Comments/Ratings for a Single Item

Actually, I'll stick with the current graphics sets for the new Chu preset and add a Mnemonic set based on your Mnemonic pieces that matches the size of the Chu kanji images. I will still make a new preset for the game though.

Speaking of which, I corrected a few mistakes in the alfchushogi set file. Your welcome to the author of said file.

Edit: I've added the Mnemonic set, which is based on LiShogi's set.

The fairychess-style code I was going for was giving me serious grief, so I decided to instead try to implement the Lion-trading rule display in this preset instead. No luck so far, though its plausible I am just being an idiot. For now, I will leave this preset as the default one, and leave it in its current state, which I know works properly. It might not be ideal, but it is good enough. Sometimes I get so wrapped up in whether I could, that I don't stop to think whether I should.

Question. In the position below (from LiShogi), Black has just captured White's Lion. The recapture of Black's Lion via White's Soaring Eagle via a Lion move, which is allowed in LiShogi, clearly violates Chu's counter-strike rule, right?

If this is true, then both the Interactive Diagram and the Jocly implementation overlook this restriction when a Lion is captured on the first step of a multi-move.

The GC preset enforces the rule in such situations, but this is not reflected in its display of legal moves.

P.S. I have already notified LiShogi of this via Discord.

Indeed, the counterstrike should be forbidden here.

The Interactive Diagram does implement the counterstrike rule: when you place both Lions in front of an enemy GB, switch on the AI, and play GBxLn, it will not capture the Lion. In the highlighting you won't see it, though, as the I.D. does pseudo-legal highlighting. Moves that expose you to check will also be highlighted. (If it is the King making such moves these are highlighted in grey, but betzaNew.js doesn't even do that anymore. Perhaps one of these days I will subject the highlighted moves to a legality test, and use another highlight symbol for the illegal ones.)

The same test appears to work in Jocly, though. And then it doesn't even highlight the counterstrike GBxLn.

Indeed, it implements the counter-strike rule for single captures. However, the Soaring Eagle and Horned Falcon can move again after stepping in certain directions.

Upon further testing, it seems that the counter-strike rule is not being invoked when a Lion is captured and then the capturer moves away. For example, in this game in the I.D., after promoting the Dragons on h9 and e4, if Black's Soaring Eagle captures White's Lion and then moves to another square afterward, the counter-strike rule will not be invoked and the AI will capture Black's Lion with White's Phoenix.

1. h5 e8 2. Qh4 Qe9 3. Qe7 Qh6 4. Qxe8 Qxh5 5. Qxf9 Qxg4 6. Dd2 Di11 7. Ti2 Td11 8. Oh2 Oe11 9. Tf2 Tg11 10. e5 h8 11. Xe4 Xh9 12. Xg6 Xf7 13. Xxi8 Xxd5 14. Qf5 Qg8 15. Dxh8 Dxe5

The moves you need to make manually are:

16. Dh9+ De4+ 17. +Dxg10xh9

If you make the same moves in Jocly, the same thing will happen.

The same thing happens in LiShogi, since the same behavior is taking place. I have already notified them about what I have found. One of the developers apparently thinks non-lion refers to non-Lion moves though, rather than non-Lion pieces.

Indeed, it seems that locust capture of the Lion does not set the 'iron lion' flag. Counterstrike through locust capture is forbidden, though.

Good catch!

[Edit] In the I.D. the AI does this correctly. But the problem is caused by the way the latest user move is passed to the AI. This derives the move from the origin and destination highlights, so it completely misses any locust capture.

If you make the same moves in Jocly, the same thing will happen, just so you know.

One of the LiShogi developers responded to my bug notificaton, but apparently thinks non-lion refers to non-Lion moves though, rather than non-Lion pieces, which leaves me wondering what the general consensus is. I would assume it to be piece types, based on the explanation of the another square clause.

but apparently thinks non-lion refers to non-Lion moves though, rather than non-Lion pieces, ...

I am not even sure what that would mean. Of course the consensus is that non-Lion refers to a piece type, and until now I have never seen any other interpretation. The 81-Dojo server (when it was still featuring Chu Shogi) interpreted it as types. No phrase that would not explicitly contain the word 'move' or 'capture' would ever refer to a move. Do there really exist people that seriously would want to argue that when the rules specify "the Queen is not not allowed to capture Pawn" that this means Rooks, Bishops, Kings and other Pawns are not allowed to capture a Pawn either, because they would do it through a Queen move? This is insane.

