Comments by FergusDuniho

The only thing I am still in doubt about is whether the current betza.txt include file on which the PTA-generated code relies tests for a move being the final one in a safe and reliable way. Because it uses mln for the test rather than turn.

I have just added thismovenum and maxmovenum. These will return $mline[$mln]->movenum and $mline[$maxmln]->movenum. It's likely that maxmovenum will be returning the same value as movenum, which returns $movenum, but using it will add clarity to your code.

You're using k/K for the general, but the code you copied is using g/G. So kpos and Kpos are not being updated.

The print color scheme now has its own small logo too. Since the big logo included 19th century black and white artwork of Chess pieces, I chose a more retro style that displays fairy pieces by changing the orientation of Chess piece images. According to a section of Anthony Dickins' A Guide to Fairy Chess, the inverted queen is the grasshopper, and the reclining bishop is the unicorn.

if == 0 $mline[$mln]->movenum:

That is mixing GAME Code with PHP. GAME Code does not support classes with properties or write out array elements with brackets, and it doesn't provide read access to $mline. But you can use the value of turn with your knowledge of which side you're concerned with to determine where you are in a game.

I am not familiar with this technique, but a problem could be that it is not just about positioning the images, but also about translating mouse clicks to cell coordinates. (Which must also work for empty cells, for entering non-captures.) In the table version I have attached event handlers to the cells for this.

With grid, you would use a <div> tag instead of a <td> tag for each space, and you could attach event handlers to it just as easily. What I have done in Game Courier, though, is place a transparent image over each empty space, as I have attached event handlers to the piece images.

set all == mln $maxmln;          // indicates last move

Note there is no $ there, so we might be talking about different things, and this one might be a legacy variable.

This operator just outputs the value of $mln.

I cannot image how this would work if mln was not the current move number.

It works because you are comparing mln with its maximum value. When there are comments, $maxmln will be greater than any move number, but it will be no higher than mln will eventually reach.

The I.D. represents the board through a HTML table,

Game Courier used to use HTML tables for hexagonal boards, but I gave up on this in favor of using either CSS with a board image or drawing the board image. This reminds me that I was working on getting CSS to work with a generated board image so that you could view each position of the game without reloading the page, but I haven't completed that. I think I had two different approaches and wasn't sure which one to go with. I either wanted to change how I generated the board so that the image would be perfectly symmetrical, or I wanted to modify the code to work with the asymmetrical boards I was already generating. I lost sight of what I was doing when I got caught up with other projects.

in theory it should be possible to create a table with a masonry-like tiling, shifting each subsequent rank by half a cell.

I imagine this would be easier to do with CSS grid, because, as far as I know, table columns remain vertical.

I have tried this, though, and it works to some extent. But for reasons that I do not grasp yet it also changes the height of the ranks in a way that I could not control.

Using grid would probably help you avoid this problem too. With grid, you should be able to make one column or row diagonal, though I have not yet tried using grid with hexagonal boards. My CSS code for hexagonal boards uses absolute positioning, as grid was not yet part of CSS when I wrote the code. Anyway, absolute positioning is another alternative to using tables.

I think the variable 'mln' gives you the current move number in the Post-Move sections.

It does not. I have made that mistake too and had to correct it. To use $mln to get the move number, you need to get $mline[$mln]->movenum. $mline is an array with a separate entry for each line of the movelist, but besides containing actual moves, it contains comments. $mln is the index that a line has in this array, but when some lines are comments, it will go out of sync with the move number.

GAME Code has turn, which will return $mline[$mln]->turn, but it doesn't currently have anything returning $mline[$mln]->movenum.

The Official Glinski Coordinates preset has CSS rendering which isn't working right for hexagonal games

The board image it was using had two separate elseif clauses dealing with it in a long series of elseif clauses in image_dimensions.php, and changes I was making to the second one had no effect. But when I noticed that it had two and combined them into one, I was able to set a value for $offy that worked for the board.

Is it possible to make an Interactive Diagram use a hexagonal board?

I may make a logo for it with some of Tenniel's black and white illustrations from Through the Looking-Glass, though I'm open to other ideas.

