MatPlus.Net

 Website founded by
Milan Velimirović
in 2006

19:39 UTC
ISC 2021
 
  Forum*
 
 
 

Username:

Password:

Remember me

 
Forgot your
password?
Click here!
SIGN IN
to create your account if you don't already have one.
CHESS
SOLVING

Tournaments
Rating lists
1-Oct-2021

B P C F





 
 
MatPlus.Net Forum General Our incomplete problem chess world
 
You can only view this page!
Page: [Previous] [Next] 1 2 3 4
(41) Posted by Ian Shanahan [Tuesday, Mar 19, 2013 17:41]

@Frank. "Punctiliousness" is better! *grin*
 
 
(Read Only)pid=10003
(42) Posted by Kevin Begley [Wednesday, Mar 20, 2013 02:18]

@Dmitri,

First: you pontificate whether my motives may be based upon a "personal favoritism."
I don't suppose this is intended as a personal slur upon my good character... so, I'll just state that my primary motive has always been to encourage a more complete, formal description and classification for chess problems, based upon definitive distinctions for all fundamental elements, in the context of a CODEX which includes both FAIRY and RETRO considerations.

I have spent some time working on a blueprint for an improved CODEX, intending to provide this to PCCC, for consideration. Though I am a long, long way from completion (completion seems farther away now, than when I started), my intent has been to leave the matter entirely at PCCC's feet, and invite anyone interested to make contributions (or provide feedback). The whole idea was to identify where decisions are necessary, provide all options to the PCCC delegates (asking for a simple vote), and seek a healthy period of discussion (without any further input from me!) to inform their vote.

I would not challenge that my subjective preferences *may* influence some proposals I have outlined -- nobody can be completely impartial.
However, you will find no trace of motivation, whatsoever, to benefit from any of the remedies I intend to suggest (for a wide variety of existing classification failures).
Honestly, should I find the desire to compose further (which has been missing for some time, and deeply exacerbated by events of late), I'm highly likely to seek non-competitive publication.
I question the merit of providing judges the indulgence of free comment space above diagrams (good or bad).

I may be shaped by my own experience, but I foresee NO personal benefit accruing from my honest efforts to provide what I am confident would be a SIGNIFICANTLY better path forward.


That said, let's move on to two of your specific claims...

A) You state that aims/goals may be "absolutely redundant" because fairy elements (like AntiKings) may redefine check...

I do admire your brevity, but I'd much prefer that you make explicit your intended points.
You leave me burdened to infer your meaning, and restate your case for you (for the benefit of others, following the discussion).
Inevitably, this leaves me vulnerable to the charge that I have misrepresented your case -- I would prefer that I were not obliged to make it for you...

Luckily, I think my efforts to divine your intended meaning will prove satisfactory, in this instance.
If you don't agree that my clarification is an accurate representation (in fact, if you don't agree that I have IMPROVED your case), then it is your duty to RESTATE your intended argument, in a follow-up post (which I will be happy to address, promptly).

So, the following represent my best guess as to your position:
1) Some fairy conditions (e.g., AntiKings) may alter the rule of check + (/checkmate #).
2) Any stipulation relying upon these aims (check + / checkmate #) would be altered by the presence of such a fairy condition.

So far, all true -- by default, the aims are malleable to definitions provided in the rules of play (which includes any fairy conditions).

3) A composer may wish to stipulate using an absolute-aim (which is non-malleable by condition).

e.g., using an AntiKings fairy condition, a composer nevertheless desires to stipulate an objective centered upon the aim of an orthodox check (or checkmate).

4) This may require additional aims (e.g., invention of an aim denoting orthodox check: "+ortho+").

All so far stated is certainly true -- presuming that this happens frequently enough to require the sanctioning of such an additional aim.
And, that hypothetical can not be disputed -- because, I have offered clearer examples of the same possibility (years ago, which I will come to soon).

Finally, we approach what may have been Dmitri's case:

5) In the context of "orthodox" chess (rules of movement provided by FIDE, with no fairy conditions) the additional aim (+ortho+) would appear "absolutely redundant" with an existing aim (+).

How am I doing so far, Dmitri -- that is your intended argument, correct?

My ESP must be off, because this is all complete nonsense...
In the context of aims, both would have an "ABSOLUTELY UNIQUE" meaning!!

Years ago, I stated that there should be two separate aims describing capture -- the orthodox capture (x), and an additional capture aim requiring annihilation (useful for circe problems).
Same exact issue, Dmitri (ignoring that my suggestion is more likely to find application).
You are just a few years late in claiming that I haven't taken these steps.

This doesn't even require a fairy condition -- to a royal dummy (+ and #) might seem redundant aims.
There is no validity in taking an "orthodox" or "fairy" perspective in considering the meaning of AIMs.
The meaning of an AIM must be viewed in the context of its definition (which is where redundancy should be eliminated).

Because this refutation of point 5 seems too easy, I must presume that Dmitri's case runs deeper than my inferences have yet appreciated.
NO PROBLEMO -- using a Ouija Board, I have uncovered the possibility of an additional inference (greatly improving his case)...

6) Dmitri might be troubled (quite rightly!) about categorical redundancies, especially as they appear in a database -- which would render anticipation queries more difficult.

If this is indeed your intended point, three things are certain:
1) My clairvoyance remains unchallenged on seven continents, because I alone extracted this argument from your original post,
2) Database redundancy is a legitimate concern (but, it extends far beyond AIMs), and
3) This should come as no surprise, to anyone!

