Old Castle Chess

Cavalry Shogi

I modified the games. Please check them.

Ok, but will these make a game more entertaining?

Well, individual pieces that can do something only N times is currently already possible: you define different piece types for each number of attempts they have left, and let those demote to the type that has one fewer attempt. The demotion can be specified by the morph parameter, and their permanent move can be exempted from it by marking it with an apostrophe in the XBetza description. Duplicating the piece type that way is not very elegant, though.

What is a problem is a 'global' budget, shared by all pieces (of the same, or a number of types). Then, if one piece uses the move, all other pieces that are capable of using it would have to be demoted. But implementing it that way would be quite inefficient.

I suppose a generic feature that is efficient could be this: a counter for each player is added to the game state, which by default starts at 0, but can be made to start at another value by a parameter counter=N. Moves in the XBetza description could be marked (e.g. with a t modifier); such a move would then only be allowed if the counter for that player is non-zero, and when such a move is performed it decrements the counter.

Yes, your no pieces idea is nice!

Sounds like some sort of limitation like "can only use  # times," or perhaps "only when there are no enemy pieces within # spaces," could be of help, from the perspective Aurelian proposes.

The former sounds to me like a possible use for t in XBetza; for example, t2mNN would allow two non-capturing Nightrider moves per game. I don't know how I'd code the latter, but I do know that I've come across several pieces with that type of limitation (and its inverse).

HG,

I'm not sure if the 0-1-infinity applies here. Having two aces in the pocket could be useful. The way I see it if you give full hippogonal directions to some of the pieces at least, the second move could be useful for a mid game relocation if the region where the first special move occurred got cleaned out. 3 will overdo it, most likely. But think about a phoenix with 2 just moving nightrider powers or a Kyrin with 2 just moving camel rider powers (in order to preserve the coloubounding). They could be useless after their first crash into the enemy forces. But if there is a second relocation move they can stay relevant. I think the downside of this is that such a piece may return homw making the game two defensive.

An even crazier suggestion would be to have an overall budget of relocation moves for all short leapers. Food for thought.

Indeed, the most-forward 4 moves of the Nightrider, as non-capture. Allowing it a single time seems cleaner that twice (0 - 1 - infinity principle).

Perhaps it is indeed no problem, as you say. It certainly would simplify the rules, which is a good thing.

@HG.

What is fhmNN? If I had understood correctly is the four foreword most directions of the nightrider.

I don't see a problem with saving the move for endgame. This is another strategical choice which can make the game better. I 'm thinking that even allowing a 2 times move should be fine. But yes, not too much.

Another way you could do it is with mandatory demotion upon capturing, or on entering some region of the board.

Lunapawn

Lunapawn is able to use a light magic. This is a low cost and 2 Materials. And stronger than a pawn.

I just recently developed an Inquisitor piece, moving BDC.

(I just thought I'd mention it, in case it matters to you.)

Other Variation 2

Jinmaku

Jinmon

Material Cost

Transformed Side Other Variation

Captain

This is a non-royal piece but moves as a King.

Cavalry transforms to Bishop(b1 and h9) or Captain(h1 and b9).

My Original New Shogi

Transformed Side

King

Soldier

General

Admiral

Cavalry

Knight

Bishop

Inquisitor

Barrier

Gate

Rook

Castle

Material Cost

Well, what sucked was the fact that for shortage of piece images I had to assign some images with a strong association to a move to a completely different piece. Such as the Camel and Knight that you mention.

