Can you watch Fluidity please, I can’t add new variants but I have 5+ of them in keeping ready.

So how can I add my piece to Piececlopedia?! If I can;)))

Here are the rules of Chess66 as I've written them up for Game Courier. Apart from a change to the naming convention for the new spaces, is this complete and fully accurate? Note how I've divided this into five sections. These are the geometry of the board, the difference that Switches make, how spaces in a Switch count as separate spaces, how spaces in a Switch are sometimes treated as the same space, and how the individual pieces move.

Rules of Chess66

Chess66 is an adaptation of Chess to a board with a different geometry. Where no difference from Chess is noted, it follows the rules of Chess.

1. Its board has two extra spaces that each overlap with a neighboring space and push the rank it is on over by half a space. A4 overlaps with a4, pushing the 4th rank half a space right. H5 overlaps with h5, pushing the 5th rank half a space left.
2. Because its ranks and files no longer line up neatly on a Cartesian grid, movement is based only on geometrical relations and not on the coordinate system.
3. Lateral movement is allowed through the sides of spaces, and lateral movement in the same direction goes through the opposites sides of spaces. Because of this, vertical movement may sometimes shift the file named by the coordinate system.
4. Diagonal movement is allowed through the corners of spaces that do not share any sides, and diagonal movement in the same direction goes through the corners at opposite ends of spaces.

A Switch is a pair of overlapping spaces that come together at one end and branch off at the other. In doing so, they allow some new movement options:

1. At its narrow end, each space in a Switch shares the same side with an adjacent space. This allows a Rook or Queen to move through the narrow end to one of two different files. In this way, a Rook or Queen can checkmate a King without assistance from another piece.
2. From the broad end, a Rook or Queen can move through the Switch to the same file, providing a new way for double attacks to work.
3. Also at the narrow end, each space in a Switch shares a corner with another adjacent space that neither one shares any sides with. Through this corner, diagonal movement is able to change color. So, a Bishop that moves to A4 or to H5 may move away on a subsequent turn on a different color. This allows Bishops to switch color, giving a Bishop the power to reach every space on the board.

Although the two spaces in a Switch overlap, they count as separate spaces.

1. A piece moving to a Switch must move to one space or the other. For example, a Pawn on a2 can make a double move to either A4 or a4.
2. While some paths can lead to either Switch, others lead to only one space or the other in a Switch. For example, a Bishop on e8 can go to A4 or h5, and a Bishop on d1 can go to a4 or H5.
3. The movement options available to a piece on a Switch depend upon which space of the Switch it is on. For example, a Pawn on A4 can proceed only to a5, and a Pawn on a5 can proceed only to b5. Also, a Bishop on A4 can move away on either light or dark spaces, but one on a4 can move away only on light spaces.

Because the two spaces in a Switch overlap, they are treated as the same space in some respects:

1. The two spaces in a Switch cannot be occupied simultaenously.
2. It is illegal to pass through a space in a Switch if the other space is occupied.
3. A piece may move to a space in a Switch only if both spaces in it are empty or one is occupied by an enemy piece. If a piece moves to the empty space in a Switch, and the other space is occupied by an enemy piece, that piece is considered captured.
4. It is not legal to move from one space in a Switch to the other. Instead, A4 is treated as though it were horizontally adjacent to b4, and H5 is treated as though it were horizontally adjacent to g5. This affects the movement options for Rooks, Queens, Kings, and Knights

The King leaps one space in any lateral or diagonal direction. It may castle with a Rook on its first move so long as it is not in check, there is nothing in between it and the Rook, it doesn't pass through check while castling, and the Rook hasn't moved. In castling, it moves two spaces toward the Rook, and the Rook moves to the space the King passed over.

The Queen may move as a Rook or a Bishop.

The Rook may move any number of spaces in any lateral direction until it reaches an occupied space or Switch.

The Knight can leap directly to any space that could be reached in two one-space moves except for those reachable by two in the same direction. Since a piece on A4 or H5 can move directly to b4 or g5, a Knight on one of these can leap further away than a Knight can normally leap.

A Bishop may move any number of spaces in any diagonal direction until it reaches an occupied space or Switch.

The Pawn may move one space straight forward without capturing, or it may move one space diagonally forward to capture. On its first move, it may move two spaces forward without capturing so long as it isn't blocked. If this move takes it over a space an enemy Pawn could have captured it on, that Pawn may immediately capture it by en passant, moving to the space it passed over. On reaching the last rank, it must promote to another piece. This may be any piece except a King or another Pawn.

This has been published. I have edited the text to improve the grammar and clarity. It could still be better, but is acceptable. Thank you for your patience.

@Greg,

I'd like to make chessV play custom variants against another chessV copy where the small parameter would yield different results. May you write a small guide on how to do that?

The definition of the Knight move is not correct. On an undistorted board it would allow the Knight to also move like Wazir or Ferz. This could be fixed by also excluding moves that are reachable in a single step.