No surprise whatsoever -- the same thing occurs with fairy conditions, which may reduce to redundant rules.

Consider a parry-mover, with a h#n stipulation.
If black checks white on every move in the solution, the condition *might* reduce to that of another idle-mover condition (e.g., "black idles, except to give check"); and, in some instances, it might even reduce to a h#n without any condition!

Consider that r#2 can be reduced to semi-r#2 (though, as is, neither constitute a valid stipulation).

A few years back, this forum had a protracted discussion concerning such categorical redundancies, specifically focusing on how to resolve these issues in databases.
Precedent is one criteria which must be considered, but an intelligent algorithm requires many considerations.

There are methodologies to minimize database redundancy.
I intend to present some ideas in a CODEX proposal (sometime in the future).

The trivial remedy is to classify elements (especially aims and conditions) such that a more complete search is easily obtained (e.g., search for all instances of checking aims -- including: +, +ortho+, etc).

Another solution would be to require PCCC to sanction a hierarchical remedy (e.g., to intelligently rank elements according to an intelligent algorithm), and expect compositions to utilize the highest "correct" ranking element possible (e.g., use + because it ranks higher than +ortho+).
Software tools could provide assistance here...

More sophisticated remedies require a fundamental breakdown in the classification of elements (e.g., a hierarchical family structure for fairy conditions, fairy units, aims, etc).

B) You claim that a helpmate can be a fairy condition.

This is obviously false -- h#n is a stipulation, which can not be considered a "fairy condition."
I provided a clear definition for what a fairy condition is: anything which ALTERS (or constrains) THE RULES OF MOVEMENT (the physics of the chessmen -- which is distinctly separate from an OBJECTIVE), as provided by the FIDE RULE BOOK.

Either you are still desperately searching for a redundancy flaw, or my powers of perception are failing.
You will just have to make this case explicit yourself -- very likely, I have already covered it, above.
There is no way to misconstrue the meaning of a helpmate, as a fairy condition.

My best guess: you are trying to say that a fairy condition might provide the white player an ability to move black units; thereby reducing a directmate into a helpmate.

Is that substantially the gist of what you are trying to say?

There are methods to reduce (and sometimes eliminate) redundancy:
First, PCCC intelligently eliminates any redundancy, in its proper context, via sanctioning authority.

If one element (condition, unit, aim, etc) is EXACTLY defined as redundant (in all cases) with another, of the same type (respectively: condition, unit, aim, etc), then PCCC only sanctions ONE such element.
The result: no redundancy within context of each element classification.

Second, they can intelligently eliminate any redundancy, in a hierarchical manner.

Starting from the lowest fundamental element -- the valid AIM (the ultimate objective of a formal stipulation: #, +, =, etc).
PCCC will not sanction any higher order element which can be defined as a valid aim.
There are sub-categories of AIMs, and simple rules to completely determine any valid aim.

I have covered this in previous discussions -- that is why I am struggling to put this all together into one package, which will provide elementary explanations that any beginner can follow.

From aims, you can build valid formal stipulations of the 1st order.
Only from 1st order stipulation, can you build 2nd order stipulations (all based upon iteration).
All the way up to Nth order -- a complete description of all formal stipulation, capable of describing problems FAR beyond what the present formal system allows...
And, also capable of providing the composer an ability to define the conditions for soundness (by eliminating duals from the solution).

In other words, you may stipulate a "s#2" problem, which requires no dual (for soundness), all the way through the solution, iterating unto checkmate.

Alternatively, you may stipulate a "s#2" problem, which ends just prior to the final mating move.
In other words, you stipulate to fulfill the goal (realize a position with a compelled-mate), but stop prior to the stated aim (mate of wK).

Once formal stipulations are defined, this all fits into a hierarchy.
You do not sanction a fairy element which can be exactly described by a a formal stipulation.
This reduces redundancy across all elements.

So, when PCCC encounters a condition which attempt to allow one player to move the units of another player, for the duration of the game, they reject it -- if this can already be exactly treated as an aim, or a formal stipulation of the 1st order, or 2nd order, all the way up...

But, remember: there are RULES to test the validity of AIMs, and formal stipulations (of all orders)...

One such rule: no formal stipulation may alter (or constrain) the rules of play.
In other words, stipulations can not force a deviation from a legal game.

Helpmates are an objective, which may be formally stipulated, because they do not alter the rules of play in the game of chess.
If you think otherwise, you would be required to provide ONE example where the helpmate stipulation causes a deviation from a LEGAL GAME.
Try, and you will fail.

Series-movers do deviate from a legal game (move alternation is altered).

Reflexmates also deviate from a legal game -- by constraint (much like the maximummer constraint)!
In a r#n, the movement option (as a law of physics governing the chessmen) PREVENTS certain moves (when a player has a #1 available).

That is why you first lock in precise definitions for the most fundamental elements.
AIMs, first, then formal stipulations.
You can not build a 1st Order Formal Stipulation (OFS), without an AIM.
You can not build a 2nd OFS, with a 1st OFS.
...
You can not build a Nth (OFS), with a (n-1)th OFS.
Now, you've covered the full range of formal stipulation.

All else must be a fairy element.
If it can apply to ALL units, it is a fairy condition.
Otherwise, a fairy unit.
 
