No.737 by Oliver Sick – Wormhole condition with specific e.p. capture. To my opinion, the logic of this solution is debatable, however it is C+ by Popeye. Still, the animation by Py2Web seem to be impossible for such e.p. capture. I’d appreciate reader’s comments! (JV)
Definitions:
Wormhole (sent by author): Wormhole is property of the squares. If a move of the piece ends on a wormhole, the piece is transferred to any other unoccupied wormhole. A pawn moving to a promotional square, which is a wormhole itself, promotes and the promoted pawn is transferred to another wormhole. If a pawn is transferred to its ground row, it can make a single step or capture. If a pawn makes a double step onto a wormhole, it still can be captured en passant, even if the pawn has been promoted into any figure. Zero moves are not allowed. (translation of the definition from Die Schwalbe)
Circe: Captured units (not Ks) reappear on their game-array squares, of the same colour in the case of pieces, on the file of capture in the case of pawns, and on the promotion square of the file of capture in the case of fairy pieces. If the rebirth square is occupied the capture is normal.
|
No.737 Oliver Sick
Germany
original – 27.02.2015
|
Solution: (click to show/hide)
|
neutral Pc4
black Kf5
white Pd2 Ke1
h#1,5 (2+1+1)
Circe
Wormholes d3, d4, g1, g8
|
Animation is impossible for this kind of e.p. capture
(According to the wormhole rule en passant capture is possible,
even if the captured pawn was transferred and promoted (as in this case the wRg8). (JV)
Solution with detailed author's comments:
1... d2-d4[wPd4->g8=R]
- Wormhole rule: This is a doublestep onto a wormhole, so the white pawn has to be transferred to another wormhole, in this case to g8. Applying the wormhole rules the white pawn then also has to promoted (into a rook in this case)
2. nPc4*d3 ep.[nPd3->g8=nQ][+wRh1]
- Wormhole rule: According to the wormhole rule en passant capture is possible, even if the captured pawn was transfered and promoted (as in this case the wRg8).
- Circe rule: According to the Circe rule the captured wRg8 must be reset to its origin, in this case h1.
- Wormhole rule: According to the wormhole rule the nPd3 has to be transfered from wormole d3 to another wormhole (in this case g8) and the nP must be promoted (in this case into nQ).
2... 0-0[wKg1->d4] #
- Circe rule: According to Circe castling is possible even if the wR is reborn to h1.
- Wormhole rule: Acording to the Wormhole rule the wK has to be transfered from the wormhole g1 to another and unoccupied square (in this case d4)
And finally it is mate!
The en passant rule for wormholes ("en passant capture is possible even if the captured pawn is tranfered and even if it is promoted into a rook, queen, bishop or knight..) is quite weird, but it is "common sense" (and so defined in Schwalbe and other tournaments) and the way it is implemented in popeye.
{1... d2-d4[wPd4->g8=R]
2. nPc4*d3 ep.[nPd3->g8=nQ][+wRh1]
0-0[wKg1->d4] #
(C+ by Popeye 4.69, no solutions by WinChloe)}
Summary comments:
- 1st halfmove: promotion and doublestep
- 2nd halfmove: en passant plus another promotion
- 3rd halfmove: castling and mate
=> Valladao with four pieces in 1.5 moves. I assume this is minimal at least in time. (Author)
|
No.737 
Can somebody from Popeye team answer to the following questions:
1. Why the neutral pawn played by Black on the 8th rank is promoted into a neutral Queen? I think neutral pawns played by Black can promote still only on the 1st rank.
2. Why the short castling is allowed, although the [intermediate] square g1 is under the observation of the neutral Queen?
Please also let us know if you consider the above mentioned behaviors as features or bugs. Thank you very much!
Hello Vlaicu,
I am not part of the popeye team. But I see the following :
On 1.:
The definition of neutral pawns is that they promote on the 1st or 8th rank. The rules for neutral pawns I know (from the net or chess magazines) say nothing about the party who is promoting the pawn (=> promotion on g8 is allowed ?!?) Popeye 4.69 uses this definition consequently (eg. also for supercirce…)
On 2.:
The square g1 is attacked in deed, but only when a moves ends on it. In this case the move ends on d4. But in fact I also feel uncomfortable with this, because of the following observation in Popeye 4.69:
Castling is now allowed in the following situation
Condition Wormholes d4f1
Pieces
white Ke1 Rh1
black Qf8 Ke6
but is allowed here
Condition Wormholes d4g1
Pieces
white Ke1 Rh1
black Qg8 Ke6
This feels ambigous (what is the difference between f1 and g1?!?). Of course to avoid this ambiguity I could add a bPg4, but I’ve looked for a minimal….