I have now done that except that I cropped and divided one illustration instead of using two separate illustrations. Also, most of the small piece images I used are based on characters Alice encounters in Lewis Carroll's books.

Glinski's Hexagonal Chess is the featured variant for May, 2024.

The color scheme code I was working on yesterday and into today is now on the site. As I mentioned before, it will now change the theme with JavaScript as long as JavaScript is enabled. This is done mainly as a backup for older browsers that do not support :has() in CSS. It uses both JavaScript and CSS to change the color scheme so that if one method fails, the other one might still work. Generally, the CSS method will fail on older browsers, and the JavaScript method will fail if someone has JavaScript disabled.

One new feature is the Print color scheme. This is a more minimalist version of the Light color scheme. I may make a logo for it with some of Tenniel's black and white illustrations from Through the Looking-Glass, though I'm open to other ideas. For now, it uses the same logos as the Light color scheme.

Internally, I have moved the color scheme CSS to colors.css, where it is better organized and more maintainable than before. First, it creates individual custom properties for each color in each color scheme. The light ones begin with --light-, the dark ones with --dark-, etc. Since it has to define each color scheme multiple times for different selectors, it defines them in terms of the individual custom properties already created. So if I decide to change a color, there is now only one place it has to be changed.

Also, this provides page authors with the ability to customize color schemes for a particular page. By adding a style section in which you add custom properties to :root, you can rewrite individual values for particular color schemes.

I could not figure out what closing tag you were mentioning.

I already fixed that bug, and it was in my code, not yours.

This ID still does not work properly!

I tested it in Safari, Edge, and Firefox this morning, and it was working in all of them, though I did have to reload the page in Firefox before the board would display.

When I go to any person information page and try to view unpublished submissions, it just shows my own submissions.

Okay, I corrected that.

There was a typo in implementing a new property I had added. I added the displayPieces property, which when set above zero will display the piece table when the page loads. This much worked as it should, but when it was not set, it omitted the closing > in the tag, which caused what followed it to not be displayed.

It's important to maintain backwards compatibility, and anyone who doesn't like the name of Aanca is free to use a different name with an alias.

For some reason, it cannot find the description even though the path to that file is correct, but I'm not really concerned about that.

It wasn't prepending ROOT to the filename. So it was actually incorrect. I have now fixed that.

Therefore Fergus created separate HTML pages for individual variants, which embed the Jocly applet preconfigured to run the corresponding variant, with a rule description. The overview page you refer to contains links to those individual game pages.

While I used to make separate HTML pages for individual variants, I replaced this practice with a PHP script that gets called through a semantic URL whose relative part starts with "/play/jocly/". If you want to review its code to see how it works, you can find the script at "/play/jocly/control.php". It relies on certain files being in place to display the description, rules and credits. It should work with games that have not been indexed yet, and if you do use it with unindexed games, you can, as an editor, add them to the index.

Please remove these unneeded lines from Edit Metadata also. I’ve written new page, thank you, but when I tried to revert it to Members-only, error repeated. It shows that I’m veteran contributor with nine or more submissions, so I cannot publish more if I have nine submissions to review (even if it’s not).

While it was giving you the same error message, it wasn't due to extra code again. This time, another variable was undefined, which caused its counts of the database to return the wrong values, and the solution was to add a new line assigning its value.

I'm sure there's a reason for this, but why is it that having legality defined for move highlighting isn't enough to have legal moves enforced?

1. Rule enforcement was a feature before showing legal moves was. Instead of replacing the standard method for rule enforcement, calculating all possible legal moves ahead of time was done as an extra step.

2. Additionally, moves sometimes had side effects that were not explicitly mentioned in the notation for a move, such as en passant capture in Chess or changing the side of a piece and placing it in hand for Shogi. This imposed the requirement of running code for the actual move when it was made.

3. Because calculating all possible legal moves is more expensive than just calculating whether one move is legal, it is done only in the Post-Game section. So, if rule enforcement depended upon the calculation of all possible moves for each turn, it would work only in on-going games made one move at a time, or it would have to be repeatedly done in the Post-Move sections, which would prolong the calculation of whether a series of moves were all legal. The way it currently works, it evaluates the legality of a series of moves much more efficiently than it would if it had to calculate all possible moves after each move. For the sake of efficiency, which kind of matters in an interpreted language written in another interpreted language, it makes sense to calculate all possible legal moves only when there is a choice of which move to make. For past moves, calculating all this is just overkill.