   
(Read Only)pid=10004
(43) Posted by Kevin Begley [Wednesday, Mar 20, 2013 03:26]; edited by Kevin Begley [13-03-20]

@Ian,

> In a nutshell, you're saying that "Series.[whatever]" is the fairy condition, and "whatever" is the aim.
>I can agree to that. So an elided *combination* like Ser.H=17 is indeed both a fairy-condition (Ser.) and an aim [H]=17.

Close, but not exactly right -- replace "stipulation" wherever you said "aim."

In "Series.[whatever]", "Series" must translate to a fairy condition, yes (a rule, which alters move alternation).
"Whatever" is *probably* the stipulation, presuming it is a valid formal stipulation (e.g., r#n = s#n stipulation plus an additional fairy condition).

An AIM is the most fundamental building block of simple (1st Order) formal stipulation.
It is the ultimate objective: #, =, +, x, -ep-, -00-, etc...

If the stipulation is h=17, the PLAY is "h" (help achieve the aim!), the AIM is "=" (stalemate bK), and the DEADLINE is "17" (moves).

But, just as any valid formal stipulation must obey rules (no alteration of the chessmen's physics), a valid aim must also obey rules.

There are two types of AIMs:
a) Move Attributes (x, +, -ep-, -00- etc), and
b) States (#, =, etc).

Move attributes are easy to define (a move which accomplishes something).
States (#/=) are defined by an attribute of the prior move (check/not-check), plus the absence of legal moves.
States require careful rules, so as not to disturb the entire classification system!

To understand GOALs, consider:

1st order formal stipulation = Play-[Aim]-Deadline.

h=17 is a 1st Order Formal Stipulation, which I would rewrite as: "h-[=]-17."

To build higher order stipulations, you begin with some 1st order stipulation (say h=17), and make this your GOAL.
The goal then functions exactly like an aim, but is actually a sub-stipulation.
Goals are like state-AIMs -- they represent the realization of a position with some state.

In the 1st order stipulation, state aims are the realization of a position with a game state (#, =, etc).
In the 2nd order stipulation, state aims are the realization of a position in which the goal (sub-stipulation) can be met by (or before) the deadline.

2nd order stipulation = two types:

a) play-[goal]-deadline, and
Where the goal = some 1st order stipulation, which must be solved by iteration, down to its ultimate aim (#, =, +, x, etc).

b) play-(goal)-deadline
Where the goal is the same, but realization of the goal terminates the problem, you do not solve the goal by iteration, and never reach the final aim.

Easy, right?
But, you're probably thinking, "who would ever require a higher order stipulation?"

You have probably already used it, many times, without ever realizing it.
Almost everybody requires this.
And with a formal stipulation which describes this option, you'll have an fundamentally sound infrastructure, which facilitates progress in stipulation exploration, far into the future...

The key idea is not only to stipulate the weird "h-[#2]-3" problems -- though I was inspired to provide for this, by a composition by a favorite composer.

The big idea is to provide a mechanism which can terminate prior to the realization of the final aim.
You can now redefine a s#2, such that you stop just short of the final coup (thereby eliminating any dual).

Cap-Zug -- which is essentially a "sxn" stipulation, plus some fairy condition(s) -- requires (at least) a 2nd order stipulation, to prevent any dual from occurring in the solution (e.g., multiple capturing moves may be possible, but are ignored by explicit stipulation).
If no dual appears in the solution, there is no dual.

This enables you to stipulate a whole class of "s-(goal)-n" problems, in which YOU may stipulate the conditions of soundness (rather than having to rely upon any dogmatic convention, which may change over time).
And, it enables you to preserve many problems of the past, which was composed under an alternate convention.

Nth order stipulation = play-[goal]-deadline -or- play-(goal)-deadline
Where goal = some (n-1)Order stipulation.

That defines all formal stipulation, using only a few fundamental elements (Play, Aim, Goal, Deadline, Goal, and termination), which are based upon only a few distinct rules.

It is important to understand that the PLAY is always dictated in relation to the present goal.
In "h-[#2]-3", black helps for 3 moves (to realize a position allowing a directmate within the deadline).
After reaching the goal, black's play changes, to oppose the aim (mate) within the deadline.

The same character may help you to the store, then fight you for the last pint in the cooler, then help you get home.
That's the nature of problems!

This forms the backbone of a categorization system I would suggest.
Except, there are some rules to consider, which I have neglected here, with respect to aims (especially of the STATE type).

This should enable any beginner to roll-up the most complex stated problem into a formal stipulation; and conversely, they to unravel the highest order formal stipulations into simple wording.

They will not lack a fundamental grasp of what separates a stipulation from a fairy condition; nor will they be left wondering about categorization.

The trouble is, WFCC/PCCC have created album divisions for which there are no fundamental definitions.
And, they've handed out FIDE titles, for these undefined divisions (thereby creating multiple International Judge Titles, without providing any jurisdiction).

Worse, they will almost certainly prefer the present state of incompleteness and unawareness, because the political fight to maintain favored divisions is likely to be fierce.
People want their titles -- even if based upon flawed and biased divisions, which segregate problems into a number of autonomous albums.

That is why our chess problem world remains incomplete.
That is why nobody cares to provide a logical path for improvement.
Ask the "masters" fundamental questions -- for which any beginner should have an answer! -- and they will only make you feel unwelcome.

The divisions are an earned fraud.
Call it what it is: a pyramid cheating scheme.
I refuse to be party to my own theft.
I urge others to insist on a return to honest fundamentals.
Doesn't have to be my suggestions -- just honest.

The title is just a goal.
Follow through to achieve the ultimate aim: preserve and improve the integrity of all chess problems.
 
