he then
would have to sue V i suppose.

Example, as some guys seem to be fond of this:
if an airplane is using a Rolls Royce engine, well
then they pay for it, isnt it (and yes, they even keep
calling it a Rolls engine). Now if a Chinese company
would steal the drawings from Rolls Royce and
rebuild the same engine, use it in a supermodern
airplane, as technology has to improve, isnt it, and
then sell it to others as the new Airbus, without paying
royalties to Rolls, what would you think of that ? (well
imho it would be industrial espionage and fraud).

Sure, in this mechanical world there is no 'open source' movement,
but wait a minute, there is science. If you make some drawings in
university for a new airplane, publish it in your Phd paper,
and subsequently on internet, there is no big deal if you don't
sell it as a new airbus and publish your references.

Other example: just like the open source guys simply are
giving away their knowledge apparently for the kick of
it, or recognition, or being proud of improving technology
etc, someone like Bill Gates is giving away billions
to continents like Africa. Now assume a guy there,
lets say the Idi Amin guy, or Mubarak or so grabs a few
millions of the whole, and then claims its his own. hmm,
corrupt you would say ? Well basically its the same.

So the discussion seems to be going in this black&white
direction, you either have legal or an illegal engine.
And the legal engines can be open source (stockfish)
or secret/original (Hiarcs, Shredder, Zappa probably etc)

Most of the 'illegal' engines (ippo, invanhoe, most likely houdini
are *not* commercial, so who cares you would say,
its a hacker world on the internet, it like buying a rolex
in thailand for a few bucks, (almost) nobody cares.
The open source engines usually also are not commercial
(but could be, in compiled way) but thats another point again.

But.. if an 'illegal' engine (and there are indications R is
at least partly illegal) then they certainly shouldnt be
commercial, or at least sold and advertised as original
stuff, ie for big bucks like eg via Chessbase.

Now that's certainly an issue i would say, whether
the Fruit author or not would sue V. And it certainly would
be reason enough for the distributors (Chessbase and
Convekta) to reduce the price considerably i would say;
in the meantime i only can hope for the legal commercial
guys they can surpass Stockfish again, whereby i also
gave a hint to the open source compilers/ distributors
(compile & sell) although they wont care about it i suppose.

jef
PS what i would do if i would be mr V. ?
well i guess i would hire some smart & knowledgeable
programmers, you know the hacker type/Romanian
whatever, and let them work their ass off translating
the Fruit 'modules' into similar, but different stuff..
Question to Bob Hyatt:
dr Bob: do you think this would be possible in C ? my impression of this lousy language is that its possible, apparently there are lots
of different ways to represent certain algorithms, and i don't mean simply changing some parameters no i mean really change the basic coding, but
not the underlying thoughts (you cant patent these like alfa/beta, or more sophisticated Fruit ideas as they became common knowledge, isnt it).
Yes, difficult , but not impossible, in fact a nice challenge as they might try to make it more efficient coding again (then in Fruit.. )

First, I don't think C is a "lousy language". But the answer to your question is yes. Simple theory would tell you that there are an infinite number of ways to syntactically represent the same semantic behaviour. And even if we quibble about the "infinite" word, you could replace that by "large" and still be standing on pretty solid ground.

And note that even in assembly language, there is an infinite number of ways to syntactically present a program that produces the same output. And not just different variable names and different data structures or overall program structure. You can combine different instructions in different ways, to produce the same final result. So any similarity between two programs, once you get down to source code level, really becomes indicative of a common background connecting the two programs...

[jefk]

[quote="bob"]
.. the answer to your question is yes. Simple theory would tell you that there are an infinite number of ways to syntactically represent the same semantic behaviour. And even if we quibble about the "infinite" word, you could replace that by "large" and still be standing on pretty solid ground.
[/quote]

ok thx again for the reply, it confirmed what i suspected, but..
as most of these 'modules' , variables, declarations, etc are pretty
much interlinked in a chess program (yes ,and certainly in C),
it not only would be lots of work, replacing the Fruit parts
to original parts, it would be a huge amount of work i suspect.
which is the reason i wrote that its 'common sense' to suspect
R2/3/4/ also have parts of Fruit in it. Doesnt mean that R4 is a Fruit
clone, but its probably also not a legal engine anymore, imho,
although ofcourse then there would be more proof needed i guess

jef
PS i wonder in which language this IBM Watson thing was made,
probably some scripting language(s) i suspect..(?)

[De Vos W]

I don't think any Chess programmers really care much about this. It is the non-programming, non-testing,non-tournament-organizing basically no-good-for-anything people that are incessantly making trouble about this issue. Come to think of it, they are actually behaving a lot like they are eaten by jealousy. Not sure what they could be jealous about, though. If they were cloners, they might envie the recogntion authors of original engines get for their hard work, and their 'own' programs merely are shrugged off as clones. But they don't manage even that...

What annoys here is your arrogance and pointing finger, but you must not forget that you lulled yourself into a false sense of competence.

HGM is extremely competent. I think you are the one with the inflated sense of self-worth and his status in computer chess outshines yours by far.

Extremely competent Yes i see...Fairy-Max.... Spartan-Chess..more
dangerous than ever...Oh Well! Cheers

[hgm]

I see that you are still doing your utmost to prove my point that it is mainly the not-good-for-anything members of the computer-Chess community that seem to be jealous.

Of course I can imagine that it hurts when people remark that the status of the author of a humble engine like Fairy-Max outshines yours by far. But you won't have to be concerned that it will inflate my ego. After all, it is hardly an accomplishment to outshine you, and the author of Chad's Chess manages quite well in that respect too.

Yet I am curious: What would you consider to be _your_ greatest contribution to computer Chess? (Other than making nuisance of yourself in forums, by telling others what they should do, and smearing them with libelous insults if they don't...)

[De Vos W]

I see that you are still doing your utmost to prove my point that it is mainly the not-good-for-anything members of the computer-Chess community that seem to be jealous.

Of course I can imagine that it hurts when people remark that the status of the author of a humble engine like Fairy-Max outshines yours by far. But you won't have to be concerned that it will inflate my ego. After all, it is hardly an accomplishment to outshine you, and the author of Chad's Chess manages quite well in that respect too.

Yet I am curious: What would you consider to be _your_ greatest contribution to computer Chess? (Other than making nuisance of yourself in forums, by telling others what they should do, and smearing them with libelous insults if they don't...)

Feed me ... Feed me... Lets throw a little party...

http://www.youtube.com/watch?v=BGRN39oifsE

Next time i try cheese ...