[Edit: combined 3 and 4]

That’s not what I meant. I was looking for Reche's Super Faceoff Masquerade and could not find it. Please be clear about what you were doing and walk me through it step by step.

Happy Birthday.

This is for the Darker color scheme, which has a black background. Blue will be fine, since it shows up well against black, and it contrasts with another piece being white. The tallest piece in the Dark scheme's logo is 427, and the height of the logo is 432. Without the pieces, the text part of the logo has a height of 415. If I do like I did with the Light logo, a piece height of 285 would work.

Since the fairychess include file uses a shortcode to display pieces, I changed how the shortcode works instead of any code in the fairychess include file. I modified it to include WIDTH and HEIGHT values in the IMG tag when the image is an SVG image. These will be equal to $width and $height when these variables have values, as they usually do in Game Courier, and they will be 50 otherwise.

I have now updated Mobile Detect to 4.8.06. Since it works differently than the old version, mobile detection was a bit wonky until I got it working correctly.

The inside part should be colored #f9f9f9. Using the background color for inside the piece makes it transparent and impossible to recolor properly.

I modified the Dark color scheme to use a darker color for the text that works well with the chocolate background color. This is to make reading more comfortable on brightly lit screens. However, it was too low contrast on my Likebook Mars.

So I have added a new high contrast color scheme called Darker, represented by the New Moon emoji in the color scheme selection form. This is intended mainly for eink devices for which the Dark color scheme may be too low contrast, though you may still use it on other devices if you prefer it. In general, its background colors are darker, and its text colors are lighter. For the main background and text, it uses black and white. Unlike the Light and Dark color schemes, this one is not available through your browser settings. It is available only by selecting it on the website, and you may have to refresh your cache before it starts working.

I am not sure where these set files are stored,

They are stored in

/home/chessvariants/public_html/play/pbm/sets/

nor am I sure how to upload the new versions to the site.

Do you know how to use SCP or SFTP? Do you still have the passwords I sent you by email?

My iPad has iOS 17.4.1, but I do have access to an old iPhone with iOS 15.8.2. The dark scheme partially works on it, but I see that the logo is not changing. It shows the elephant and unicorn logos instead of the dragon horse and (fairy) princess logos even when it is in dark mode. By substituting an older CSS file in index.html, I could tell that color-mix was not working, and blue but especially indigo did not contrast well with the dark background. I found this iPhone had the same problems in both Firefox and Chrome. At least using custom properties instead of color-mix has fixed the problem with the link text color.

To get the logo to change, I tried using the <picture> tag. As I feared, it would not work unless dark mode was selected at the browser level. This is why I normally use CSS to switch the logo. Chrome and Safari did not provide me any means to select dark mode at the browser level, though Firefox did. But when I selected dark mode in Firefox at the browser level, it wasn't working. Using test pages, I got it to work for browser selected dark mode by removing a test using :not. But doing this disabled the ability to select the light scheme from the menu when the browser's dark mode is turned on. I tried to fix this, and while my fix works on my desktop, it does not work in the Firefox app on this iPhone. So, I have no solution for getting the right logo to show up on this older iPhone that would work equally well with both methods of selecting the color scheme.

I still don't get why the previous way wasn't working since I wasn't changing the pawn subroutine which seems to modify promotion options before the last rank.

The Pawn subroutines are only for actual moves, but nothing was being done to handle variable values of wprom or bprom for potential moves. The stalemated subroutine was supposed to provide a list of all legal moves, but it was omitting moves where someone declines promoting a Pawn. By having functions that dynamically calculate what a piece may promote to, it is now able to provide an accurate list of legal moves for games with variable promotion rules, such as Gross Chess has.

Anyway, this reveals another, but more minor, problem, which is that when a pawn gets to the promotion zone the text in the option selection box is large enough that the lines overlap slightly.

I increased the line-height and made some changes to the borders.

To get access to the CVP pages I logged in with an account of a friend.