   
(Read Only)pid=10005
(44) Posted by Dupont Nicolas [Wednesday, Mar 20, 2013 05:00]

@ Kevin,

The point where we disagree is that you consider orthodox chess rules (i.e. CODEX) as a paradigm, and you try to classify the various concepts through it: roughly speaking, the helped genre is orthodox because close to chess game rules, while the series genre is fairy because far from those rules.

I can’t follow you on that way because, more or less, this is just like assuming that the earth is the center of the universe, the Sun an orthodox star while Antares is fairy…

Mathematicians and Physicians are making enormous efforts to avoid singularities. A singularity is a nightmare for a scientist because general theories are scratching on those points. Orthodox chess game is full of such singularities e.g. castling or en-passant captures, so CODEX can’t be the right candidate to be the starting point of a well-made chess problem theory.

My point of view is that “rockagogo” is much more “orthodox” than the classic castling, just because it is much more general. We need to start with “universal” definitions, to end with particular cases, among them orthodox chess game. Your attempt is just the opposite, trying to find universal definitions (what is a stipulation, a condition, etc.) from a particular case.

You recognized that the standard capture concept is not enough general, that a division has to be operated between annihilation and possible rebirthing. Why don’t you follow this correct way until the end? Trying to setup the general concept of “capture” so that both the orthodox captures and the various fairy captures appear as special cases?
 
   
(Read Only)pid=10006
(45) Posted by Dmitri Turevski [Wednesday, Mar 20, 2013 07:06]

@Kevin

The reference to personal favoritism was intended as a friendly pun, i never even thought of questioning the nobility of your motives.

 QUOTE 
I do admire your brevity, but I'd much prefer that you make explicit your intended points.

I'll try:
If fairy elements are allowed to redefine what check/checkmate/stalemate is then any problem that uses aim different from checkmate (eg +, =, ep, 00 ...) can be reformulated as a checkmate problem (with additional fairy condition(s) that account for the new definition of checkmate). If all problems are reformulated as checkmate problems then indication of the goal/aim is redundant.
The above is also true for a problem that wants solver to deliver orthodox check under AntiKings condition.

 QUOTE 
4) This may require additional aims

This is sort of a logical mistake. This may, but also this may not. Additional fairy conditions (that deal with defining orthodox check as a new checkmate in AntiKings) may serve as well.

 QUOTE 
you are trying to say that a fairy condition might provide the white player an ability to move black units; thereby reducing a directmate into a helpmate. Is that substantially the gist of what you are trying to say?


Yes. This is exactly how you have replaced "Series-" with a fairy condition. And this is where i lost you. Instead of addressing the issue "it may be done" you have switched to "should it be done?" and "PCCC". Perhaps it shouldn't, the point is that it may be done, which proves (well, at least to me) that the system you propose is just a different kind of syntactic sugar. I see no point in discussing preferences for certain kinds of syntactic sugar.
 
   
(Read Only)pid=10007
(46) Posted by Kevin Begley [Wednesday, Mar 20, 2013 10:31]

@Nicolas,

>Orthodox chess ... is full of ... singularities e.g. castling or en-passant captures, so CODEX can’t be the right candidate to be the starting point of a well-made chess problem theory.

I had this same idea, years ago, Nicolas.
All things universal...

Nothing based upon an arbitrary starting position (no more Circe Modalities)...
Later, I removed all conditions ... just Leapers, Riders, Hoppers...
Not even pawns (which require a board orientation)!
That gets rid of promotions, double-steps and en passant.
No more castling...

A universal minimalist form, which nobody plays, and almost nobody composes for -- but so what, right?
It provides a minimal basis, for which all other rules should refer.
This, I thought, would serve as the optimal form.

I even carved out a section of my database, which had only such problems in it.
You wouldn't believe how small such a database can be; or, how boring -- but so what, right?
I actually did consider composing entirely within such a framework (no joke!).

Then, I started hacking into the board size -- 7x7, 6x6, 5x5... and on down I went.
For a brief moment, I considered holes, but that was wrong -- information theory abhors their entropy.
I began studying 2D holograms projected onto the event horizon of micrscopic black holes -- here was a chess board, satisfying the most fundamental law in all of physics: the conservation of information.
The simplest possible rules for Chess that you could possibly describe... The Caissa Particle was in my grasp!

I crunched rules of movement into genetic bit patterns, and cinched them with iterative consensus.
Minimalism took me, and I began dreaming of Royal-Vizirs posted on the moebius-edge of dimensionality itself.
Compressed hoppers wheeled overhead, completely defined by error correction itself.
But, it was not the end -- I felt complexity in me.
Until at last, I threw down minimummers, and smote their ruin upon the plank square!

I've been sent back to elaborate the problem with your minimalist approach.

Your basis is a waste of space.

Think of it this way: In the English language, "e" is the most commonly used letter.
Knowing this, have a careful look at a Morse Code chart.

e = •
q = ––•–

What is the most commonly used basis for chess problems?
Here is your answer: FIDE 1997 = •

Minimalism, for print journals, leaves room only to reference the dot, and a few alterations.
And, online, all rules should be completely self contained.



>My point of view is that “rockagogo” is much more “orthodox” than the classic castling, just because it is much more general. We need to start with “universal” definitions, to end with particular cases, among them orthodox chess game.

I like Rockagogo. But, it doesn't describe all castling options in Chess960.
A minimalist description of movement patterns is not economized on a macro scale (of castling).