[Edit] In the I.D. the AI does this correctly. But the problem is caused by the way the latest user move is passed to the AI. This derives the move from the origin and destination highlights, so it completely misses any locust capture.

I did test the position I gave you by having the AI capture the Lions instead of the user, but this too seems to not properly invoke the counter-strike rule, though I suppose you still have to update the source file. I'm guessing the problem is that the way the AI gets the last move in general is through the origin and destination squares.

I can see the way you interpret "non-lion" in the counter-strike rule becoming a point of contention. The rule does not explicitly say "non-Lion piece" when it says "a non-Lion cannot capture a Lion...", which is the source of this confusion.

I can see the way you interpret "non-lion" in the counter-strike rule becoming a point of contention though.

I cannot see that, and never have spoken to anyone who thinks that. (Which includes the Chu-Shogi Renmei, the 81-Dojo community, the German Chu-Shogi Association, the TSA, Hidetchi, the Wikipedia maintainers...) If it would have said "A Lion cannot capture a protected Lion"... Would that also mean that a Rook cannot capture an adjacent protected Lion, because 'Lion' refers to 'Lion move', as it did not explicitly contain the word 'piece'? And if 'Lion' by default refers to a piece, why would 'non-Lion' be different?

I read in LiChess 'issues' (in particular about Ln x P x Ln without recapture possibility) that the developers want to follow CSRM rules, even though most people consider some of the deviations from historic rules there non-sensical. And the CSRM definitely considers 'Lion' a piece type, and not a move.

Now that you say that, it clears up all the previous confusion. I guess my autism got in the way.

I have cleaned up the code for the Chu Shogi preset, and have made a separate preset for testing the Lion-trading rules for the legal moves display.

[Edit]: Good news! The GC Preset now successfully displays the counter-strike rule. Still working on the other Lion-trading rule though.

It's official. The Lion-trading rules are now properly displayed in the GC preset.

[Edit] I've made modifications to the Chu Shogi GC set files for the Japanese and Latin Chu Shogi pieces so that the size of the highlights matched the size of the piece images, allowing for more user-friendly legal move displays.

It seems to me that capturing a protected lion by a lion is an available move as long as the diagram understands. Anyway I won't do it and I imagine the AI does not do it either.

The AI understands the Lion-trading rules, but this is not reflected in the highlights when the player clicks on a Lion.

I made some paper cutouts of Chu Shogi using the Mnemonic pieces and a checkered board for learnability purposes, complete with the rules.

The pieces are designed so that if you only use paper, folding them as instructed will make the pictures face the opponent, like real Shogi pieces.

@H. G. Muller,

It's a shame that the position isn't given for tsumeshogi B23 in your MSM Errata B page. It sounds like it would be a fascinating problem, especially since it involves a Kirin deferral, but all that is given are tiny snippets of the solution. If given the FEN or a picture of the position, I can have HaChu solve it like I did with B15, and then compile the problem in the Chu Shogi Applet I made, which compiles all the tsumeshogis from your MSM Errata pages that give the position and have known solutions under the historic rules.

I've made 2 tsumeshogis involving Kirin deferrals (series MD in the Applet's tsumeshogi folder), but they are embarrassingly simplistic. Having B23's position would eliminate the need for the MD series, as it would showcase a much more elegant example of Kirin deferral.

I did not fully discuss the problem because the solution given for B23 was essentially correct. I only mention it because it failed to point out the essential need for deferring promotion to Lion, and I did not think the discussion in the MSM did that fact justice.

At that time the MSM was my only source for these problems, and it is a copyrighted source (which also contains the FENs). Many of these problems are presented on the website of teh chushogi-renmei, though. The numbers of the B series seem to correspond with the numbers in parentheses of set 1 there.

I took a look at my file with FENs, counting lines. Can this be the one for B23:

4+A1gk2f1/6s2+P1l/1+H1n4r1e1/5N3mp1/5xpO4/11+L/9Q2/12/3+v8/12/12/12 w - 0 1

?

[Edit] Indeed it is. HaChu cannot solve it, though, as it does not consider Kirin deferral. You have to first play the Kirin move; then it finds mated in 8 within a second.

4+A1gk2f1/6s2+P1l/1+H1n4r1e1/5N3mp1/5xpO4/11+L/9Q2/12/3+v8/12/12/12 w - 0 1

Thanks. Once I find the correct line to in order to reach get the mate in 8, I can compile this problem and delete the MD series (which is embarrassingly simplistic by B23's standards).