You didn't have to do that. You can post comments without signing in.

I have not been able to access the CVP pages for a few days now. ... Something seems to have gone wrong with my password.

I have sent you a new password by email.

In looking into this, I tested Gross Chess to see if it had the same problem, but before I could tell, I encountered another problem with it. I realized that for a game like Gross Chess it wouldn't do to use bprom and wprom as though they had static values. So I rewrote the fairychess include file and the Gross Chess code to support dynamic values for what a piece is allowed to promote to on a given space.

This makes use of some new functions that end with "-Promote". Here are the default functions for the Pawns:

def White_Pawn-Promote var wprom;
def Black_Pawn-Promote var bprom;

For backwards compatibility with the original way of handling promotions, these just return the value of wprom or bprom. And for additional backwards compatibility, the stalemated subroutine will use these functions only if the piece is not in the promotable array. So, to enable the use of these functions for providing dynamic values for what a piece can promote to, you should unset promotable or set it to an empty array. I added this line to Gross Chess after including the fairychess include file.

unset promotable;

Since the default functions return static values, they need to be rewritten for the particular game they are for. Here are the functions I wrote for Gross Chess:

def White_Pawn-Promote merge intersection var cap elem - rank #0 9 ((B N V W) (B N V W C R S) (B N V W C R S A M Q)) elem - rank #0 9 ((P) (P));
def Black_Pawn-Promote merge intersection var cap elem rank #0 ((b n v w c r s a m q) (b n v w c r s) (b n v w)) elem rank #0 (() (p) (p));

Since what a Pawn may promote to in Gross Chess depends upon the rank it is on, I used the rank value for black (or a value calculated from the rank value for white) as the index for a couple of arrays from which it extracted a particular value. For example, black can promote on ranks 0-2, as they are designated internally. So, this code will return the element of the array with the same index as the rank value:

elem rank #0 ((b n v w c r s a m q) (b n v w c r s) (b n v w))

Since white promotes on ranks 9-11, I subtracted 9 to get a value from 0 to 2 for any rank promotions are allowed on or a number that is out of range for any other rank. So, this works similarly:

elem - rank #0 9 ((B N V W) (B N V W C R S) (B N V W C R S A M Q))

Since the last rank for black is 0, and the last for white is 11, and 11-9 is 2, these list sets of promotion options in the reverse order from each other.

Since promotion options are limited to captured pieces, each function calculates the intersection of the value above with the captured pieces. This looks like this for black:

intersection var cap elem rank #0 ((b n v w c r s a m q) (b n v w c r s) (b n v w))

Finally, I get to the part that is relevant to Obento Chess. Whether it can promote to a Pawn as a way of declining promotion depends on the rank but not on what has been captured. So Pawns were not included in the main lists of promotion options. Instead, it merges the intersection calculated above with the value of another array element. Again, the specific array element is a function of the rank. Here is what it looks like for black:

elem rank #0 (() (p) (p))

Since declining promotion is not an option for black on rank 0, an empty array is provided for the element with an index of 0. This is not necessary for white, as the rank it cannot decline promotion on has a higher index.

elem - rank #0 9 ((P) (P))

For Obento Chess, you might use functions like these:

def White_Pawn-Promote elem - rank #0 9 ((FP) (FP) (F));
def Black_Pawn-Promote elem rank #0 ((f) (fp) (fp));

You could handle promotion for the other promotable pieces with similar functions for each specific piece.

Much time passed, but still. On dark theme.

Indigo font on brown/dark-grey/purple background is hardly readable. Blue is just slightly better.

It's not like I've known this to be a problem and have just done nothing about it. Different browsers have different capabilities, and if what seems like an obvious problem to you goes on for a long time without being fixed, this could be because I am not experiencing the same problem as you are.

This particular problem is probably because I am using color-mix for the link colors instead of custom properties like I am for all the other color changes, and custom properties have better support than color-mix. So there may be an occasional browser that supports custom properties without supporting color-mix. However, I have not been able to locate one. While I have some old devices with old browsers that do not even support dark mode, every browser I have found to support dark mode supports it fully with both custom properties and color-mix.

