Comments/Ratings for a Single Item

Does anyone know how to run chessV under linux ?

mono ./ChessV.exe

results in "Directory of piece set graphics could not be found"

Greg is not here, but has anyone else managed to implement the R2thenBishop and the B2thenRook in chessV?

No matter I solved it!

@Greg,

Are you here?

Hello Greg,

How would chessV will understand pieces like R2thenBishop or B2thenRook? What about ignoring the first few step of these like in the Tamerlane chess giraffe?

@Greg: Could you imagine to map my variants 'Conquer' and 'Bull's eye' in your ChessV? I would be glad.

A much better player than Zillions. I notice that version 2.2 is scriptable, and the scripting language looks easier than Zillions' Lisp-based scripting language (which is a monumental pain in the a** to debug--all those nested parens). Does ChessV have a scripting reference? I'd love to see it rather than ask a thousand questions in this thread.

Hi Greg,

In the latest version I have the FEN does not contain anything about the last piece moved. The AI does not know how to checkmate in KJK.

[EDIT] I don't think I have the last version!

Hi Greg,

A while ago you have said that maybe you'll try to make two ai play on different threads, so that they'd play different parameters. A nice feature paired with that, would be the ability to force the set of parameters. Maybe some could still remain random, if they are hard to do.

Hi, Greg! Thanks again for putting some of my games in ChessV. I deeply appreciate it. Gilded Grand Shatranj was a spur of the moment design, over the course of some minutes, quite literally exactly as it appears on the Grand Shatranj rules page, and almost that fast. That's the only write-up of it that I recall. If I actually wrote up all the games I've done while here as separate game pages, I'd show twice as many as appear here, or more. Many of the games are part of a series or variants of a particular game and/or style. Truthfully, many of my games are designed to illustrate the ideas as simply and easily and most familiarly as possible.

Thanks, Fergus. Been tied up by work as of late. Some of these have been pending description for a long time. Here's a quick summary:

• Duplex Chess -- a new double-move variant that I've invented and need to document.  I wanted to make a balanced, double-move variant that wasn't too overwhelming.  This is managed in this game by the fact that no piece has a range more than 2, and the same piece cannot move on both of a player's moves ... with the exception of the King.  The King can move twice making him a viscous attacker.  (He can hit-and-run, unlike any other piece.)  But exposing him has risks, since if he falls, the game ends instantly.
• Gilded Grand Shatranj -- this is definily a Joy Joyce original, but doesn't have it's own page.  The text on the Grand Shatranj you mention is probably the source.
• New Zealand Chess -- From Pritchard's Encyclopedia of Chess Variants.  "Rook becomes Knight and N becomes R when capturing only.  R & P endings described as 'bewildering' (BCM Sep 1903)." Is all the info I have.
• Nightrider Chess -- Also from Pritchard's Encyclopedia of Chess Variants.  V. R. Parton (1950s)
• Odyssey -- This is mine, it's a complicated, ambitious invention that needs a page. 
• Opti Chess -- This is a differnet capablanca-array with flexible castling.  Derek put it forward in comments probably close to 20 years ago now.  It turns out that this array works really well (in terms of most options for good openings.)
• Relative Royalty Chess -- This is weird.  It grew from ICS's TwoKings variant, for which the rules were a random outcome of how ICS handled the presence of more than one king.  But "TwoKings" had problems (not surprising, since the rules were kinda random).  Relative Royalty Chess was an attempt to fix those issues.  I made it only because when first published, an enthusiastic user on Talk Chess loved TwoKings and wanted this.  I don't know if anyone plays it, or knows, or even cares.  Maybe it should go away.  But, if not, it should be documented.
• Unicorn Grand Chess -- This is the Unicorn Great Chess pieces with a more Grand Chess setup.  (Pawns on the 3rd ranks, Rooks start connected on the 1st, pawns promote on the 7th by replacement under Grand Chess rules.)  This was my idea to adapt Uncorn Great in this way.  David ok'd the invention.  As much as I like Unicorn Great, I think this is even better.  Hasn't really been played here, but it has been built into ChessV for nearly 20 years.
• Warochess -- This was emailed to me.  It was also submitted here, but if memory serves, it wasn't published on the basis that it was too close to other variants.
• Wild Castle -- This is another thing to spring from ICS.  It's basically like FRC except there are only 18 setups.  (Rooks are always in the corner.)

Since Greg hasn't responded to my earlier question, I went ahead with adding a ChessV tag to games playable on ChessV.