And indeed, Jocly slices the desired part out of a file that contains all pieces. The instructions for which area to take out of the file are in the game's *-view.js file. It appears that I indeed did not backport Chu Shogi to source code yet, but in the library file http://hgm.nubati.net/jocly/jocly-master/dist/browser/games/chessbase/chu-chess-view.js you can see how the association is made by searching for 'sprites' on that page. There you find something like (reformatted for the purpose of showing here):

View.Game.cbFairyPieceStyle=function(e){
return $.extend(!0, {
      1: {default:{"2d":{clipy:0}}},
    -1": {default:{"2d":{clipy:100}}},
  default: {
        "3d":{display:this.cbDisplayPieceFn(this.cbFairyPieceStyle3D)},
        "2d":{file:this.mViewOptions.fullPath+"/res/fairy/wikipedia-fairy-sprites.png",
              clipwidth:100,
              clipheight:100
             }
  },
"sh-pawn":{"2d":{clipx:0}},
"sh-shuttle":{"2d":{clipx:2700}},
"sh-copper":{"2d":{clipx:2400}},
"sh-silver":{"2d":{clipx:900}},

The clipwidth, clipheight, clipy and clipx parameters specified there determine which part of the image should be used. Apparently the square size is 100 x 100 pixel here.

Rather than replacing some of the pictograms in the file, I would recommend extending the file with some new pictograms on the right, and then modify the code in chu-chess-view.js to use those image for the corresponding pieces by changing their clipx.

Thank you. I'll try to pull the png. The 2d does not suck, it is pretty good. There are some pieces that I just want to adjust, like the existing camel, and knight. I see the camel and expect 1x3, and the knight 2x1 leap. There a no knights in chu, so the computer often makes a move that I don't expect because I am thinking knight. I really dislike kanji tiles because I can't easily identify one from the other, and if you flip it, whoa! :)

Question: the png file seems to contain everything. at CV the PHP code expects a seperate png/jpg file for each piece and there is some file that glues it all together. Do I need to make about 80 png files including flipped images, or is the code so advanced that somehow jocely can dice and slice the png file and make each piece off that one file?

Jocly on github won't include Chu Shogi; it is one of the games I added. I do have a Jocly repository on my own website hgm.nubati.net, which does contain some of the modifications I made. Not all, though, as I originally started by hacking the compiled code, and never finished backporting all that to source code. So I am not sure whether that source would do Chu Shogi.

In any case, Jocly does not contain a 2d piece set especially for Chu. What I remember is this: each game can specify an image file as 'resource', which then contains all the 2d piece images it uses, side by side on a transparent background. For each piece type the game definition contains the number of the piece in that file; apparently the Jocly generic code cuts the piece out of the image, and pasts it on the board it displays.

My Chu implementation uses the standard file with Jocly 2d pieces, ( http://hgm.nubati.net/jocly/jocly-master/dist/browser/games/chessbase/res/fairy/wikipedia-fairy-sprites.png in the Jocly library on my website) which I had extended with a few pieces for other games I implemented. The way to go seems to make an alternative to that file, perhaps by copying and modifying it, and then point the Chu game definition to that file instead of the standard one. If needed by directly editing the Jocly library.

I admit that the current assignment of 2d symbols to the Chu-Shogi pieces sucks. I really only payed attention to creating the 3d kanji tiles. To qualify for incorporating Chu Shogi in the official library I suppose it should use kanji tiles in the 2d representation as well. But I dislike kanji representations, like all westerners other than the miniscule community of non-Japanese Shogi players, so it seemed a good idea to use the 2d representation as an alternative. Personally I would prefer the mnemonic representation for that, but a poll under Chess players revealed they would prefer a pictogram representation, where the symbols are a reminder for the name rather than the move of the piece.

First of all Jocely is awesome. I am playing chu shogi on Jocely, and using the 2d symbols on the board when I play. I'd like to design my own 2d piece set. If I download jocely at git hub, does it contain the existing symbols for 2d chu shogi jocely game? If yes, it will make my life easier, I can just edit each piece one at a time. If I design a new chu shogi piece set, would you be willing to add it to CV?

I found the other comment. Fergus replied on 2023-07-10 related to a bug with Alice chess. Do I need to add the lambda in Copper_General or the "-range" function?

Can I get access to these graphics for editing and adding chessmen?

