MatPlus.Net

 Website founded by
Milan Velimirović
in 2006

20:36 UTC
ISC 2024
 
  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-Jul-2024

B P C F





 
 
MatPlus.Net Forum General Sentinelles PionNeutre
 
You can only view this page!
(1) Posted by Joost de Heer [Wednesday, Aug 16, 2023 15:05]

Sentinelles PionNeutre


Popeye uses the implementation that white pieces leave white pawns, black pieces black pawns and neutral pieces neutral pawns.
Winchloe uses the implementation that any piece leaves a neutral pawn.

Does anyone know which implementation is the correct one? For me, the Winchloe implementation seems more natural.
 
(Read Only)pid=25075
(2) Posted by Mu-Tsun Tsai [Wednesday, Aug 16, 2023 17:51]

There's a relevant Popeye issue on this:
https://github.com/thomas-maeder/popeye/issues/352
and accordingly Popeye has the correct implementation.
 
(Read Only)pid=25077
(3) Posted by Thomas Maeder [Thursday, Aug 17, 2023 15:47]

FWIW, Popeye's implementation is conforming to https://juliasfairies.com/fairy-terms/fairy-classification-project/fairy-classification-tabular/

The reason may be that the text in the Fairy Chess Classification Project was inspired by what Popeye does, though :-)
 
 
(Read Only)pid=25089
(4) Posted by Kevin Begley [Friday, Aug 18, 2023 23:50]

Apologies, but this will be a long post.

It is circular reasoning to use a definition from the classification project to argue that a Popeye issue is not a bug.
That's like the FBI leaking a story to the press, then using the story in the press to justify cause for a wiretap.

The argument made by Mu-Tsun Tsai -- that Popeye has a correct interpretation based upon the Fairy Classification Project -- is not well founded, and can be rejected on its face.
This project is not a legitimate authority, and in their capacity as arbiter of such disputes, this project has demonstrated a total lack of integrity.
They based their conclusions entirely upon precedent and favoritism (specifically a favoritism for the free software tool).
Going by their own hierarchy (precedent), that project has no basis (other than the entitlements of raw favoritism).

Problemesis had defined this condition (see http://christian.poisson.free.fr/problemesis/condra.html) LONG before the classification project.

 QUOTE 

Sentinelles en Pion neutre - When a piece (not a Pawn) moves, a neutral Pawn appears on the vacated square if it is not on the first or the last rank, and if there are less than 8 neutral Pawns on the board.


Thus, we can (and should) reject any claim that the classification project can be considered an authentication agency.


To resolve this dispute, we need a honest authentication agency to consider three questions:

1) What were the rules offered by the inventor of this condition?
2) What were the rules offered by the first composer to use this condition? and (even more importantly),
3) Is there some compelling reason to alter/discard the definition (or naming convention) which was first provided for this condition?

I would encourage giving programmers the right to filibuster any decision made by the authentication agency (if we can't reach an amicable agreement about the naming and rules definitions, then the fairy element in question can not be fully sanctioned).
In the absence of any sanctioning agency, programmers (of solving tools, databases, and other software) are on the front lines of making sense of fairy elements.
I would also give filibuster power to a group of editors (especially from the most established journals).

Excuse me if I can not take the fairy classification project seriously, but they could not even provide a functional definition for what is a fairy element (which would distinguish this from a stipulation).
And, for the record, I was HIGHLY supportive when I first learned of this effort to logically classify problem chess elements (while still a valuable resource, and a contribution worthy of appreciation, that project failed utterly in achieving its primary mission).


Rules (including default rules) and naming conventions should be resolved in the most consistent, most logical manner -- with a long view toward what is best for the artform.

Problems which question the rules should be encouraged -- if any problem group deserves their own genre, it is problems which question our rules/conventions (I can't stress that enough).

Full Disclaimer: every time I make a problem in the condition Circe Parrain (quite often), I worry that the solving tools are inadequate (the rules of this condition -- a condition I love -- are so poorly considered, I often fear I'll be cooked by the very holes I have poked in this condition).
No honest authentication authority could never sanction this condition (I'd still make problems with it, but I'd advocate for some compromise to make this something that can be sanctioned).

Generally, the programmers have resolved these matters, because fairy chess has no authentication agency.
Thus, I would encourage the programmers (Popeye and Win Chloe, specifically) to seek a mutually beneficial resolution (if the agreed naming convention does not align with the first definition provided for this condition, a compelling reason should be provided).

A sanctioning authority would discourage inventors from mislabeling their inventions.

For example, Reflexmate is clearly a fairy condition, not a stipulation (same goes for series movers and parry series movers).

Furthermore, Take & Make is clearly a subform of Anticirce, just as Optimal Replacement is a subform of Circe (their naming convention should be corrected, but only a sanctioning authority can right these wrongs).