Here are the games on ChessV I did not find a page for here:

• Duplex Chess by Greg Strong
• Gilded Grand Shatranj by Joe Joyce (This is briefly mentioned on the Grand Shatranj page, but the description is too brief, and the Game Courier preset it links to has no landing page, does not describe the rules, and is not programmed to enforce the rules.)
• New Zealand Chess by Unknown
• Nightrider Chess by V. R. Parton (maybe it's in one of his documents, but it doesn't have its own page.)
• Odyssey by Greg Strong
• Opti Chess by Derek Nalls
• Relative Royalty Chess by various
• Unicorn Grand Chess by David Paulowich and Greg Strong
• Warochess by Eric Warolus
• Wild Castle by Unknown

With this complete, I have edited the header to use the ChessV tag to tell whether a game is supported in ChessV, and I have removed the Chessv column from the Item table.

As a demo for how a ChessV tag could be used, I created and populated a Zillions OTB tag for games that come with Zillions-of-Games, and I added code to add an item to the Play menu for games that come with Zillions-of-Games.

Greg,

It appears your site doesn't support https. An easy way to fix this is to use CloudFlare for your DNS, as this site does. I have a free CloudFlare account that lets me use flexible SSL/TLS for free without hosting any files on this site.

Also, do you think it would be better to use a ChessV tag for games supported by ChessV? I created a column in the Item table for you to mark games supported by ChessV, but you haven't been doing that. I just marked Cylindrical Chess as supported by ChessV, and now the Play menu lists it as playable on ChessV.

The "variation" setting should not result in it making bad moves. The "weakening" setting certainly can if you turn it up. Regarding endgames, when the material gets down to only a few pieces, ChessV can have difficulty determining how to close the deal. I need to post a new build - I have made some definite improvements in this area.

I have observed that, too. I think it because it chooses among positions that are already winning!

I've noticed ChessV making some obviously bad moves in endgames. Could that be due to the variation setting?

Thanks!

If you have at least one colorbound piece on each shade, there is a half pawn bonus. If you have two or more colorbound pieces on one shade with none on the other, there is a large penalty.

@Greg,

Have you noticed my previous comment!

@Greg, I have searched the source code and the documentation but I did not understood is the evaluation for the colourbound pieces pair is a bonus or a penalty. I need to know this in my bishops vs knights (for example) experiments.

I'll look into it more deeply hopefully!

Presumably, there was some reason that was not apparent. No way for me to say.

I've seen a game where a pawn has promoted to joker and not to rook for no apparent reason. Any insight towards why?!

The king has a value of 0.

Greg, what is the strength of the king used by the joker value evaluation?

Ok, HG, thanks!

Short follow-up on posting diagrams: You must only write the board part of the FEN between the tages; the other stuff contained in a FEN would not be expressed in the image anyway. When I place your FEN between the tags I get this:

These are not the pieces you want; apparently the standard meaning for C is Cannon, for S Squirrel, and for J it is Camel. But that doesn't have to discourage you: you can either try an alternative likely letter (like M for Marshall), or replace the letter by the full name of the Alfaerie image in braces ({berolinapawn} and {fool}:

[ fen]3{FOOL}3k1w/10/10/8p1/2P5Pp/1{SERGEANT}7P/5{sergeant}4/10/4m4{fool}/7BK1[ /fen] gives

And because it was facing a difficult position white decided to go for a draw. Thanks Greg!

That's right. If white's last move was a chancellor, then it is a stalemate by either the old rules or the new rules.

White's previous move was by a chancellor from h8 to h10. That chancellor got captured. I now think I know what you mean, is what's the black joker's current power. The stalemate/checkmate status depends on that. And because the last moved white piece until white moves again is a chancellor that does not attack the king. We have a stalemate here.

No matter about my previous comment.

I have no idea what you mean

Agreed, but I'm not sure if it is a different stalemate.

Ok, diagram updated with Chancellor.

What was White's last move? The version you have doesn't have the updated rules yet, so the black joker can't move. In the updated version, the black joker will have the ability of white's last move, which could make it a checkmate if it attacks the white king. Otherwise, it is still a stalemate under the new rules. Any piece white moves would lead to white's king being in check by the black joker. Even white moving joker (imitating a king) would make black's joker imitate a king and therefore it is still check. Since white cannot move, it's a stalemate.

The short story is that black has just captured an chancellor on h10 and white is to move. The version that I have (I think is the last one published) gives now a stalemate. The long story is that any piece moved by white leads to a king capture by black in the next move. Here actually the black chancellor on e2 has a role to play. The difference from the studies we already have is that a role is played by the white joker who inherits and transmits a king (more actually a man I think) power. Is that in the second case a checkmate or a stalemate? It looks like ChessV as white has sacrificed it's chancellor for an easy draw, but is this a way for black to checkmate the opponent? And most important how should it be? I am confused by this myself. Depending on what you can do Greg, I am ready to adjust the rules in these rare cases. The most important thing is though, that the rules are clear for everybody.

And c is a chancellor, but this actually has no importance!

This should be close enough for discussion:

The fen of the position, as is found in chessV, is this:

3J3k1w/10/10/8p1/2P5Pp/1S7P/5s4/10/4c4j/7BK1 w - - 0 78

but I did not manage to do it. Here J stands for joker, w stands for wizard, s stands for maasai pawn. The rest are self explanatory.