Hi Kevin

Sorry, I haven't checked my email for a while. At the moment I can't commit to starting new games and I really appreciate your interest in playing with me.

Hi Carlos

Please check your incoming email.

I've sent you 3 emails including on 12 July, 2023 and I'm wondering if you are receiving emails properly.

Regards, Kevin Pacey

Sorry if this is posted in the wrong place, I didn't know where else to post about this. I've recently finished (or at least almost) a chess variant that I've created, (I have had other ideas in the past and some I might figure out in the future) my problem is how to test it? I don't know enough about there various programs (like Zillions of games), I don't know their capabilities, limitations etc (I've only foumd brief info on that) so I'm not sure if my game could be made with any, also I do have a little coding knowledge, but not sure of it's enough. The main issue is whether or not it's worth investing in one or more programs if they're going to be useless for me and I don't think it's a good idea to submit without testing it first. So I feel like I've hit a brick wall.

I mean International Day of Chess (variants)

Congratulations Neill Armstrong!

Congrats! It’s 20th of July! Thank you all for extending the culture of the humanity!

This is mind-blowing, Jean-Louis. Well done.

Hi. Just announcing that I have designed a Metamachy set, available for anyone having access to a 3D printer. Some photographs done with cheap PLA. The squares on the board are 50x50mm2.

They can be found on https://www.thingiverse.com/kazo65/designs

There is little I can do about that. Sometimes it helps to try a few times.

But, HG it says my connection is not secure!

Oh, yes!

It appears your login name is catugocatugocatugo.

HG,

Maybe you can help me!

I was running out of ideas for how to speed up such large games. And for the variants I had in mind (14x14 and 16x16) the ideas I already discussed seemed sufficient for making those playable. A fundamental issue is that it takes at least as many moves as you have pieces to move all your pieces.

I've always liked the concept of the transport pieces in Jetan Jeddara as a way to speed up large games.

My argument isn't about shortening the total number of individual unit moves, but rather about giving all the pieces more opportunities for movement and involvement in the game. Just for starters, every opening pawn move is matched with a piece move. Right there, you get all the turns that are just opening pawn moves to move a piece, also. That gives you a few extra moves right in the beginning of the game. I believe that will force both players to use more of their pieces in a game. And that is doubly true if you allow any 2 pieces to move per player-turn. I do agree that each turn will be longer, and the game overall might go on longer, but that was not part of my considerations. And for what it's worth, it's a lot better than the only other option for "forcing" more use of all the pieces that comes to mind now, which is to just remove the pawns and play without them, which is in some ways very instructive, but does not give you anything close to a game of skill.

I was thinking about pieces that could only be used once to inflict great damage. E.g. like a Fire Demon, but which also disappears itself on capture. In Tenjiku Shogi a Fire Demon is 'priceless', since it can continue to inflict damage for the entire duration of the game, and this will add up to more that it could ever destroy in a single trade. But with a kamikaze piece the damage stays limited, albeit large. If there is no defence against it, you must simply accept that a certain fraction of your army will be destroyed by  this without having been used. But it could be a different fraction, depending on the defensive strategy you employ. After the kamikaze pieces are gone it becomes a normal chess game. And if you wait too long using the kamikaze pieces chances are that you cannot do as much damage with them anymore, because population density went down.

Perhaps a Grasshopper that explodes on capture, destroying itself and all adjacent enemy pieces would be an idea. It has great forking power, so that the only defense against its attack is deciding which pieces you sacrifice to it.

@HG,

The non-capturring moves are non necessarily about shortening the games. Giving the fact that it complicates calculations, the game could be very well actually longer on the clock.

But, maybe having this rule reduces the locality of the turn. Things could happen easier in multiple parts of the board so many more pieces contribute to the decision making.

Your choice of aggressive capturing pieces (probably lion and fire demon) it shortens the games but I don't think it makes it more interesting. It could very well make it barren of strategic concepts!

