Check out Glinski's Hexagonal Chess, our featured variant for May, 2024.


[ Help | Earliest Comments | Latest Comments ]
[ List All Subjects of Discussion | Create New Subject of Discussion ]
[ List Earliest Comments Only For Pages | Games | Rated Pages | Rated Games | Subjects of Discussion ]

Comments/Ratings for a Single Item

EarliestEarlier Reverse Order LaterLatest
Play Chess Variants with Jocly. Missing description[All Comments] [Add Comment or Rating]
A. M. DeWitt wrote on Mon, Feb 5 02:24 PM EST in reply to François Houdebert from 02:17 PM:

Here's an install link to Shogi, seireigi, kyoto, mini, but note that this may change soon.

The link is broken, so you'll have to send a new one.

For the record, I tested the feasibility with chu seireigi but couldn't make the lion/eagle and falcon moves correctly. But it's still interesting to see what it could become.

The Seireigi Lion moves are basically the exact same as Chu Shogi, except the second step can only be done when the first captures something. I sent you a comment on Chu Seireigi's page explaining what I found. You can recruit H. G. Muller's help with the Lion moves.

Hectochess should be easy to implement. It's basically just Chess but with 5 new piece types on a 10x10 board and a few tweaks where needed.


François Houdebert wrote on Mon, Feb 5 02:38 PM EST in reply to A. M. DeWitt from 02:24 PM:

The same link in https

For the Lion we probably could make it work with HGM (it needs to combine drop model + fairy move) but we might finish what we've started first.

Indeed we probably have already everything for hectochess. I might give it a try.


A. M. DeWitt wrote on Mon, Feb 5 02:52 PM EST in reply to François Houdebert from 02:38 PM:

For the Lion we probably could make it work with HGM (it needs to combine drop model + fairy move) but we might finish what we've started first.

Yeah, I would finish what you started first. No need to overwhelm ourselves.

I don't think I will be able to figure out a lot of key aspects for Jocly b/c it is such a complicated interface to program, so I think I it is more prudent to work on other things and let you two handle the Jocly implementations.


François Houdebert wrote on Tue, Feb 6 03:53 AM EST in reply to A. M. DeWitt from Mon Feb 5 02:24 PM:

Here's a draft of hectochess, which can be accessed from this dashboard so that you can check the rules and thumbnails.

I've chosen to keep the usual representation for the Leo so that people used to Jocly can play directly without consulting the rules too much.


A. M. DeWitt wrote on Tue, Feb 6 10:04 AM EST in reply to François Houdebert from 03:53 AM:

The link to the Hectochess implementation is broken, please fix.


François Houdebert wrote on Tue, Feb 6 11:19 AM EST in reply to A. M. DeWitt from 10:04 AM:

The link to hectochess is restored. I'm going to make some corrections to the shogi model, but I'll probably wait until I'm sure it's fully feasible before finishing it properly.


A. M. DeWitt wrote on Tue, Feb 6 12:20 PM EST in reply to François Houdebert from 11:19 AM:

Okay, the link is fixed. However, it goes to the games dashboard instead of the implemetation itself. I checked, and I didn't see Hectochess listed (or maybe I missed it); however, I did find the implementation's URL.

I have inspected the Hectochess implementation, and it is quite stellar, but there are a few mistakes to fix, which should be pretty easy:

  • Black's kingside Knight and Champion have been switched (Champion should be next to the Rook, and Knight should be next to the Bishop)
  • White's promotion zone is two ranks too far forward. It should be at the farthest rank.
  • A King should be able to castle two or three squares to either side. Currently, the implementation only allows the latter.
  • Your board has the white corner square on the left (it should be on the right). Because of this, you ended up swapping the King and Queen (from White's perspective, Queen should be on left side, and King on right side).

Also, some things worth mentioning:

  • Hectochess uses a 64-move rule instead of a 50-move rule, due to the increased board size.
  • Since there are so many Pawn promotions, it may be beneficial to split the options into two rows or have scrolling on the promotion dialog.

François Houdebert wrote on Tue, Feb 6 01:20 PM EST in reply to A. M. DeWitt from 12:20 PM:

It should be better but castling is not yet fixed.

If you want to check the rules : try again on here and follow this image


A. M. DeWitt wrote on Tue, Feb 6 01:36 PM EST in reply to François Houdebert from 01:20 PM:

The board looks better but there is a weird anomaly in the pattern on the last two ranks from White's perspective, which makes it so that two of the rows appear to be merged into a single row, and that Black's white corner square is on the left.

The other problems are all fixed (including King and Queen setup).

As for castling, it shouldn't be too hard to program...I hope.

P.S. I think your update broke the Werewolf Chess image on the dashboard, just so you know.


François Houdebert wrote on Tue, Feb 6 05:17 PM EST in reply to A. M. DeWitt from 01:36 PM:

I think it's working well now.

Any objections to the rules?


A. M. DeWitt wrote on Tue, Feb 6 08:57 PM EST in reply to François Houdebert from 05:17 PM:

The only problem I can find (which probably isn't that big of a problem in the first place, but may still be worth fixing) is that it takes 50 moves w/o moving Pawns or capturing to draw in the implementation, whereas Hectochess increase that threshold to 64 due to the increased board size.

Apart from that, everything works properly.


François Houdebert wrote on Wed, Feb 7 03:11 AM EST in reply to A. M. DeWitt from Tue Feb 6 08:57 PM:

I think the 64-move rule could be operational now(not tested).