Indeed, it automatically invokes the Diagram Designer, with default settings. For a static image that should be good enough. You can also use the Interactive Diagram as a game viewer, by including a parameter moveList= , followed by all the moves of the game. This preloads the game in the Diagram, and you can then use the buttons in the AI panel to navigate through it.

Oklh, that is using the diagram designer, isn't it?

You just write a FEN of the position between [ fen] and [ /fen] tags, somewhere in the text of your comment.

I think I have encountered a new situation involving the joker. I want to post a png with the position but I don't know how. Can someone tell me how to add a picure in the comments?

Ah, I see. That is 79%, which is way outside any statistical error bar.

So for 106 games it would be around 4%, so that the 62/106 (=58.5%) score is about two standard deviations above equality

What I meant was that white won 62 games more than black. The actual numbers were 80 white wins, 18 black wins, and 8 draws. This was very different from the first test, where black won more, but with a much smaller margin.

This is an anomalous result; if the games were really independent, the statistical error should decrease as 1/sqrt(numberOfGames). So for 106 games it would be around 4%, so that the 62/106 (=58.5%) score is about two standard deviations above equality, while the other score pointed to equality with a 3% standard error. The standard error in the differens of the two results should be about 5%, so the 58% is off a bit more than you would expect, but not extremely so.

What I often did to make the games more independent is play them as shuffle games. If you shuffle white and black independently (as seems natural for CwDA) you can create a lot of starting positions even when you leave King and corner pieces in place.

How many is a lot? I did one set of 200 games and had a final score of -2, but then I tried a second set and by the time it got to 106 games the score was 62.

What would be good settings for evaluating the balance of a different armies variant? I've tried some tests with 0:30+10 and Medium variation but got conflicting results.

Your results are understandable.  Although the armies are not balanced, they are close, so you need to play a *LOT* of games to get valid results that are not swallowed up by noise.  At a time control of 30+1, you can't play many games.  And such long controls aren't necessary since with games on an 8x8 board ChessV can reach a search depth of 10 in a fraction of a second on a modern computer.  But the other thing to watch out for is that you aren't playing the same game over and over.  The search variation setting helps, but I also use different pre-calculated opening lines.  I will post more about this soon ...

What would be good settings for evaluating the balance of a different armies variant? I've tried some tests with 0:30+10 and Medium variation but got conflicting results.

Thanks!

May you specify if the new version is ready? I am on the tip of my toes for that!

I'm sorry, I know you're anxious.  I will post as soon as it is ready, but I can't say when exactly that will be.

May you specify if the new version is ready? I am on the tip of my toes for that!

Hello Greg,

Have you an idea about when you will release the next ChessV version?

@Greg, If you allow customizable low material evaluation like KMK (M is the CW), may you make it with lists like in KWWK(W is the CF)? I'm sure you though about that, too, and it is a bit late to raise this problem, but it could turn helpfull!

But it should become easier to find cut moves when the evaluation is constant. If all leaves would be evaluated as 0, the first random sequence of moves will be the PV, and every move you randomly pick in a cut node would immediately be a cut move.

What is the problem with a high branching factor? You can search less deep in a given time, of course, but that also holds for any opponent.

I was surprised by just how shocking slow it became.

With alpha-beta the EBF only grows as the square root of the typical number of moves. So even if the latter is 6 times higher, the EBF would only go up by a factor 2.5.

The square root growth is a theoretical value, and while it may be commonly achieved, it is certainly not guaranteed.  What I suspect is happening is this.  The board is very large, the armies start far apart, and no additional evaluation parameters have been added.  So I suspect that there are just a lot of moves where the evaluation is the same so we get way less beta cut-offs.

What is the problem with a high branching factor? You can search less deep in a given time, of course, but that also holds for any opponent.

With alpha-beta the EBF only grows as the square root of the typical number of moves. So even if the latter is 6 times higher, the EBF would only go up by a factor 2.5.

More than 16 files is also problematic. It would increase the size of a data structure that stores pawn information.

Thanks for looking at it anyway

Ok, I think we can forget about 16x16 for now.  I tried ChessV with the 16x12 Double Chess...  OUCH.  It barely works at all.  The branching factor is just too high.  Even the opening position has 72 legal moves (compared to 20 for Chess).  I suspect midgame positions could easily have over 200 although I didn't get that far.

More than 16 files is also problematic.  It would increase the size of a data structure that stores pawn information.

I would only add that H.G.'s #1 can correspond to either my Option 2 or Option 3, depending on how the Joker is defined. (Whether it emulates the last piece moved by the opponent or the last piece moved period.)

Since Aurelian has opted for Option 2, and I'm not hearing any objections (I don't think H.G.'s post really deviated from this), it would be nice to consider the matter settled - at least for Apothecary and for the "default" definition of the Joker. Other game inventors may decided to do different things of course.