One should not confuse moves with turns. Allowing multiple moves per turn doesn't shorten the game in terms of moves. You just re-order the moves.

@HG,

I agree with you that very long leapers are bad design because of being very hard to defend against. On the other hand if you give 16-24(I'm more for 16 and maybe accept 20) path mover (lame leaper) moves they become useful. The reason for that is that moves from the opponent in the mid game need to be calibrated so that it won't unblock such pieces. These path movers (lame leapers) in the mid game threaten terrible forks but are quite useless in the endgame providing an interesting strategic choice for the players. I like the strategic choices that are offered by unblocking paths (this is why I use all sort of bent riders). In exchange I don't like having many leapers. A disadvantage though as Jean-Louis always says is that these moves are difficult to see by humans.

Hi, HG.

I've been following this thread off and on, as I have a fondness for what others see as huge variants. I would argue that when you come to something like this: "A fundamental issue is that it takes at least as many moves as you have pieces to move all your pieces" you are seeing a restriction that may not need to be there.

Chessplayers as a group seem to be inherently conservative, and highly resistant to significant changes to any version of chess, and even minor changes. So something as radical as suggesting a multi-move approach to speeding the game up, which it will, leaps past heresy directly into the depths of anathema. So be it. Use 2 moves per player-turn to speed up the game. If that is too radical, allow an optional pawn move each turn. Move a piece and a pawn each turn, with no requirement to make the pawn move.

I was running out of ideas for how to speed up such large games. And for the variants I had in mind (14x14 and 16x16) the ideas I already discussed seemed sufficient for making those playable. A fundamental issue is that it takes at least as many moves as you have pieces to move all your pieces. And it sems silly to have pieces you would never get to moving. I suppose that this would not be so bad if there was such a large variety in strategies that you would use different subsets of the pieces in different games, so that in the long run all pieces get used equall often.

I guess another way would be to incorporate aggressive multicapture, which initially would quickly destroy the majority of the pieces before they moved. So that depending on how the opponent would use them you would be left with different subsets of the initial pieces, small enough to use them all without making the game unduly long. E.g. a piece that would capture as in Atomic Chess, destroying itself in the process. Players would be under pressure to use these pieces quickly, while the population density is still large and the strongest pieces have not yet been traded, to cause the most destruction.

I don't like leapers with very long leap. There is basically no defense against those; they can attack you with forks from within a safe location that you cannot reach. If you make them lame leapers, they get too easily blocked.

HG, You have sort of abandoned this thread for a while, and to me it is a very interesting discussion. The min 3 steps idea seems very good to me. I was contemplating something these days in relation to centennial chess. In that game you can move twice before the first capture. Maybe for such a large variant you could think about introducing the double non-capturring moves at any time. The games tend to become rather localized otherwise. Or in contrast to that you could use more leapers with long leaps like say a (4,1) leaper. Or you may replace the wizard (as you already have a camel) with (1,1)&(4,1). This will give more of an over the whole board sensation. On the other hand too many forks from such leapers could lead to a sentiment of randomness. And also easily jumping the pawn chains is rarely a good idea. What I'm proposing the is to have long path movers (like the falcon from falcon chess) with more (say 16 or 20) destinations!

I probably understand why this site sleeps...

...simply check out Webmaster's page...

{if he will see this comment, I will delete it}

@HG, I'm using the Tamerlane giraffe but with the caveat that it can stop at camel square, and a manticore that must do 2 steps at least (may stop at knight suquares). That in different games. What you say, I have not considered. I did consider something even weirder though, because the two above mentioned pieces are hard to develop in the opening I have considered pieces that have the first step pass through pieces another of a few others blockable and then ordinary. Also, now you got me thinking about sky rooks combined with Tamerlane picket like pieces.

Please expand a bit on mezors! I don't understand what you mean!

 It's a category of chessmen who have mating position but cannot reach it.

Not on this category, but I also will note that Commoner also has a power near the Rook.