(I’m not in the Popeye team, so these are just my personal thoughts.)
Q1. The Q promotion on g8 looks to me like a bug. It is indeed Black who plays the P to his own first row, which by the wormhole rules should imply that it stays a P, being able to make a single-step or a capture from there.
(But on the other hand: How are neutral Ps in other fairy forms – PWC, Einstein or some other – treated when they appear on row 1 or 8? You could argue that on the following half-move, the nP is on the promotion rank of that side, and why then is it not promoted?)
Q2. This seems to me like a feature! g1 is not the intermediate square in the castling (that would be f1). A king can move to a guarded (observed) wormhole if he is then transported to another, unguarded wormhole. That should be legal also when the king moves onto the wormhole by castling.
(The follow-up question is harder; what if f1 is a wormhole, is castling legal or illegal? Does it depend on whether there is another unguarded wormhole or not (i.e., whether a direct K move to f1 would be legal or not)?
I would say that castling is illegal.
FIDE rules:
Castling is prevented temporarily if the square on which the king stands, or the square which it must cross, or the square which it is to occupy, is attacked by one or more of the opponent’s pieces
So, white king is not allowed to visit g1 during castling.
But the king is not going to occupy g1 … but d4!
Such explanation is, of course, possible. It is then necessary to consider the fact that the pawn did not move on d4 , but g8 – is then e.p. possible?
Kjell, I agree with you the king is not going to occupy g1. But does it cross that square during the castling?
There is a bug in Popeye whereby neutral pawns can promote on their own first rank. The bug was introduced in version 4.63. For more details see:
http://sourceforge.net/p/popeye-chess/bugs/
156 Neutral Pawns can promote on home rank
What should happen to a neutral pawn brought to the 1st or 8th row is at least debatable. After all, a neutral pawn is both white and black at the same time, so “their own first rank” is not a clearly defined concept.
One option is to let the role of the nP when it gets to the first or last rank decide what happens: if a nP is treated as white when it gets to row 8 (i.e., played by White or captured by Black), then it promotes; if it is treated as black when it gets to row 8, then it stays as a nP – mobile for Black but immobile for White.
But another option is to simply say that both row 1 and row 8 are promotion ranks for neutral pawns (as they are both black and white), no matter how the nP gets there.
As far as I know, until Popeye v4.63, the first option described above was the only option considered by both programmers and chess problem composers (there are tens of such compositions). This is aligned with T.R.D’s view on neutrals: “neutral pieces […] may be played by either player – moved if he regards them as his own, or captured if he regards them as adverse” (Caissa’s Wild Roses). When moved or captured they are regarded as ‘either’ white or black but not as ‘both’ white and black at the same time – in the end, they are ‘neutrals’ not ‘grey’. So, if neutral pawns are regarded as white or, respectively, black, at the beginning of the move, they should not be regarded as black or, respectively, white, at the end of the move. I personalty find the fact that the behaviour was changed starting with Popeye v4.63 (ironically, when Wormholes was implemented) disregarding definition, precedent and without being formally announced, as unfortunate.
WinChloe seems to disallow the possibility, for a nP moved by the white side, to reach the first row.
As an example with nPc4 bBd5 in T&M, cxd5 with make move to a8 or g8 is legal, but make move to h1 is illegal.
Maybe such a simple and logic rule should be kept without T&M, but instead with a couple of wormholes (d5,h1). In that case maybe the capture cxd5 should be declared illegal.
With wKe1 wRh1 and wormholes on f1 and d4, Popeye allows 0-0 but without the additional Rf1->d4. If we put the wormhole f1 on g1, Popeye still allows 0-0 but with the additional Kg1->d4.
To my eyes the underlying logic is correct here, if we remember that 0-0 is only a King move. In particular the involved Rook doesn’t “move” to f1 during the castling procedure, it is “transferred” to this square – a kind or rebirth.