EDIT: For the record, these are the potential disadvantages to this option that I've heard: (1) It requires storing state information for the type of piece last moved by both players, and (2) There are circumstances where a Joker can emulate a piece that is no longer on the board (although only when the other side is on the move).

2. The turn pass is considered a move of the King

Option 2b: the turn pass can be made by any piece (or any piece w/ a pseudo‐legal move) on the passing side. And is thus, for mate‐detection purposes, equivalent to the intersection of all such moves.

This has the advantage of needing less off‐board state to be maintained: you only need to record what the opponent's last move was, not your own (as with e.g. en‐passant or lion anti‐trading); also a Joker cannot then give check as a piece no longer on the board, which I find a mildly surprising behaviour, and the position in Greg's diagram is unconditionally checkmate. It also matches my proposed update‐on‐touch‐move semantics, which covers castling out of or moving through check, even in the presence of multiple differently‐moving castling‐capable royals, as well as a possible rule for interacting w/ e.g. Orphans

Conversely, opt. 1 has the advantage of being considerably easier (I imagine) to implement, and probably to explain, at the expense of in some ways exhibiting more surprising behaviour. And Daniel's equation of it with a double move makes some sense (though given the context of actual double‐move variants there are possible quibbles). I imagine it'd be the most popular option.

I agree that of the options H.G. listed, 1 seems most natural; I find my opt. 2b a touch moreso, but opt. 1 is not far behind, so people's mileage may (and probably will!) vary

I was thinking of trying some very wide boards, like 8x20 or 8x24, which would fit the 192 squares limit. Is the reason for the 16 file and rank limit that larger dimensions might be more difficult to display clearly?

No, this was not the primary reason.  It has to do with the size of various pre-allocated arrays.  I don't actually remember the implications of more files very clearly ... it might not be that bad.  I will take a look.  Exceeding 192 squares would be far more costly.

The way I have always looked at this is that to know whether a square is attacked by the opponent during your turn you should pass that turn, and see whether the opponent then can move there. So the question then is what the turn-pass would do for the movement capabilities of the Joker.

1 seems most natural to me. I would think that passing a turn shouldn't count as moving a piece, so it shouldn't change the Joker's move if it is defined by the last piece moved. In a game that doesn't allow passing, passing should be considered the same as a double move by the other side.

Both. You are limited to a maximum of 16 files and 16 ranks. But there is also a limit of 192 squares, so you can't really get 16x16. The largest size is essentially 16x12 or 12x16. Someday this might be increased, but not any time soon. Going to 16x16 would not be too large a code change, but it would have a performance cost for every game, even small ones.

I was thinking of trying some very wide boards, like 8x20 or 8x24, which would fit the 192 squares limit. Is the reason for the 16 file and rank limit that larger dimensions might be more difficult to display clearly?

Thank you Greg, indeed I didn't have it. Now it's alright. Smiling, I say that I will never use a joker. Too much controversial. :=)

The way I have always looked at this is that to know whether a square is attacked by the opponent during your turn you should pass that turn, and see whether the opponent then can move there. So the question then is what the turn-pass would do for the movement capabilities of the Joker. There are several cases that could be argued for:

1. The Joker keeps its move from the previous turn, because no piece was moved during the turn pass, leaving the 'last-moved piece' just as it was. In Greg's diagram black would be stalemate if the Joker had moved to b2 through a Knight move (e.g. imitating a black Knight that just moved to b2 while capturing it). But checkmate if it had been capturing a Bishop that just moved there, as it would then still be imitating that Bishop, and thus check black.
2. The turn pass is considered a move of the King (like in Chu Shogi it would be a move of the Lion). The Joker would then always check like a King during the opponent turn. In Greg's diagram black would always be be checkmated, even if it had moved there through a Knight move (e.g. imitating the Knight that was on b2 while capturing it).
3. The Joker must imitate the null move. It would then not be able to deliver check at all, and the position would always be stalemate. (During its own turn it could still capture a King, though, through whatever move it has then.)

I have not understood how the joker is moving. I understand Option 2 and 3, but what about 1? Let's imagine the Owner moves a Rook, then the opponent moves a Pawn. How is the Joker moving now if it has no moves at all?

Ok, one last try.  The question is how a Joker moves when it is not that player's turn.  When it is white's turn to move, how does black's joker move?  After white moves, of course black's joker moves like whatever piece white moves.  But when it is still white's turn, how does black's joker move?  Under option 1, it has no moves at all.

This usually makes no difference because black pieces don't move when it is white's turn anyway.  The only time it matters is for deciding if the current player is in check.