>Your attempt is just the opposite, trying to find universal definitions (what is a stipulation, a condition, etc.) from a particular case.

These are not opposite scales of a spectrum -- the real trouble is, you're on the wrong scale.
If you want to categorize problems, focus on the problem scale!

What is a chess problem?
What is a problem?
What are the elements of a problem?
What are the elements of formal stipulation?
What are types of aims are there?

This is how you find the atom of a problem.
You'll never find a Caissa Particle hidden in the rules of an evolved game, like chess.
Just try to define a chess problem such that it logically excludes checkers!
You can not do it.

Rockagogo? No, Tic-Tac-Togo!

Game rules can only define the physics of the chessmen.
The focus of a all problems is the aim.

>You recognized that the standard capture concept is not enough ... that a division has to be operated between annihilation and possible rebirthing.

No, I recognize that two unique aims are required.

I have a circe problem, with a twin:
a) help-capture-2
b) help-annihilate-3

These are not rules of the chess game -- they only define the ultimate objective of a PROBLEM.
They are two unique AIMs, which form a STIPULATION.

Just think of this as INFORMATION, which must be conveyed to the solver.
To understand ANY problem providing "complete information" (see game theory), you need to know what?
1) The starting situation (a diagram),
2) All available options (rules of movement described by a book and/or fairy conditions),
3) The deadline (# of moves beyond which all success has expired),
4) The motivations of all characters involved (relative to your success, they either help or oppose),
5) The ultimate objective (the AIM -- the direction for the solution), and
6) The definition of success (either fulfill the aim completely, or achieve some goal along the way).

If the AIM is a vital component of the information which must be provided to the solver, then we certainly must provide a formal differentiation for the two AIMs seen above.

They can not be the same, Nicolas.
From Dmitri's improper perspective, they appear redundant -- but, this can not be true.
We can not eliminate an aim which is uniquely capable of conveying the necessary information.

What do you want to do -- define the relationship of the two aims?
That's not necessarily a bad idea.
As I indicated to Dmitri, a hierarchical relationship may provide important information, in helping target queries which might otherwise require searching for all aims which might be redundant.

What would you do with their relationship?
You want to make "capture" and "annihlation" two children who share a parent?
OK -- "take" begets both "capture" and "annihilation."
Now what?

They are nothing more than unique pieces of information (dots and dashes).
You would prefer that related aims all rhyme with one another?
You want to set the two apart as some function one dot/dash?

The only thing important here is, you need some dot or dash to distinguish the two pieces of information.
The two unique aims.

You can not reduce them to a single symbol, and still convey the same information.

>Why don’t you follow this correct way until the end?
Trying to setup the general concept of “capture” so that both the orthodox captures and the various fairy captures appear as special cases?

I am not a chess piece -- those physics only apply to chessmen.
A problemist interacts with the atoms of the problem.

But, indulge me ... what is your end game here?
 
 
(Read Only)pid=10008
(47) Posted by Kevin Begley [Wednesday, Mar 20, 2013 12:47]; edited by Kevin Begley [13-03-20]

@Dmitri,

>If fairy elements are allowed to redefine what check/checkmate/stalemate is then any problem that uses aim different from checkmate (eg +, =, ep, 00 ...) can be reformulated as a checkmate problem (with additional fairy condition(s) that account for the new definition of checkmate).

There are steps which can be taken to minimize redundancy.
The first step I offered was to prevent redundant definition of any element (including aim) which exactly matches an existing aim.
This must apply to redefinition of aims, as well.
A fairy condition which attempts to redefine "X" as stalemate can not be sanctioned.

>If all problems are reformulated as checkmate problems then indication of the goal/aim is redundant.

Can't happen -- see above.
Intelligent methods of sanctioning problem elements (aim, play, fairy unit, fairy condition, etc) would prevent the reformulation of one aim to exactly match an aim already in the quiver.
So far, even without any sanctioning authority, I am only aware of two aims which have ever been redundantly defined (to exactly match some existing aim).
And, in fact, they are not identical (because in both cases, they conceal different fairy conditions).

That's the big problem -- not redundancy, but concealment of fairy conditions.
Invalid aims and invalid stipulations consume vast amounts of precious information.
And, if they are popular in certain circles, they multiply very quickly.

Look at how many unique parry-mover stipulations are required today (and still, they can't fully cover it)!
Now, think about this: 1 pair of fairy conditions would provide 100% complete coverage.

>The above is also true for a problem that wants solver to deliver orthodox check under AntiKings condition.

Not if the "+" aim is malleable to the condition.
In that case, you require an absolute aim (which is not malleable to the condition).
Off the top of my head, I suggested "+ortho+" -- a poor name, but it makes the point.
These two aims would have an absolutely unique meaning.


>[Requiring additional aims] is sort of a logical mistake.

In actuality, I said the very same thing, many many years ago...
I thought an independent formal AIM was necessary for AntiKings.
If memory serves, I wanted to call it "!+" (not check) -- and I insisted that it was a "fairy aim."
Back then, somebody had to tell me that my symbology ("!+") was already taken.

Without an sanctioning authority, the obvious question was: "taken by whom?"
The software owns the symbols, and the more tools available, the faster they can gobble up symbols -- especially by lack of coordination.
The good news is, a better tool can always come along, and complete redefine EVERYTHING!
That's the real ticket to improvement.

And, if you want the truth, AntiKings are not even a fairy condition -- they are a fairy unit.
A "fairy condition" must apply to all units.
We've been improperly calling this a condition, because we've been conditioned to improper conditions.
So, in fact, a fairy unit alters the aim.