Note: the inventor of Platzwechsel Circe did give NOT this condition its name. He argued his condition was not a subform of Circe (this was back at a time when all subforms of Circe had rebirth based upon the Initial Game Array -- his claim was hardly frivolous). He eventually settled for PWC (which may, or may not, be interpreted as a subform of Circe).
This was a bad compromise.
As much as we want to respect inventors, this condition should clearly be classified as a subform of Circe (if you take a broader view of what the Circe precedent had established).

Circe describes any problem in which a captured unit is reborn into the diagram (according to various rules).
Likewise, Anticirce describes any problem in which the capturing unit is reborn into the diagram (according to various rules) -- and (by default) no capture is possible unless this rebirth can occur.

Thus, Take & Make was clearly a subform of Anticirce, but nobody insisted on a consistent application.
This was resolved entirely by the corruption of favoritism.
It's not only an issue that one inventor is favored, another inventor is not -- the real issue is that the BEST INTERESTS OF OUR BELOVED ARTFORM ARE NOT FAVORED!

Instead, we are all diminished by these territorial pissing contests.

I have advocated for a solution to this corruption (and the corruption of our classification system) for around 20 years.
I don't love to argue -- I have no desire to carry on this argument, but I can't let it go because this matters -- nor do I seek a place on any sanctioning authority.
I have no desire to influence the decisions of any sanctioning body (that's precisely why I never volunteered for the classification project -- I feared my presence would have resulted in a product that reflected my own biases; I had hoped I could trust a group of intelligent volunteers to find their own way to a logical classification, never dreaming a few volunteers could persuade the rest to either quit the project or abandon their stated mission).

I only ask that fairy elements (and their default conventions, and naming conventions) be comprehensive, logical, consistent, universal, and elegant (easily understood and easily remembered by newcomers).
If not that, then members of an official agency should be allowed to deny (including by filibuster) full sanction for this fairy element.
That's the only way I see to encourage consistency.

At its inception, I had hoped the fairy classification project was a step in that direction, but instead they took two steps backwards (and essentially sanctioned a system of corruption and favoritism).

No interpretation can be considered correct based upon the output of the fairy classification project (since it was not a classification project, and it failed to produce a definition which distinguishes a fairy element from a stipulation).

Please, somebody, just have the courage to publicly admit this is a big problem, and that we need to do something about it.
If there is a better idea than a sanctioning agency, I'm happy to consider any alternative path to resolve these issues.
So long as we can agree that doing nothing has resolved nothing (it's only kicked the can down the road -- and burdened the next generation of problem enthusiasts to fix an issue we allowed to snowball).

There is no escaping the fact that this will cause some upheaval.
Databases and solving tools will need to alter (and retest) many problems.
Some will require correction, and others will be relegated to a subset condition (whereas they had enjoyed the default rules previously).

This is unfortunate. We should look to avoid such upheavals going forward (which is why I advocate for establishing clear and unambiguous definitions for the fundamental elements of a problem). But, the longer we delay this reformation, the more this burden snowballs.

By doing nothing, we are asking the next generation to either correct our mistakes, to pile on more (as we have done), or to (inevitably) abandon our flawed artform entirely.

Programmers can't always resolve these matters for us.
We could wait for AI (or some outside problem agency) to resolve these flaws for us, but enthusiasts of this artform will not like the results (if you fear the upheaval from a sanctioning body of chess problem enthusiasts, wait until you wrap your head around the existential upheaval threat posed by a foreign agency).

Either tell me why we shouldn't have an official sanctioning agency, or tell me there's a better solution.
 
   
(Read Only)pid=25103
(5) Posted by Mu-Tsun Tsai [Saturday, Aug 19, 2023 02:08]

Wouldn't the reasoning be equally circular to quote from Christian Poisson's website, since he is also the one who created WinChloe?
 
 
(Read Only)pid=25104
(6) Posted by Kevin Begley [Saturday, Aug 19, 2023 02:08]

Why is it so important to unambiguously define the most fundamental terms?

Consider the condition attribute Bicaptures (which can be associated with a unit) versus the color attribute Neutral (which can be associated with a unit)...

Problemesis defines the Bicaptures condition as follows:

 QUOTE 

Any unit can capture the pieces of its side (except its own King if 'Rex Exclusiv') as well as the pieces of the other side.


You can also attribute this condition to a subset of all units.

Why are neutral units considered to be a separate color, but bicapturing units are not considered to have their own separate color attribute?

Can anyone provide a logical definition for the color attribute of a chess unit (without mentioning neutral units or bicapturing units), such that we can distinguish the color attribute (associated with the unit) from the fairy condition (associated with the unit)?