It is black's turn to move.  Is he in check?  That depends on how the white joker moves at this moment.  Under option 1, the white joker has no moves because it is not his turn.  So black is not in check.  The game is still over!  If black moves, then the white joker will again be able to move like a king and black will be in check.  It is not legal to move into check, so black has no legal moves and the game ends.  But it ends in stalemate, not checkmate, because he is not in check at this moment.  It would be the same outcome if the white joker could move, but was currently emulating a rook.

The only game I'd like to invent of this size would be an 16x16 Chu Shogi variant, named apothecary of course. But for this it is way more important to see chu shogi in ChessV which is in itself stretchy!

Isn't there a more sensitive choice to have a ChessVBig build separatly? This would not cut the performance for chessV small and allow users to have speed in the smaller boards case and also allow to implement 16x16 games.! 

Sure, this could be done, but I am not enthusiastic about maintaining two versions.  And I am not sure this is a problem in practice.  There are other reasons it is not going to play Dai Shogi or Tenjiku Shogi.  I'm sure there are less crazy 16x16 games, but I still don't know how well it would work.  ChessV is well-written and modern computers are powerful but at some point exponential growth still overwhelms you.  I don't think 16x12 has even been tried.  You could try this and see how it goes.

Both.  You are limited to a maximum of 16 files and 16 ranks.  But there is also a limit of 192 squares, so you can't really get 16x16.  The largest size is essentially 16x12 or 12x16.  Someday this might be increased, but not any time soon.  Going to 16x16 would not be too large a code change, but it would have a performance cost for every game, even small ones.

Isn't there a more sensitive choice to have a ChessVBig build separatly? This would not cut the performance for chessV small and allow users to have speed in the smaller boards case and also allow to implement 16x16 games.! 

A (late) question to Greg. In your Option 1: Option 1: The Joker is "reinitialized" each time the owning player moves, returning to its initial state of having no moves at all. This is how ChessV currently works, and I think the Game Courier preset as well. Personally, I consider this a reasonable option, but it seems others do not, and it has the unfortunate consequence of meaning that King+Joker cannot checkmate a bare King.

I have not understood how the joker is moving. I understand Option 2 and 3, but what about 1? Let's imagine the Owner moves a Rook, then the opponent moves a Pawn. How is the Joker moving now if it has no moves at all?

What board dimensions does ChessV allow, and is that determined by the total number of squares or are there strict limits on the width and height?

Both.  You are limited to a maximum of 16 files and 16 ranks.  But there is also a limit of 192 squares, so you can't really get 16x16.  The largest size is essentially 16x12 or 12x16.  Someday this might be increased, but not any time soon.  Going to 16x16 would not be too large a code change, but it would have a performance cost for every game, even small ones.

What board dimensions does ChessV allow, and is that determined by the total number of squares or are there strict limits on the width and height?

As for a Xiangqi Joker; this doesn't seem decidable by logic. It depends on whether you consider confinement as part of the move rules, or as an independent property of the piece type. Like the Joker would not mimic the royalty either, and can use a King move to step to an attacked square any time it wants. But of course you could make rules that would mimic any properties of the last-moved piece type.

I agree 100%.  I was just making the point that the XBetza "I" atom doesn't answer all questions.  I think the XBetza page should probably note that different implementations may vary.

consider a game featuring both a joker and an orphan

Yeow.  Just as a practical matter, a game with both a Joker and an Orphan is unlikely to ever be supported by ChessV, and that probably goes for most any universal chess engine that wasn't built specifically for that.

ChessV has fundamental movements that are handled natively by the internal move generator, but some capabilities that you would like to be intrinsic to the piece itself are implemented by special Rules that are plugged into the game.  For example, the internal move generator does not understand capture-by-advance.  So the Advancer in Butterfly Chess gets this capability via the CaptureByAdvanceRule.  The ImitatorRule doesn't "know" anything about the CaptureByAdvanceRule, so if you drop a Joker into Butterfly Chess, it will not be able to imitate the Advancer's power to capture by advance.  This is unfortunate, but it is just a practical reality.  In order for ChessV to do everything that it does as well as it does, and without bugs, some functions need to be encapsulated in a strict separation-of-concerns.

Besides, if we did contemplate a game with both of these pieces, I bet we could come up with a LOT of complex corner-cases where it isn't clear how they should interact anyway.  (But that's just a hunch.)

Ok, if there are no objections, this will be the behavior of the ImitatorRule in ChessV

No objections from me

as I said, this whole issue is a moot point. You could probably play hundreds of games before you encounter a situation where a Joker check would or would not make castling illegal or would checkmate or stalemate depending on this.

In general I don't think much of rules or rule complications that have next to zero effect on actual game play. I would always go for simplicity when it does not matter.

It's effectively moot in the Apothecary games, sure. Nevertheless it's not so difficult to contrive games where the issue would carry greater importance: consider a game featuring both a joker and an orphan. Such a game would have an equivalent problem with determining under which conditions the joker threatens — and thus relays moves to — the orphan, which one would imagine would have a substantially greater effect on gameplay (especially if it's a gimmick game with several of a few different imitators).