Second, there's a really really bad flaw in your analysis...
So, you are marching along with two AntiKings on board, right?
Doing your thing (either by force, or with some help from the opponent)...
Then, suddenly, you say you want to check/checkmate the opponent's AntiKing in an orthodox manner -- correct?

Think about what you're saying, Dmitri...
It is your move, prior to playing your orthodox version of check or checkmate.
That means what?

You guessed it: the black King must ALREADY be in a state of check!!
You can not check a King (a move attribute) if it is already in check (a state), in any "orthodox" fashion.

If you want to checkmate the black King, instead, but in an orthodox manner, you still can not!
There is NO orthodox checkmate where the mated player was in check prior to the mating move.
So, this whole AntiKings chapter has been a huge mistake.

I provided you some better examples (e.g., capture/annihilation in circe).
I suggest you use those.
But, there is nothing redundant there.
It is a far stretch to complain about redundancy, when you have some imaginary character going out of their way, to create a new fairy condition, which redefines a redundant aim.
C'mon -- who are we kidding? -- you actually insist on protections against this?
May as well declaw the Easter Bunny, while we're at it.

What you have to appreciate is: there is no good reason to redefine en passant.
The good reason for redefining check is because it you really prefer to AVOID redefining state aims.
The proper way to redefine state aims (checkmate/stalemate) is to instead redefine ONLY check (a move attribute).
If you start defining new states, it causes unexpected problems.


>Yes. This is exactly how you have replaced "Series-" with a fairy condition.

Exactly? Ha! No, not exactly!
There is no redundant stipulation (or aim) which already provides coverage for the "series-mover" condition.
As I said, redundant elements get no sanction.

>Instead of addressing the issue "it may be done" you have switched to "should it be done?" and "PCCC". Perhaps it shouldn't, the point is that it may be done, which proves (well, at least to me) that the system you propose is just a different kind of syntactic sugar.

Ooh, "syntactic sugar."
That's a fine term.

Hey, I have an idea, let's make perfection the enemy of the good.

I proposed a considerable number of definitions for fundamental terms (fairy condition, fairy element, stipulation, aim, goal, etc), along with a functional methodology to distinguish one from another (in response to the inability of experts to determine whether series-movers were a fairy condition).
These terms form the basis for divisions in albums, and limit the jurisdiction of judges.
You don't think these are important now, because these things don't stop a renegade antiKings specialist from destroying our beloved non-redundancy???

Recall that you attempted to define "orthodox" into "zerodox" (completly detroying its meaning, by making it REDUNDANT to the CORE with an already existing term: "studies").
Suddenly, you have an explosion of zeal to eliminate -- can you believe this -- REDUNDANCY?!!!!!
Dmitri -- AYFKM?????

Explain how redundancy in aims (being redefined by rogue elements) is a threat.
Your entire AntiKing fantasy was syntactic sugar high -- these things can not even happen!
If there were a big problem here, you would have no difficulty in recounting some trace example.

My purpose in this thread is not to propose a system which eliminates redundancy -- do you get that?
I proposed better (and non-redundant!) definitions for fundamental terms.
In fact, I detailed several SANCTIONING METHODOLOGIES, already, which would greatly reduce existing redundancy.
And, I have deliberately withheld some better ideas, because I'd prefer to implement them myself, and send them as a proposal for PCCC.

I also proposed a stipulation methodology, which formally describes countless new stipulations.
Not only that, it allows you to stipulate where the problem ends (thereby dictating your own convention for soundness).

>I see no point in discussing preferences for certain kinds of syntactic sugar.

Don't pour that syntactic sugar on me.
I see what's going on here.
 
   
(Read Only)pid=10009
(48) Posted by Georgy Evseev [Wednesday, Mar 20, 2013 13:18]

@Kevin

It is extremely difficult to read your very long posts.

Let's just just check the following simple practical example.

Which from the following records are correct from your point of view?

1. #3; fairy condition: Circe (traditionally - fairy threemover)
2. #3; fairy condition: Black are idle (traditionally - series directmate in three)
3. #3; fairy condition: White choose the the move for black (traditionally - helpmate in 2.5)
 
   
(Read Only)pid=10010
(49) Posted by Frank Richter [Wednesday, Mar 20, 2013 16:09]

In generally: What would be at least one benefit, if we would make the problem world "complete" as suggested?
 
 
(Read Only)pid=10011
(50) Posted by Torsten Linß [Wednesday, Mar 20, 2013 18:51]

GE: "I do not quite understand the general reason for this discussion."

Neither do I.

Have a look at a typical Popeye input and you'll get a glimps of how things are. What a stipulation is, what a (fairy)condition is, a fairy piece etc. Except for retros that covers at least 99.99% of all chess problems. In practice that's sufficient. Everything else is Haarspalterei.
 
   
(Read Only)pid=10012
(51) Posted by Dupont Nicolas [Wednesday, Mar 20, 2013 20:55]

@ Franck, Torsten and anybody else who doubt that a structured chess problem world is needed.

I already provide a benefit, a new condition coming naturally from a piece of organization [Tuesday, Mar 19, 2013 16:50]

One could imagine plenty of other benefits, the first one being clarity. Is it logical that the number of moves in a series help-self problem is two-fold (counting or not the terminal black move)? Is it logical that, for Popeye, the sign “+” stands for a checking move, but stands for a checking non-mating move for WinChloe?