As for the size of the promotion panel, I can’t change it easily, but as the number of large pieces is not limited, I can't see a situation where a bishop, rook or knight would be chosen, so the question arises of keeping them or not.


H. G. Muller wrote on Wed, Feb 7 04:17 AM EST in reply to François Houdebert from 03:11 AM:

I think it should be possible to control the size of the promotion panel in the view file for the game. I remember I have seen that once. Unfortunately I don't remember where. Probably in a variant with lots of choice.


François Houdebert wrote on Wed, Feb 7 05:38 AM EST in reply to H. G. Muller from 04:17 AM:

thank you for the advice, I'll look into it but from my point of view, as it stands it's acceptable


A. M. DeWitt wrote on Wed, Feb 7 10:28 AM EST in reply to François Houdebert from 03:11 AM:

I think the 64-move rule could be operational now(not tested).

From my testing it took 62 moves to reach a long game draw, so either you are slightly off. or I may have miscounted.

I did find the line of code you will need to update.

In hectochess-chess-model.js it is this block of code:

				// check 50 moves without capture
				if(this.noCaptCount>=100) {
					this.mFinished=true;
					this.mWinner=JocGame.DRAW;					
				}

It should be changed to this:

				// check 64 moves without capture
				if(this.noCaptCount>=128) {
					this.mFinished=true;
					this.mWinner=JocGame.DRAW;					
				}

As for the size of the promotion panel, I can’t change it easily, but as the number of large pieces is not limited, I can't see a situation where a bishop, rook or knight would be chosen, so the question arises of keeping them or not.

I would keep the Rook, Bishop, and Knight, for consistency reasons.

Hectochess has been out too long for me to change this. Several other programs, such as ChessV, and Ai Ai, have implemented this game, and I suspect that all of them allow these promotions.


H. G. Muller wrote on Wed, Feb 7 10:58 AM EST in reply to A. M. DeWitt from 10:28 AM:

It is a bit strange that this test is made in the game-specific model file. It would be much more logical to have it in the chess base model, which also keeps track of the noCaptCount, and compare it to a variable moveLimit that by default would be 50, but could be changed in the game's model file.

This 50-move stuff was a weak spot in Jocly anyway, as the noCaptCount was not reset on Pawn moves before I fixed that.


François Houdebert wrote on Wed, Feb 7 12:06 PM EST in reply to H. G. Muller from 10:58 AM:

It's true, especially since I thought I'd done it at the model level, but it must have been overwritten during compilation, so I'll have to dig deeper.


François Houdebert wrote on Wed, Feb 7 12:26 PM EST in reply to A. M. DeWitt from 10:28 AM:

That’s the old name : the processed version is  hectochess-model.js

the sources are hectochess-model.js

So I suppose It might have worked this way.

To reproduce it : you play 64 times a piece? and then you should get a draw, right?


A. M. DeWitt wrote on Wed, Feb 7 01:18 PM EST in reply to François Houdebert from 12:26 PM:

After maybe moving up a Pawn on each side, you move a random non-Pawn piece on each side without capturing anything, ensuring as few positional repeats as possible to avoid 3-fold repetition.

I found the problem. You have the wrong number of plies in your check for the 64-move rule, as shown below:

                if(this.noCaptCount>=124) { // this draws after 62 moves (62 * 2  = (50 + 12) * 2 = (50 * 2) + (12 * 2) = 100 + 24 = 124)
                    this.mFinished=true;
                    this.mWinner=JocGame.DRAW;                    
                }

The number 124 should be 128 (64 * 2  = (50 + 14) * 2 = (50 * 2) + (14 * 2) = 100 + 28 = 128). Like so:

                if(this.noCaptCount>=128) {
                    this.mFinished=true;
                    this.mWinner=JocGame.DRAW;                    
                }


François Houdebert wrote on Wed, Feb 7 01:29 PM EST in reply to A. M. DeWitt from 01:18 PM:

you're right, sorry to have wasted your time on this.


A. M. DeWitt wrote on Wed, Feb 7 02:41 PM EST in reply to François Houdebert from 01:29 PM:

It's not a waste if the problem is fixed.

Currently, I still see the 124 in the noCaptCount comparison in your hectochess-model.js file though.


H. G. Muller wrote on Wed, Feb 7 02:45 PM EST in reply to François Houdebert from 05:38 AM:

Well, in case you ever run into trouble with the size of the promotion popup: in terachess-view.js there are the following lines:

  45         // Reducing the promo frame which was overflowing the board screen
  46         View.Game.cbPromoSize = 1100;

So I suppose this is the way you can resize it.


François Houdebert wrote on Wed, Feb 7 03:04 PM EST in reply to A. M. DeWitt from 02:41 PM:

It should be fixed.

But the file may be in a cache. Press [ctrl] + f5 to force js reloading?


François Houdebert wrote on Wed, Feb 7 03:15 PM EST in reply to H. G. Muller from 02:45 PM:

indeed, it works well


A. M. DeWitt wrote on Thu, Feb 8 11:34 AM EST in reply to François Houdebert from Wed Feb 7 03:04 PM:

It should be fixed.

But the file may be in a cache. Press [ctrl] + f5 to force js reloading?

I'll give it a while.


25 comments displayed

EarliestEarlier Reverse Order LaterLatest

Permalink to the exact comments currently displayed.