The latter case is (much like the orphan itself, and perhaps even the joker) ofc of even greater interest to problemists than variantists, and they don't really tend to have much of a presence here

Those sound great!

Option 2: The Joker retains the move capabilities of the last piece moved by the opponent.  Therefore, both Jokers always have moves (except on the very first move of the game), and the two Jokers will not necessarily have the same moves.  This is consistent with the description on the Apothecary Chess Modern page.  But I don't think the description on Wikipedia is sufficient to distinguish between this and:

I think this is the implementation I want for my games! 

Ok, if there are no objections, this will be the behavior of the ImitatorRule in ChessV.  (I do not really want to support different flavors of imitators if I can help it.)

As long as we are revisiting the Joker, I would also like to address the pawn issue.  I would like to avoid the Joker imitating the pawns differently in different Apothecary games if we can.  I think the easiest and cleanest way is for me to add an ability to specify a position-dependant move by a Rule so that it won't be imitated.  This way the game designer can have it either way -- as a regular MoveCapability of the piece (if you wanted it imitated) or as a Rule (if you don't).

That means that the Interactive Diagram isn't compliant with other rules of Apothecary.  That said, I think that is fine.  It is unreasonable to expect a simple "prototype" testing engine like the ID to enforce very nuanced rules exactly for all games.  But I do think the question is not as clear as you make it out to be.  If you add a Joker to Xinagqi, how does it imitate the King?  Is it restricted to the palace?  If it is not currently in the palace, can it move?  Does it "check" the opponent king across an open file?  Only in the palace?

Indeed, I don't worry about that. The purpose of the ID is not to provide an exact implementation of any particular variant. Just to provide a set of tools that makes it possible to get a fair approximation. And, as I said, this whole issue is a moot point. You could probably play hundreds of games before you encounter a situation where a Joker check would or would not make castling illegal or would checkmate or stalemate depending on this.

In general I don't think much of rules or rule complications that have next to zero effect on actual game play. I would always go for simplicity when it does not matter.

As for a Xiangqi Joker; this doesn't seem decidable by logic. It depends on whether you consider confinement as part of the move rules, or as an independent property of the piece type. Like the Joker would not mimic the royalty either, and can use a King move to step to an attacked square any time it wants. But of course you could make rules that would mimic any properties of the last-moved piece type.

Yes, I'm sorry, this is more complicated than I'm prepared to implement

I expected as much :)

I'm concerned people won't sufficiently understand it

That is my main reservation with it as well

Look how hard it has been to even get everyone understanding the current issue, and we are all experienced players of chess variants

To be fair, most variants are in this respect noticeably simpler; temporal imitators raise some very subtle timing‐related issues that ‘normal’ pieces can easily ignore

H.G. did present another idea which we could call Option 4 - for purposes of check determination when the other side is on the move, the Joker is always considered to move as a King

Indeed, I saw. For moving out of check, this proposal is equivalent to mine as H.G. has since noted; for mate, it differs only in that the null move (per H.G.'s explanation) can be performed by any piece (with a pseudo‐legal move — though I suppose it'd be a valid simplification to allow it to simply be any piece), not just the King.

(But I don't think this will be a popular option.)

Indeed; it seems a tad artificial to me. My proposal eliminates a bit of the artifice at the expense of some definitional clarity. Which is a tradeoff that I can understand one might be reluctant to make (especially if, as in your case, one finds the J distasteful in any case ;‌) )

If you add a Joker to Xinagqi, how does it imitate the King? Is it restricted to the palace? If it is not currently in the palace, can it move? Does it "check" the opponent king across an open file? Only in the palace?

That depends on how much of the restriction is considered to be a property of the General and how much is considered a general game rule. There's no general consensus on where that line lies; I think all of the particulars you list have been interpreted in a variety of ways by different extrapolations

If you add a Joker to Xinagqi, how does it imitate the King?  Is it restricted to the palace?  If it is not currently in the palace, can it move?  Does it "check" the opponent king across an open file?  Only in the palace?

I was thinking just that, too! 

NOTE: When this conversation is concluded, I will move comments to more appropriate places (Apothecary, XBetza, etc.)

The ‘castling out of check’ case suggests to me another definition again ... 

Ofc this is probably horribly inefficient to program

Yes, I'm sorry, this is more complicated than I'm prepared to implement.  And I'm concerned people won't sufficiently understand it.  Look how hard it has been to even get everyone understanding the current issue, and we are all experienced players of chess variants.  H.G. did present another idea which we could call Option 4 - for purposes of check determination when the other side is on the move, the Joker is always considered to move as a King.  (Sorry if I'm not understanding that right.  But I don't think this will be a popular option.)

I don't think the XBetza definition is ambiguous. It says "all moves of the most-recently moved piece type". 'All' really means 'all', so including double-pushes and e.p. captures.