We could define the color attribute such that it determines which player has the right to move a given unit.
Alternatively, we could also define the color attribute such that it determines which unit has the right to capture a given unit.
This choice appears completely arbitrary.

How many colors are possible?
How many colors can a diagram reasonably convey?
If our diagrams can not adequately convey all colors possible, how shall we distinguish color attributes from fairy conditions?

These are the fundamental questions we never address.
We have no definitions for the most fundamental elements of a chess problem, and as a result, our most experienced composers can not distinguish a fairy condition from a stipulation from a color attribute.

Maybe you've created a new stipulation, maybe it's a new fairy attribute, or maybe it's a new color.
The inventor is free to decide (whatever he or she favors -- consistency be damned).
But, the editors and the programmers are also free to overturn the inventor (whatever they favor -- precedent and consistency be damned).

This results in splintering. Alternate interpretations are incompatible with our need to establish a universal problem convention (this was the whole point of formalizing chess problems).

If we have no official sanctioning authority, who can we get to convince a rookie inventor that their new color, or their new stipulation, is really a new fairy attribute?
We certainly can not appeal to the Fairy Classification Project as an authority -- they classified nothing (they claim classification was never their intent).

If we have no definition that distinguishes the color attribute from a fairy attribute, our standard for deciding is raw favoritism (inherently a corruption of our artform).

Suppose an inventor prefers the perspective that bi-capturing is a color attribute -- maybe they want to make a new class of half-neutral units, where a moved neutral toggles its color attribute from Neutral Unit to Bi-capturing Unit.

If we don't establish definitions which clearly distinguish one fundamental element from another, we have surrendered any basis by which we might dismiss any alternative perspective.
We have also surrendered any hope to establish a mechanism to logically, consistently, and universally convey a problem (again, that was the ever loving point of formalized problem stipulation, but failed to follow through).

If no official body can sanction the definition of the unit's color attribute, we are inviting chaos (inventors are free to invent an entire rainbow of new colors for units, and programmers/editors are free to favor colors, from whatever inventors they favors).

Providing definitions for the fundamental problem elements is the first duty of a sanctioning agency.
I asked the Fairy Classification Project to do just that, but they refused to provide definitions for the simplest terms (including for terms they repeatedly use).

I asked the members of that project to provide us a foundation, upon which some sanctioning authority (perhaps it would be them) could be constructed.
They refused.

So, if anyone wants to argue that I should recognize that project as a legitimate authority, please take note: I encouraged them to claim authority for themselves (and allow for others to petition -- and special groups to filibuster -- their decisions), but they consistently refused to define the most fundamental terms (including several terms they used excessively, which remain undefined).

If we can't even agree on the need for a body to resolve interpretation disputes (and publish a consistent set of definitions describing the fundamental elements of a chess problem), then problem chess can only hope our next generation of enthusiasts will display a greater resolve.

More than likely, nothing will happen -- that's been the pattern, which can not be interpreted as a favorable sign for the longevity of our artform.
 
   
(Read Only)pid=25105
(7) Posted by Kevin Begley [Saturday, Aug 19, 2023 02:27]

@Mu-Tsun Tsai,

You ask a good question:

 QUOTE 

Wouldn't the reasoning be equally circular to quote from Christian Poisson's website, since he is also the one who created WinChloe?


This would be circular, if anyone argued that Problemesis was a sufficient basis to declare Win Chloe's interpretation correct.

That is NOT what I have argued. I reject that argument (and that circular reasoning) completely.


The Classification Project based their decisions upon two things:
1) Favoritism, and
2) Precedent (when this could be used to justify their favoritism).

Note that the latter item includes bad precedent (read: admittedly inconsistent rules, conventions, and naming conventions).
They can't provide any rationale for their decisions, when they themselves admit the precedent is bad, nevertheless, they argued against any effort to correct any bad precedent of problem chess (other than precedent that, for better or worse, has been overturned in the past, and precedent which might be altered by favoritism in the future).

They classified nothing.
They provided no definition for the fundamental elements of a chess problem (in some cases, they could not define their own terms).

Poisson's website provided a definition LONG BEFORE the Classification Project was ever conceived (and remember: this project was conceived with a specific purpose -- to classify problem chess elements -- the original purpose was totally abandoned, I am told, by a small group of volunteers who refused to change the name of the project).

If you accept the Classification Project's reasoning, then Poisson's website wins any precedent argument.
Or, at the very least, the burden is on the Classification Project to demonstrate that their interpretation has precedent.

My argument takes neither side. I favor neither interpretation, and neither solving tool.
I argue that a sanctioning authority should be established, and it must ask three key questions:

1) What inventor first defined this condition?
2) What composer first demonstrated this condition? and (MOST IMPORTANTLY),
3) Are there compelling reasons (for the benefit of our artform -- especially to make it more consistent, more logical, more universal, etc) to reject the rules, default conditions, or the naming convention as it was first established?