Have you thought about lame versions of the Slip Queen or Skip Queen, or mixtures of the two (e.g. the compound of a Slip Rook and a Skip Bishop)? These have similar mobility to a Rook on any size board.

@HG,

Things I love!

The idea of the Silly Sliders was born out of my search for a piece that would be almost a Bishop, but not color bound. I wanted that because I wanted to investigate whether the Bishop pair bonus had anything to do with color binding, or whether it is just a consequence of the move sets of two Bishops on opposit shades have so many orthogonal contacts. (But I still haven't gotten to doing that...) First I considered replacing just one of the F moves by a W move, to have minimal difference. But because it then attacks orthogonally adjacent squares it has mating potential, which could make a difference. One way to avoid that was replace all F moves by W. And then I liked the resulting piece.

But for the actual test it would probably be better to take a divergent piece that replaces only one of the (backwards) mF moves of the Bishop by an mW. That also avoids the problem of mating potential.

The Daring Dragons were made for show-casing a piece with an unusual area binding (odd/even files instead of square color), also in relation to pair bonuses this might involve. I started to like the vRsN (the move pattern of which reminded me of a Dragonfly) because it turned out to have good mating potential despite the fact that the usual method for checkmating with orthogonal sliders (forcing the bare King to take opposition along the ray that traps it, and then make one slider step to check and force the King back) cannot be used there. But there is a very peculiar method for forcing the bare King to take opposition when your King is on the orthogonal, so that you can force it back by checking through a double step.

In the case of afyafWZ even if 4 of the moves are not blocked as they were in the second rook squares, the Z squares are longer and more often off board. I think you are correct though in assessing the this piece is a bit stronger than a rook.

BTW: I like the silly sliders and the daring dragons quite much!

Well, taking away the first few steps of a slider (but keep the remaining moves blockable on those squares), and replacing those by leaper moves away from the path is a good way to make new pieces without altering the value too much. I did the same in the Silly Sliders army for CwDA, replacing the W steps of the Rook by F, and the F steps of Bishop by W, and the K steps of the Queen by N. If you also replace the second move of each slide by a direct leap, the value would probably increase (by about a Pawn), as those moves were blockable, and the leaps no longer are.

Pieces without any stepping moves are rather cumbersome to manoeuvre, though.

BTW, better to write afyafW than yafafW: the latter already changes to Rook after the first step, so it does W+R+R. This is still a minimum of 3 steps, but for larger distance the R+R part can be realized through many diffeent intermediate squares. (It really describes a 'degenerate' hook mover, with two sliding legs.) This would slow down the Interactive Diagram's AI, which is not clever enough to recognize the moves as the same, and would try to search all of them. With W+W+R there is always only a single realization.

I had never thought about bend hook movers. Thanks for pointing these pieces out. Anyway, the thing that I was asking myself is how weird would be the pieces I was talking about for a more regular chess player. Probably pretty much, given the fact that I have met players who see the Capablanca pieces too complex. I don't understand why, but anyway!

If you want edge-seeking pieces, you could use hook movers that can only turn a corner after some (large) number of steps. E.g. B-then-R that can make a 45-degree turn only after making 6 or more diagonal steps. In the center of a 12x12 board this is just a normal Bishop. But in the corner of an empty board it would cover 1/4 of the board.

On a 12x12 board camel and zebra are as short as the knight is on a 8x8 (the same ratio between the longest move and the board length). I'm sure that there exists a large enough board where Giraffe=Rook. The pieces I mention are maybe more like a giraffe that can move to dababah and threeleaper spaces.

It's arguably a little bit of a stretch to call Camel/Zebra leaps genuinely short; normally that refers to anything within two king‐moves away (The Short‐range project differs a bit on this, but partly because a lot of those games (especially Joe Joyce's) tend towards even larger boards). So your proposed pieces have the same kind of personality (broadly speaking) as Tamerlane's Giraffe (which in turn leads to the question of that piece's value — might it be comparable to a rook?)