That means that the Interactive Diagram isn't compliant with other rules of Apothecary.  That said, I think that is fine.  It is unreasonable to expect a simple "prototype" testing engine like the ID to enforce very nuanced rules exactly for all games.  But I do think the question is not as clear as you make it out to be.  If you add a Joker to Xinagqi, how does it imitate the King?  Is it restricted to the palace?  If it is not currently in the palace, can it move?  Does it "check" the opponent king across an open file?  Only in the palace?

in effect that the K can be captured en‐passant on its starting square

This is exactly how I implemented castling in Fairy-Max and the Interactive Diagram. So these would not castle away from an adjacent enemy Imitator, even if it just had landed there through a Knight move. This is an ambiguity in the definition of castling, however. The condition for legality of castling could have been formulated as "when the King could be moved to any square between and including its origin and destination without exposing it to capture". As this describes a stay at the origin as a (null-)move of the King, a Joker should use a King move to capture it even there.

I don't think the XBetza definition is ambiguous. It says "all moves of the most-recently moved piece type". 'All' really means 'all', so including double-pushes and e.p. captures. (I do not consider promotion part of the move.) It also mentions that it imitates the piece type, rather than a particular piece. So it doesn't matter whether the piece that moved still was able to make its initial move; it is the Imitator that should satisfy the conditions for making the initial move.

This is also how I implemented it in the Diagram: the I atom in the move decription of the Imitator's move is simply replaced by the move description of the piece it imitates, as if the Imitator was of that piece type.

Problems could occur with piece types that do not have a fixed move, like the Elk in Elk Chess. Because the Diagram treats that as two piece types, that promote to each other depending on where they land. The implementation would then imitate the type the Elk promoted to, even if it was on a square shade where a 'unified Elk' would have moved differently. I see this more as a problem / ambiguity in the implementation / definition of the Elk than in that of the Imitator.

Option 2: The Joker retains the move capabilities of the last piece moved by the opponent.  Therefore, both Jokers always have moves (except on the very first move of the game), and the two Jokers will not necessarily have the same moves.  This is consistent with the description on the Apothecary Chess Modern page.  But I don't think the description on Wikipedia is sufficient to distinguish between this and:

I think this is the implementation I want for my games! 

The ‘castling out of check’ case suggests to me another definition again: if we consider the restriction on castling out of check to be an extension of the restriction on moving through check — in effect that the K can be captured en‐passant on its starting square, upon having decided to move — then check for castling purposes would always be calculated with an effetive K move for the joker. Iow the move of the J, and whether it gives check, is defined at ‘touch move’ time.

For the case of determining checkmate or stalemate, that would mean that the J would give check with the intersection of the moves of all pieces able to move pseudo‐legally. As such KJK would still lead to checkmate, but with most other combinations of material the J would not give passive check at all. But in e.g. KJKQ the J would give passive check as a K. With more complex pieces (esp. those that can be blocked — particularly the Vulture of the large Apothecary games) this would potentially be position dependent. (And I think it's a little subtler yet in the hypothetical case of a game with both a joker and two royals with disjoint movesets)

Ofc this is probably horribly inefficient to program, provided it's even deemed to make sense (I like it for castling‐out‐of‐check restrictions, but I'm ambivalent between it and Option 2 (by Greg's numbering) for check‐/stalemate — it avoids the surprising(?) behaviour of HG's example where the J checks as a piece that's no longer on‐board, in exchange for arguably slightly greater opaciity of definition), so take it or leave it :‌)

Ok, I think we are all on the same page now.  Yay!  So we have three different options to choose from.  The difference is what powers of movement a Joker is considered to have when it is the other player's turn.  This only matters for purposes of deciding if that player is in check, or moves through check when castling.

Option 1: The Joker is "reinitialized" each time the owning player moves, returning to its initial state of having no moves at all.  This is how ChessV currently works, and I think the Game Courier preset as well.  Personally, I consider this a reasonable option, but it seems others do not, and it has the unfortunate consequence of meaning that King+Joker cannot checkmate a bare King.

Option 2: The Joker retains the move capabilities of the last piece moved by the opponent.  Therefore, both Jokers always have moves (except on the very first move of the game), and the two Jokers will not necessarily have the same moves.  This is consistent with the description on the Apothecary Chess Modern page.  But I don't think the description on Wikipedia is sufficient to distinguish between this and:

Option 3: The Joker imitates the last piece moved by either player.  This is how the Interactive Diagram currently works and is consistent with the description of the 'I' atom in XBetza.  (That said, I assume there is still ambiguity in the XBetza definition because of questions like whether it imitates the two-space pawn move, can capture en passant, castling, or any other game-specific rules.)

[Edit]: The diagram I posted is a checkmate in both options 2 and 3, but a stalemate in option 1.  The situation H. G. posted is a checkmate in option 2, but a stalemate in options 1 and 3.