Further, I argue that major programmers (Win Chloe, Popeye, and some large database programmers), along with a group of editors (for well established journals, who may vote), should have filibuster power -- such that any controversial fairy element is not fully sanctioned.
And anyone filibustering an element should make an argument why this element should not be accepted.

Oh, you can still compose and solve problems that contain elements which are not fully sanctioned.
But, I argue that the desire to obtain a full sanction will encourage inventors, editors, and programmers to do what our artform requires (allow us to convey problems in a more logical, more consistent, more universal fashion).

I never argued for (or against) Win Chloe's interpretation.
I merely rejected your false argument -- the classification project classified nothing, it is not a legitimate (nor even a reliable) authority, and their decisions were never based upon any logical criteria (not even precedent -- it's just a statement of raw favoritism; a better name for it would be the Favorite Color Project for problem chess).
 
   
(Read Only)pid=25106
(8) Posted by shankar ram [Saturday, Aug 19, 2023 04:24]

Thomas, Stupse nicht in den bären!
 
 
(Read Only)pid=25107
(9) Posted by Mu-Tsun Tsai [Saturday, Aug 19, 2023 07:31]

@(7)
To clarify, that wasn't “my” argument either. I was merely pointing to what Thomas had to say. I make no bones to the fact that I'm not knowledgeable enough in this subject.
 
   
(Read Only)pid=25108
(10) Posted by Kevin Begley [Saturday, Aug 19, 2023 10:25]

I have very little knowledge about Sentinels (neutral or otherwise).
I do know that you can not claim that "Popeye has the correct implementation" if that claim rests entirely upon a reference to a definition which appears in the Classification Project.

If you referenced the inventor of the variant, the original definition, or the first problem which used this condition, I'd be the first to admit you have strong support for your claim. And in that case, if the party opposing your interpretation can not dispute your historical information, the burden would be theirs to demonstrate why this interpretation should be rejected.

As I've often stated, the best interests of the artform should come first (consistency is more important than precedent), and we need some official body to look after those interests (to encourage best practices by withholding their full sanction when necessary).

I have no idea if your claim is correct. I'm no expert on this condition, I know almost nothing about its history, and I haven't the foggiest idea why one interpretation might be better for this artform than another.
So, I take no position.

My only position is that we should be mindful how we resolve these disputes, and we should create an official agency which attempts to render an impartial judgement.
Also, I think we need to allow key parties an opportunity to filibuster that process (to effectively withhold sanction), to discourage such an official agency from arbitrarily mowing down interpretations which might benefit us all in the long run.

I also know we should not put too much stock in a source which can not distinguish (with clear definitions) the fundamental elements of a problem (including terms they themselves use) -- sadly, at present, that means we must be skeptical about virtually every source we can find!


Finally, let nobody discourage you from challenging the views of anyone in this forum (especially mine -- I appreciate alternative viewpoints, and I consider lively, respectful debate to be the best methodology for learning).

Bears are dangerous creatures which can harm you. I an not such a creature. My ideas are not bear claws. I resent any contrary insinuation (which I consider a sinister attempt to both personally offend me, and to deprive me from receiving honest feedback).
I consider that insinuation to be a savage mauling of my good character. Am I attacking anyone personally for it? No. I am not.
I merely point out that this is shameful behavior, from one of our good companions. I'm willing to forgive...

I understand that some parties involved in the misnamed classification project believe I don't appreciate their efforts. That's entirely false.
I respect them all. I appreciate their efforts. And I have often stated that good things did come from it (despite the failure to deliver what was originally promised).

But, I will not pretend this project merits a greater reliability than a definition which appeared on some old website. It does not.
This project could not provide definitions for their own terms (for the most fundamental elements of a chess problem), and failed to find a primer by which the reader could determine which problem elements belong in which fundamental buckets (they couldn't even define their buckets). That's just the honest truth.
I offered an honest criticism of that project's output. It was never personal.

I try to avoid the topic, frankly, but in this thread, that project was mentioned as a source which validates an interpretation (against the report of bug).
I simply maintain that this bug report (which might be credible, or might be entirely incorrect) can not be invalidated by any classification project definition.

And I have advocated for an official body which could help amicably mitigate such disputes (especially when they turn personal and/or territorial).

I reiterate: if anyone has a better idea (than a sanctioning agency) to discourage the splintering of our terms (we all have an interest in seeing that the terms which describe a chess problem are universally accepted), I would be happy to learn of your alternative approach (either in this thread, or in a new thread -- whatever is deemed an appropriate venue).
 
 
(Read Only)pid=25111

No more posts


MatPlus.Net Forum General Sentinelles PionNeutre