At that time the MSM was my only source for these problems, and it is a copyrighted source (which also contains the FENs).

I did make sure to cite my sources for the tsumeshogis (which you can find in the  "0 - Chu Shogi Tsumeshogi Sources" PDF File in the Applet's tsumeshogi folder.) However, since there is a legal precedent that a Chess position cannot be copyrighted, I did not bother asking for permission from anyone before compiling the tsumeshogis, since the ruling, which presumably extends to Chess variants, would effectively consider the positions to be in the public domain.

I also fixed certain problems that needed it.

Oops, I misinterpreted HaChu's output. It is not mate in 8 after playing the Kirin to h10; the score is from black POV then, and meant that white could not mate.

I ran the position throu Hachu, and it wasn't able to find a mate with the FEN you provided, even after the Kirin move. After

1. +Ah9 Mxh9 2. +Pi11 Rxi11 3. +Lh11 Gxh11 4. +Hd12 Nxd12 5. Oh10= Xxh10 6. Qxd12 +Vxd12 (I think this is right, please correct me if I am wrong),

Gote's +VM guards the d12-l4 diagonal preventing a mate from occuring, resulting in

7. Nf11 Ki12 8. Nxg11,g12 Kj12 9. Nxh11,h12 Kk11 10. Nxi11 Kl10 11. Nxk12 Kl9 12. Nxl11,k11 Kk8 13. Nxk10,j10 Kj7 14. Nxk9,l8 Ki6 15. Nk7 Kh5 (keeps the Lion away from the vital SM guarding the 9-rank) 16. Nj7 Kg6 17. Ni8 Kf7 (guards all possible checking squares that would result in mate, with help from the +VM and PH) 18. Nh7 Ke8 19. Nxg8 Kd9 20. Nf7 Kc10 21. Ne8 Kb11, after which no mate is possible.

However, if the +VM on d4 is replaced with a +RC, it works!

4+A1gk2f1/6s2+P1l/1+H1n4r1e1/5N3mp1/5xpO4/11+L/9Q2/12/3+a8/12/12/12 w - 0 1

My guess is there was a copying error when you went to copy the FEN code for B23. The new position results in

1. +Ah9 Mxh9 2. +Pi11 Rxi11 3. +Lh11 Gxh11 4. +Hd12 Nxd12 5. Oh10= Xxh10 6. Qxd12 +Axd12 7. Nf11 Ki12 8. Nxg11,g12 Kj12 9. Nxh11,h12 Kk11 10. Nxi11 Kl10 11. Nxk12 Kl9 12. Nxl11,k11 Kk8 13. Nxk10,j10 Kj7 14. Nxk9,l8 Ki6 15. Nk7 Kh7 16. Ni5 Kh8 17. Ni6 Kg9 18. Nh7 Kf10 19. Nxg8 Ke11 20. Nxh9,g9 Kd10 21. Ne8 Kc11 22. Nc9 Kc12 23. Nc10 mate.

Sorry, I again misinterpreted HaChu's scoring. Which was indeed such that there was no mate in 16, but that the checks could only be kept up for 17 moves. So you are probably right about the Flying Ox/Whale copying mistake. The MSM clearly depicts a Flying Ox in the diagram for this tsume puzzle, though.

Sorry, I again misinterpreted HaChu's scoring. Which was indeed such that there was no mate in 16, but that the checks could only be kept up for 17 moves. So you are probably right about the Flying Ox/Whale copying mistake. The MSM clearly depicts a Flying Ox in the diagram for this tsume puzzle, though.

My guess is that the designer did not anticipate the Flying Ox's diagonal move being problematic, rather than it being a copying error, since the kanji for the Ox and Whale are vastly different. Of course, what's important is that the problem is fixed and solved.

Not  sure if you still update the hgm.nubati site anymore, but feel free to further discuss this problem in the MSM Errata B if you want (If you do, I'd recommend having a picture of the position).

P.S. I've updated the Applet, deleting the MD problems and solutions and adding B23, as well as citing our comments for B23's source (the MSM citation is already covered by the MSM Series B citation).

It seems to me that capturing a protected lion by a lion is an available move as long as the diagram understands. Anyway I won't do it and I imagine the AI does not do it either.

I am working on making display of Lion trading restrictions possible in the diagram as a side project. I recently found a way to get the coordinates and pieces from the last move of a game, which will allow enforcement of the counter-strike rule. Still not sure how to do the bridge-capture rule though...

[Edit] Successfully implemented the counter-strike rule in my prototype.