So, first, I would like you to report to me which version of which browser on which operating system or device has this problem. Then I would like you to make sure everything is up-to-date. Make sure that your device is fully updated and that your browser is the latest version. After doing this, let me know whether the problem persists.

That's now fixed. Thanks.

I have now restored your password. The problem was that your code had an unclosed <script> tag at the end, and this enveloped the code that followed it. I switched your code from WYSIWYG to HTML and added the missing </script> tag. As long as you're including scripts in your code, avoid using WYSIWYG mode.

With the arms and the tail, it is looking a lot more like a centaur

Okay, I have published this. Note that the change I made to the HTML for the Chezs pieces is actually simpler than the tables you used to display the rest of the pieces. While you spread out information about a piece among multiple table rows, my modified code kept all the information about a piece together. Doing this makes the code easier to read, and it makes it easier to add new pieces. I will leave changing the rest of the code up to you as an exercise. It remains up to you whether you want to do it, but as I mentioned, it’s advantages do go beyond how is displays on phones.

It would be nice to include some cropped versions of these images with links to the 3D printer files.

I’m sorry to hear that. You have my condolences.

Please use full names and links when nominating games, and please name every game you nominate within your comment so that nominations can be properly documented.

The minimum would be to give dates of invention to put the information in a historical perspective.

I copied your list to the page and made the URLs relative. If you want to expand that list or add more content to the page, you can post it here.

This is obviously an April Fool’s Day joke. So it’s nothing to worry about if you use this rival site. Regardless, this is not the type of subject we have webpages for. So I will delete this page after posting this comment.

42 is the age Adams was when his first and only child was born, though he was 30 when he published Life, the Universe, and Everything.

Grant Acedrex is now the featured variant for April, 2024.

If the idea is to represent the move, the logical form in a set where the Amazon is a Staunton Knight with the head of a Staunton Queen mounted on top of it, would be a King's head on a Knight. The cross could be omitted for the non-royal case.

I'd trusted that Fergus had already thought of that and hadn't found a way to make it work, for some reason

I hadn't thought of it, because a crossless crown is still a crown. Having thought about it now, the cross is the main thing distinguishing the king's crown from the queen's. So, a knight with a crossless crown could easily be confused with an amazon. My thought has been to combine man and horse in some way.

It'll be interesting to see if anyone else can, as it would avoid using a 'reverse centaur' figurine. Maybe a Man's head (from Piececlopedia page for that) on top of a knight could be even better.

As crowns and miters are something worn on the head, some other kind of headgear, such as a helmet or a hat would probably be suitable for a piece with the additional powers of a man. While this might not work as easily for rook+man or bishop+man, these are also rook+ferz and bishop+wazir and are known by other names, such as dragon horse and dragon king. So maybe they could be represented in a different manner. Or all three might be represented by human face pieces like in the Superba set.

I have now fixed this with the solution that was already working for Shogi. It adds skip as the second move when the piece doesn't promote, which it now allows as a second move, and it doesn't ask which piece to promote to again if the move contains skip.

Indeed. Slavish adherence to perfectionism is a roadblock to getting anything done. It's better to do what we can for now even if it is imperfect. A bad solution may inspire someone to come up with a better one, or feedback may lead the person who came up with the initial solution to come up with a better one. In biological evolution, science, and other endeavors, improvement comes through trial and error. So we shouldn't be trying to discourage people from offering solutions we don't like. Instead, we should just critique solutions we don't like and try to offer better ones if we can.

It looks fine on my desktop or iPad, but on my phone, it requires either a lot of horizontal scrolling or zooming the page down to a small size to see all the images. So, I recommend switching from using tables to using flexbox. What I have in mind is replacing each table with a single row that will wrap when the screen isn't wide enough for it. To illustrate how it works, I have rewritten the Chess pieces section to use flexbox. I have styled it to resemble how your tables look, but if you decrease the width of your window, you should notice that it behaves differently. To make the other sections work the same way, all you have to do is consolidate the tables cells for each piece into a single div, replace each <td> with <div> and each </td> with </div>, and replace the table tags with <div class="piecerow"> and </div>.

Does anyone want to nominate or second any more games before the month is finished?