@Bn Em

The lack of personality comes from replacing some short moves with other short moves. Most people won't like that I'm afraid!

@Lev,

Please expand a bit on mezors! I don't understand what you mean!

@Jean-Louis, Well, the pieces you mention are not rook strength though! There is also tripunch chess where bishop+manticore and rook+gryphon are featured.

I also have some Rook-strength pieces: Longdebeest (Ö in my notation, CFX) and Dove (Ď in my notation, fFvDbsNbAvHfG). Both are majors; both are also weird (especially Dove).

{Also I think about new piece category, Mezor, which are minor in most variants, but become major in drop variants (such as my Warrior, Copper, Silver, Wildebeest, several Argentinian hoppers, etc.)}

@Bn Em: yes I knew about Qilin vs Kirin, Kylin, but it's just a working name for me at this stage. I wanted to stick with something more or less dragon-like. Thanks for pointing me to 4 Linepiece Fusion.

@Aurelian

The pieces you're proposing, with, as you say, less emphasis on centralisation in exchange for greater power at range, remind me a little of a more organic way to achieve what Big Outer Chess was going for. I wouldn't worry a priori about a ‘lack of personality’ from being part‐leaper and part‐rider; even things like the Archbishop have plenty of character, and there's certainly a distinctiveness to such a dispersed pattern of movement

As for partial bent riders, if not the chiral ones what about Ships/Snaketongues or their sideways or (as yet unattested afaik) crabwise counterparts?

The compound cannon would indeed probably deserve careful handling

@Jean‐Louis

Your Godzilla is not particularly less new than your Simurgh/Qilin: Gilman gives the latter two as respectively Metropolitan and Ancress — and even uses them in Four‐Linepiece Fusion — though ofc it makes sense that you minght prefer to stick with the monster theme for the names.

That said, whilst duplication of ‘Simurgh’ is fine (Gilman uses it for the viceroy‐then‐bishop viceroy‐then‐rook which only exists in 3D (or hex)), it seems odd to me that you'd choose to duplicate Qilin, which is just the Chinese version of Japanese Kirin, i.e. the familiar FD

@Aurelian. On my side, for a giant chess (on 16x16) I am exploring, for few weeks now, two compound bent-riders that I called Simurgh: Gryphon + Bishop and Qilin: Manticore + Rook. In the same game I have also the Godzilla, not new, which is Gryphon+Manticore.

They are very strong pieces. They need a crowdy large board.

The mRpR sounds interesting too. That is a compound of the Xiangqi and Janggi cannons. I have not considered it. Now that I'm thinking it is always around rook level but more useful in the opening that the endgame. I think it could cause trouble in the opening though. I definitely don't like asymmetric bent riders. For 12x12 R3fB3 is probably below rook. Isn't this piece too blend though? But I was asking more about the pieces I was proposing.

What about FyaflF or mRpR? I imagine those might be close to a rook. Or maybe something like R3fB3. Maybe it depends on what other pieces you have

Hello guys, I was thinking about obtaining other pieces of rook strength. I thought about the yafafWZ (a piece that moves at least three rook steps and may also move like a zebra) and the yafafFHC ((a piece that moves at least three bishop steps and may also move like a camel or a threeleaper). These are pieces with a steppe learning curve but I had noticed they have something to offer. Most pieces are better when centralized but these even if they lose some leaper jumps when close to they edge the gain long rays. Also, it could be argued that they don't have personality as they are leaper on one hand and a hard to use rider on the other. I am thinking about this pieces for a 12x12 game. Any thoughts?

I've created this by simple editing and copypasting but error occured https://www.chessvariants.com/play/pbm/play.php?game=Stone+Garden+Full+Contact&settings=stonegardenvarfc&submit=Edit

Do you want to change your ID of Cryinto or your name of Rechefiltr is Fire? The former will involve replacing it in multiple tables of the database, and we have no script for that. The latter should be something you can edit yourself.