Unfortunately, there is no way to do this at present. ChessV can be used to test changes to its parameters by playing against another engine, but as far as I know there are no other engines capable of playing your games. You certainly could run two instances of ChessV but they won't talk to each other.

I need to separate the game representation used in thinking from the one in the GUI. This would not only allow testing of different parameters but would also make multi-threaded thinking possible.

The description of the King's Flight is not clear on whether the King can leap over another pieces. It cannot, based on the Zillions ZRF also by Sidney LeVasseur. I will update this page to clarify.

Let me understand this correctly: A bishop on e8 moves in the direction of switch A4/a4. Further down in your description you say that in this case the bishop can only reach A4.

Correct.

But the bishop reaches the same corner of A4 and a4; therefore the bishop can decide whether to stop on A4 or a4.

No, it cannot. Let me break the move down into steps. First, the Bishop on e8 moves to d7. At this point, movement in the same direction must go through the opposite corner is just passed through. This is the corner that d7 shares with c6. Continuing in the same direction, it can go to b5. At b5, it can move diagonally to A4, because b5 and A4 share a corner but no sides. Also, this is in the same direction that the move from e8 to b5 was in. However, b5 and a4 do share a side, which means that movement from one to the other is not diagonal. So, a move from e8 to a4 would be illegal.

So, in principle, a move into the switch, no matter from which direction and valid for all pieces, must result in a decision between the two squares of the switch - provided that the switch is not occupied.

No, that is completely the opposite of how I understand the rules.

A bishop on e8 or d1 can move to A4 or a4 respectively to h5 or H5.

This is unintelligible. If you mean what you said above, it is contrary to my understanding of the rules.

Also, a Bishop on A4 can move away on either light or dark spaces, but one on a4 can move away only on light spaces.

But a bishop on A4 cannot move to f8. For that he would have to be on a4.

That is correct, and I didn't say anything to the contrary. The light path from A4 goes through b3, c2, and d1.

3. ... or one is occupied by an enemy piece. If a piece moves to the empty space in a Switch, and the other space is occupied by an enemy piece, that piece is considered captured.

A piece can only be captured on the square it stands. Therefore, in an occupied switch, you cannot move to the empty square and capture the piece on the other square of the switch. After the move into the occupied switch, the capturing piece stands on the square of the captured piece.

Are you saying that a piece cannot capture a piece in a Switch unless it can move to the space the piece is on? Or are you saying that when a piece can move to either space in a Switch, it can move to the Switch and capture the piece, and then it must occupy the space the piece was on?

The Knight can leap directly to any space that could be reached in two one-space moves except for those reachable by two in the same direction.

I am not sure if the rule is correct. In my description I use a definition which comes from Alfred Pfeiffer (Chemnitzer Schachverband e.V.):

• The knight moves to one of the squares that a king can reach from the square in two moves, but which are not on the same row, line or diagonal. It does not move across squares that lie in between.

They are the same, but I removed the ambiguities in his description and used some technical language. Since the King cannot move into check, and a Knight is not subject to the same restriction, I replaced the reference to two King moves with one to two one-space moves. Since I would normally refer to ranks and files rather than to rows and lines, and since I have been using these words in an algebraic sense rather than a geometric sense, I rewrote it to not use them. To "leap directly" is technical terminology that implies that a move "does not move across squares that lie in between."

Email verification not working (I have checked the spam folder).

No, it cannot. Let me break the move down into steps. First, the Bishop on e8 moves to d7. At this point, movement in the same direction must go through the opposite corner is just passed through. This is the corner that d7 shares with c6. Continuing in the same direction, it can go to b5. At b5, it can move diagonally to A4, because b5 and A4 share a corner but no sides. Also, this is in the same direction that the move from e8 to b5 was in. However, b5 and a4 do share a side, which means that movement from one to the other is not diagonal. So, a move from e8 to a4 would be illegal.

This is a very special consideration. Do you think that such can be communicated to potential players?

Why do we leave out consideration of equal/different corners and sides? Can't we agree that switches can be entered from all sides and a choice must be made between the fields of the switch? Surely that would be much easier for everyone.

In the discussion at the time, I just took your position, but ran into a wall looking for a simple solution. I came to the conclusion that switches have a special status, which must lead to a choice between the fields of a switch from all sides - from above, from below and from the side.

No, that is completely the opposite of how I understand the rules.

A bishop on e8 or d1 can move to A4 or a4 respectively to h5 or H5.

This is unintelligible. If you mean what you said above, it is contrary to my understanding of the rules.

What can I say? I think my position is quite logical. At the beginning of the discussions, I was of the opinion that switches work differently when they are operated from below, from the side, or from above. I have abandoned this opinion and changed it in favor of a pragmatic solution, in that a switch must be handled the same regardless of the direction.

Are you saying that a piece cannot capture a piece in a Switch unless it can move to the space the piece is on? Or are you saying that when a piece can move to either space in a Switch, it can move to the Switch and capture the piece, and then it must occupy the space the piece was on?