Personally, I have no strong preference between these options.  Option 2 is slightly harder to implement but I can still do it if that's the concensus.

Well,HG, Even the rarest of situations have to be taken into consideration.I'll wait to see what Greg has out of all of that!

Yes, that was what I also said to Greg. There exist various kinds of Imitators. But the practical difference is very slight, as delivering check makes any difference for castling out of check (while normally you would castle before the opponent has any opportunity to check), and deciding whether a mate is checkmate or stalemate (while stalemate is usually only possible when the opponent's King is bare, in which case you will have many ways to checkmate that are not affected by the precice Imitator properties). So it is really a moot point.

Me too like Aurelian, I was not understanding the controversy. There is an old entry for the Joker in the Piececlopedia: https://www.chessvariants.com/piececlopedia.dir/joker.html where the Joker is defined relatively to the last move of the opponent. Which is Aurelian's definition I believe.

Beside this, is that discussion at the right page? It looks like if ChessV page is hijacked.

• "Check threat": after making one's move, an opponent's fool becomes the same piece for one move; if the fool would thus give immediate check, the move would have been illegal, thereby preventing the piece moving ("neutralizing" it) unless it moves to block the check or take the fool.

That is from the wikipedia article about Omega chess.And the following is from my article on Apothecary Chess Modern:

The joker (jester,fool) imitates the power movement and capture of the last piece moved by the opponent.

In both cases the joker's move is update at the begining of one's move replacing the old move that has been inherited two plies ago.

So HG, I think we are talking about differnet pieces. 

Isn't this the same situation with having the king and joker unobstructed on the same orthogonal? This pins all rooks, queens, chancellors and so on.

No, this is different: moving a pinned piece is something the opponent punishes in his own turn, by capturing the King. The question here is whether the side that moved has delivered check with this move. Even when he does, he cannot capture the King. Because it is not his turn.

Anyway, the Diagram considers this stalemate, because the Imitator moves like a Knight. This is how I defined the I atom in XBetza:

A new atom in XBetza is I. Unlike the other atoms, it does not represent a fixed move: it stands for "all moves of the most recently moved (or imitated) piece type".

The most-recently moved piece is a Knight here. It doesn't say anywhere that the piece has to be an opponent. In your own turn it would of course be one. So this doesn't affect how the Joker moves. Just how it delivers check (which again is only important in the case of mate or castling).

I think I had not understood this earlier. Now I see the conundrum.

This is a checkmate because the joker still has black queen's powers. Only after black moves the power is changed. In this case it cannot. So it is at least a stalemate. But for the reason stated above the black king is also in check.

Isn't this the same situation with having the king and joker unobstructed on the same orthogonal? This pins all rooks, queens, chancellors and so on.

@Greg,

I see why you hate the joker! And probably for similar reasons I love it!

We can summarize this as:

• To judge whether in check at the start of your turn, the enemy Joker should be imagined to still have the move of the previous turn.
• To judge whether you are in check during or at the end of a turn, you should imagine the Joker to move as the piece that is moving in that turn.

Now Greg's example was special, because the Joker itself moved. If the white King had been on c2, the Joker on a5, and a white Knight on a4, and on the move before Queens were traded at c3, where the Knight recaptured (Na4xc3)... Is this checkmate or stalemate?

The Diagram considers it stalemate, because it lets the white Joker move as Knight after 1... Qxc3 2. Nxc3.

@Greg,

The way I see the situation at hand (the actual situation that has started this discussion), the black king will be captured at the next move and this counts as a check to me, even if we are very technical about it, it is not. I'm afraid I have no further understanding on the matter

@Daniel,

This is not my piece either. I'm just using it. The joker has no movement at the beginning of the game.

I know it's not my piece, but I would think that the Joker does check the king, since the king was the previously moved piece. It seems weird that the Joker could ever have no movement at all.

Well, I think he meant that during your own turn enemy Jokers have no move (so that you can never move into their check), but that whether they deliver check should be judged in the opponent's turn (after he released the piece, but before he pressed the clock, as it were).

This does not make sense to me.  It sounds good but does not really clarify anything for me.  It might be helpful to discuss an actual situation.  In this position, black has just moved his King into the corner:

So now white's Joker moves as a King and moves to b2:

Is black's King in check or not?  If not, then that's a stalemate.  I do not think saying "whether they deliver check should be judged in the opponent's turn (after he released the piece, but before he pressed the clock, as it were)" changes anything.  Any move the black King makes, it will then be judged that he is in check, so that move is illegal.  Any move the black King makes is illegal, so the game is over.  Whether it is checkmate or stalemate is determined by whether the King is in check.  Am I missing something?

Well, I think he meant that during your own turn enemy Jokers have no move (so that you can never move into their check), but that whether they deliver check should be judged in the opponent's turn (after he released the piece, but before he pressed the clock, as it were).