As a proof game composer I feel good, everything is perfectly organized into my mind, I can work seriously. As an almost beginner in fairy chess I feel troubling. Like if I was into a huge supermarket where red wine is at an extremity and white wine at the other… I just would like to find at the same place everything I want to drink.

And what about newcomers? Don’t they have the right to be completely lost in that jungle? Do you really think that we are doing everything possible to make our discipline attractive?

That's why I feel completely in touch with Kevin's approach, even if our methods to try getting the needed organization seem divergent. Hope we shall dance on the same floor a day or another…
 
   
(Read Only)pid=10013
(52) Posted by Sven Hendrik Lossin [Wednesday, Mar 20, 2013 20:59]; edited by Sven Hendrik Lossin [13-03-20]

Why we need this discussion?

Let's ask the thread-opener.

"There are World Championships partly based on the Codex, which is in need of an update to the competitive requirements of today. The inconsistencies and unfinished matters need to be addressed one day. If it were my decision, it would be sooner; but as it is not my decision, it will have to be later. Until this day comes, we will continue to live in our incomplete problem chess world."

Obviously that is why ;-)
 
 
(Read Only)pid=10014
(53) Posted by Sven Hendrik Lossin [Wednesday, Mar 20, 2013 21:14]

@Nicolas Dupont:
I've also been wondering why our discipline is not so attractive to a lot of chess players. Especially I do not quite understand how somebody, who knows a lot about chess, spends a huge amount of time solving Sudokus when he has taken the first hurdle (that is: having a considerable strength in otb chess) to solve chess problems. But there are good reasons for that but you will certainly not find them in a lack of formalization.
In my opinion the opposite is the case: chess players see unnatural positions and observe that they use a lot of unknown terms in the annotations and comments of the solution. I had to read the solutions and comments of "Die Schwalbe" for more than a year to get a first and small impression what it is all about. It's been worth it but as long as we expect this endurance from interested people I have little hope that our hobby will be interesting for broader groups of chess players. This is certainly dissuasive.
 
   
(Read Only)pid=10015
(54) Posted by Dupont Nicolas [Wednesday, Mar 20, 2013 21:51]

@ Sven

Why we need this discussion? Let's also answer Wikipedia, article “fairy chess piece”.

“Because of the distributed and uncoordinated nature of unorthodox chess development, often the same piece is referred to by different names or the same name is used for different pieces in various contexts”

Right to the point! Coordination is needed (I do prefer the word "structuration" but never mind). It will not imply that suddenly each chess player will leave his Sudoku to enter fairy chess problems, but at least “dispatching” will no more be an excuse…

“they use a lot of unknown terms in the annotations”.

True, and indeed this is an obstacle. In my own structuration process I refer to classical series with the easy minded notations WW for direct, WB for self, BW for helped and BB for “auto-self” (the color of the series side followed by the color of the terminal side).

The BB setting is a variation of Dan’s “auto-zugzwang” series, it is the same as WW except that black has no other option than reaching the goal at its terminal move. An article about it will appear in Phénix.
 
   
(Read Only)pid=10016
(55) Posted by Kevin Begley [Wednesday, Mar 20, 2013 23:14]; edited by Kevin Begley [13-03-20]

@Georgy,

Some posts contained considerable information, but I do apologize for running way overboard on others.

>Which from the following records are correct from your point of view?
> 1. #3; fairy condition: Circe (traditionally - fairy threemover)

Pardon me, but I'm not clear what you are trying to ask. Specifically, how do you mean: is this "correct"?

We have a "traditional circe #3" -- yes?
If you are asking me if "circe" is a fairy element/condition, the answer is: yes/no.
It must contain a fairy element, because it alters the rules of movement (via rebirth).
But, it can not be considered a fairy condition, because the alteration does not does not universally apply (to all units).
The rex inclusive form of circe may be considered a fairy condition.
Circe should probably differ to that, and then exclude the Royal units.
Or, it should simply use circe units.
Let PCCC decide -- I don't have an opinion on which is better.
All I am trying to convey is the meaning of some fundamental terms.

2. #3; fairy condition: Black are idle (traditionally - series directmate in three)

The traditional "ser-#3" is not a valid formal stipulation, because it alters the rules of movement.
Therefore, it must contain a fairy element.
That element is found in the prefix "ser-", and it does apply universally.
Therefore, it may be classified as a fairy condition.

I would stipulate this as: "#3"
I would list the fairy condition: "black idles, except where to achieve a goal"
I would not reduce this to "black idles," because the above form can be used to cover a broader description.
The former condition provides better coverage, because it can be used to describe "ser-s#3."

3. #3; fairy condition: White choose the the move for black (traditionally - helpmate in 2.5)

Are you following Dmitri's failed attempt to suggest that I've introduced some "redundancy" ??

If you are, let me run this by you another way...
Hopefully, this will prove instructive, and help some still struggling to understand that the definitions I have offered would actually EXCLUDE the possibility of this redundancy attempt.

What is proposed here is not a valid fairy condition (nor even a valid fairy element).

A fairy element references or appends a "rule book" describing all the possible movements of chessmen -- just as "FIDE 1997" does.
However, the latter does not reference the complete FIDE rule book -- it only references rules which govern LEGAL options of movement.

They exist to do nothing, except to tell the units how they may legally move, in a given position.
Physics for chessmen -- that's all they are!
Narrowly defined, they can provide no information about what counts as winning, nor do they provide claims...

In fact, the "rules of movement" do not even define players in the game!
All they provide is a complete description of all legal move-options.