This seems to me to be much simpler than you make it out to be.
After all, a piece can only be captured where it is. Why should this be different for a switch?
Why should a piece be able to move to A4, and thereby capture a piece on a4 quasi en passent? That only happens with pawns. But that's where it should stay. The basic principle should be that pieces are captured on the square on which the pieces were placed. Pragmatic solution, isn't it?

They are the same, but I removed the ambiguities in his description and used some technical language. Since the King cannot move into check, and a Knight is not subject to the same restriction, I replaced the reference to two King moves with one to two one-space moves. Since I would normally refer to ranks and files rather than to rows and lines, and since I have been using these words in an algebraic sense rather than a geometric sense, I rewrote it to not use them. To "leap directly" is technical terminology that implies that a move "does not move across squares that lie in between."

I think I have understood that.

This is an interesting game I'd like to implement, but the following is problematic:

4.  As an exception to rule 1 above, you may capture a defended Golem or Half-Golem two squares away with a Golem or Half-Golem if it is the only legal move you have.

I don't have the ability to retroactively rule an otherwise illegal move legal based on the outcome of every other move.  I doubt any chess engine is going to do this.  There is a ZRF, and I've looked at it and don't believe it does this either, but I can't be 100% sure.  I have changed computers several times since I last had Zillions installed.

...because paths to and from Switch spaces are no longer symmetrical with each other. This would allow one King to attack another, and the following position would count as checkmate.

I can't see anything other than a checkmate situation. The legal moves of the black king are vertically a5 and a3, horizontally b4 and diagonally b5 and b3. The move to c5 is excluded and only with that can the black king free itself.

We had such a situation in the discussion below and it involved a bishop on A4.

Also, a Bishop on A4 can move away on either light or dark spaces, but one on a4 can move away only on light spaces.

But a bishop on A4 cannot move to f8. For that he would have to be on a4.

That is correct, and I didn't say anything to the contrary. The light path from A4 goes through b3, c2, and d1.

If the bishop on A4 cannot move towards f8, then the king on A4 cannot move to c5? On the other hand, the white king can capture the black king on A4. This is undoubtedly an asymmetry that, as I understand it, cannot be avoided when handling a switch.

I guess I didn't understand your question correctly. I'll try my best to answer.

The switch is occupied, so the black pawn can move diagonally into the switch. In my opinion, the black pawn does not have a choice between A4 and a4, but moves directly to A4 and captures the white pawn there.

This is a (still imperfect) attempt with the Interactive Diagram. Even for this I needed to adapt the Diagram script: although the new move-entry system does allow promotion to enemy (as needed when capturing a Golem), the AI did evaluate it as if the capturing piece changed into a Half-Golem of its own color. The type changes are dictated by a user-supplied routine WeirdPromotion; this was the easy part.

Of course the AI assigns the same value to Golem and Half-Golem, as it uses mobility as only guidance. Having it upgrade the value because of a user-specified demotion of pieces that capture it seems asking for too much. That raises the question how to handle this. Perhaps the Diagram should have a parameter value that can be used to overrule the automatic piece-value assignment.

The Golem and Half-Golem are subject to the Diagram's anti-trading features of the protected and counterStrike kind. These also forbid trading of adjacent Golems, though. Actually I think it is a flaw that Golem Chess allows this. For a Lion an adjacent piece would be captured through hit-and-run, and protection would be ineffective and not lead to trading. This is why the Chu-Shogi Lion-trading rules are as they are. For the Golems it still leaves an easy way to trade those, though. Of course I could have the Diagrams protected feature exempt adjacent capture. That would even give a better representation of the Lion. But I intentionally did not do that to make it more useful for pieces that cannot rifle-capture...

And no special exemption for checkmate yet... In the Diagram this is more difficult, because it only rejects the illegal captures after search, in the daughter nodes. So it becomes more difficult to communicate to the parent the move was rejected for this reason, so it can decide to redo those.

[Edit] Hand-tuning of piece values through a value=N line with N in centi-Pawn immediately after the line with a piece definition is now generally supported. (And used here to set the Golem value to about double that of what its mobility would suggest. That seems reasonable; it is like having two Half-Golems on the same square.)

I took the discussion from earlier to heart and to that extent I thought about it and adjusted the rules. The result:

I find the idea that a bishop on d1 can only move to a4 or to H5, or a bishop on e8 can only move to A4 or h5, to be completely illogical.

You can bend it with corners and sides so that it seems logical, but basically, from my point of view, it leads to complicating the rules.

The rule that a switch is a place of decision to occupy the squares of a switch uniquely, no matter from which direction the switch is occupied, seems to be a clear rule that everyone can understand. I do not think that such a rule complicates the course of the game.

The concept implies that unsymmetrical move sequences cannot be excluded. I cannot express it in any other way. 

This diagram isn't working properly. Capturing a golem is demoting it to a half-golem, but of the wrong side. The half-golem should be the same side as the golem was.