Stop looking at this from the perspective of chess -- shudder it!
Think like a PROBLEMIST -- from the perspective of a PROBLEM...
What are the distinct elements required to convey the problem information?

You can not define a fairy element which provides NO ALTERATION (or constraint) to the rules of play.
The rules of play do not provide for a "player" (of either the white or black pieces)!
They do NOT govern PLAYERS -- they are narrowly confined to describe all possible kinematics of the chessmen, (and the chessmen ALONE do they govern!)

Have ANY of the rules of play been modified (or constrained) by your proposed fairy element?
No! Zero change.

Think of the rules of play as nothing but a program which constrains you to legal moves in the game.
It does not recognize players.
It does not arbitrate any claim.
It does not acknowledge any agreement.
It does not dictate win, loss, draw (or the points earned for each).

You are essentially talking to the programmer, asking for a modification to let Wiktor move for Bojan.
The application can not govern them, and thus, the programmer must refuse.
Similarly, The official governing body must refuse to sanction this invalid fairy element.

If you want to appeal their decision, you would be advised to show how this requires any alteration to the present program.

A valid fairy element is something that requires functional reprogramming of the application.
The programmer will happily alter the program to alter the kinematics of chessmen movement.
But, will do nothing else.

With such a "fairy element" rejected, the present (and worrisome!) redundancy problem has been averted.
And, that is no syntactic sugar -- that is nutritional value, which improves our PROBLEMIST world.

I'm sorry if I ran long here -- but, it would help me if your question was better understood.
If the misunderstanding is mine, I do apologize.
And, if your communication is not perfect, I do understand (my poor understanding of Russian is the problem).
 
 
(Read Only)pid=10017
(56) Posted by Kostas Prentos [Wednesday, Mar 20, 2013 23:33]

I don't know about fairy chess, but since sudoku was mentioned, take a look at the following website: http://killersudokuonline.com/index.html with several different types of fairy sudokus one can play online. Killer sudoku in particular is so much fun.
 
 
(Read Only)pid=10018
(57) Posted by Kevin Begley [Wednesday, Mar 20, 2013 23:45]; edited by Kevin Begley [13-03-20]

@Torsten,

>"Neither do I [quite understand the general reason for this discussion]"

Why are we talking about the weather?
Because it interests some of us.

>Have a look at a typical Popeye input and you'll get a glimps of how things are.
>What a stipulation is, what a (fairy)condition is, a fairy piece etc.
>Except for retros that covers at least 99.99% of all chess problems. In practice that's sufficient. Everything else is Haarspalterei.

This is a rather disappointing attitude...
I used to encounter a computer engineering student who would routinely complain:
"Who cares about the NP-Complete classification of a problem?"

For a glimpse of how things really are, he might have told you to peruse some typical source code.

Oh, I have seen popeye's output.
I have uncovered some bugs in popeye's output.
Not ever did I see a definition for the fundamental terminology of problem chess -- including: fairy element, fairy condition, stipulation, aim, etc etc etc...

Whom do you think you're fooling? Popeye does not even define its fairy pieces in the help file!

It is probably true that 99.99% of computer programs are written without any understanding of complexity classifications, for problems.
And, the same constant may apply to chess problems -- 99.99% of judges can not define a fairy element, despite the fact that it is used to define their specific jurisdiction.

The sad fact is, programmers who do not understand the classifications of algorithmic complexity, never even manage to appreciate how this helps them to write better programs.

When will I ever use algebra in my life? === Why does a problemist need to understand problems?

The reason is to make our PROBLEM world better.
Not to nitpick -- to offer improvement!

E = 1/2 * m * v ^ 2 (Newtonian View -- valid 99.99% of the time).
E = m * C ^ 2 (Some Nitpicker).
 
   
(Read Only)pid=10019
(58) Posted by Dmitri Turevski [Thursday, Mar 21, 2013 04:55]

Circe (RI) allows white to move black pieces (1)
Ser-# forbids black to move black pieces (2)
h# is nothing more complicated than (1)+(2)

All h# changes is "a complete description of all legal move-options" - none for black, complex options for white. Physics of the game, if you like.
 
   
(Read Only)pid=10020
(59) Posted by Georgy Evseev [Thursday, Mar 21, 2013 06:37]; edited by Georgy Evseev [13-03-21]

@Kevin

OK, you have answered as I expected, so we may say that we are playing on the same minefield.

BTW, you have drawn the tradiditional well-known line between "heterodox" and "fairy".

The next question:

Dmitri has shown that any stipulation may be reworded as standard #n with a set of conditions (maybe some of them fairy, some of them heterodox).

Where is the obvious line which allows honestly and fundamentally distinguish, when we should apply the conditions and when - change the stipulation?

(I can answer this question myself, but my opinion is subjective and requires understanding beyond beginner's level. So, it is not fundamental and may be not even honest, but it is very practical)).)
 
   
(Read Only)pid=10021
(60) Posted by Torsten Linß [Thursday, Mar 21, 2013 09:33]

ND: "@ Franck, Torsten and anybody else who doubt that a structured chess problem world is needed."

Yes, I doubt that a structured chess problem world is *needed*. However, some aspects of problem chess *do have* a structure that can be *discovered*. As a programmer I had the pleasure of exploring some of these structures. But this insight never helped me as a composer.
 
 
(Read Only)pid=10022

Read more...
Page: [Previous] [Next] 1 2 3 4

MatPlus.Net Forum General Our incomplete problem chess world