Comments by benr

I filled in the board size metadata as 8x8, but perhaps it would be better as 8x15 (still with 64 cells)?

You should add the western-style piece images for the rest of the pieces in the Pieces section (to align with the interactive diagram).

Snark or not, I agree with H.G. that this is entirely too much art for a piececlopedia page. If you wanted to get feedback on the various outputs you've generated (to decide on one or two for the page), that should've been done in comments.

The promotion zone the article mentions makes no sense to me

Indeed. And the CECV and Moeser both say that the promotion zones are E*5 and A*1. (Not also D*5 and B*1 as you suggest, perhaps just because the pawns there can still advance one more.)

Also, @Fergus, the green version of the item description here is completely unreadable on the new background color of comment headers.

From Jeff's website, in the Torus Games source code zip, there is a Revisions History file that lists Jan 2006 as "Initial Torus Games 2.0 release." I've tentatively set the invention date for this game as 2005.

Hmm, I didn't have the Centaur in mind for this tag, but I guess it fits (Royal Court, Sac Chess). Any opinions about that?

What about pawn compounds (any existing games aside from some absorption games? Speaking of...should Absorption Chess be here?)?

@Fergus, clicking the links at the top of this comment (the link to the tag page or its comment list) doesn't work, again something about the + in the name. I suppose at this point we should just change the name and avoid url encoding characters in tag names?...

@BnEm: this page doesn't qualify for the Chess+Compounds tag, with the unicorn (debatable I suppose) and jester (not debatable, I think).

I've started occasionally getting full-page google ads on my phone when clicking from one CVP page to another (or coming back to a page from a sleeping screen). They seem to be between the two pages: the second page takes a moment to load after I dismiss the ad.

As the author emphasized in their last comment, the conditions apply "After your move".

The chess board is already too small. Why would you make it even smaller?

I particularly like small or mini variants.

• Clarifying the rules on the Pawn's double step that Reiniger mentioned
  • Does the double step for Pawns also apply to White Swans?

This has already been clarified in the page.

• A distinctive piece image for the Black Swan
• The movement of the Black Swan

Black Swans aren't pieces, but one possible outcome (specifically, any outcome with multiple pieces) when revealing a White Swan.

If a White Swan coming from the first rank gets flipped over on the second rank, it can be subject to capturing.

This should be in the Rules section in my opinion, if it is meant as a rule. You may want to be careful with this though, as you may encounter a problem with having to keep track of which White Swan started where, and thus which ones are able to be captured.

Actually, this seems to be nonsense and could just as well be removed. A white swan moving from anywhere immediately gets replaced; since it no longer exists, there's no point in saying anything about its being subject to capture.

But this means that the opponent having more roses isn't necessarily good for you, right? It allows them to win by blossoming more readily (even if it allows you to win by checkmate more easily)?

I think the rules are nearly complete now, with one exception: pawns' two-step. Since they only appear later in the game, do they get one, or no? If they do, does it depend on their location and/or limited to their first move, and does en passant exist?

A White Swan coming from the first rank can be flipped over on the second rank, provided there is an empty square to land on. In that case, any White Swan landing on the second rank can be subject to capturing, but only after being flipped over and replaced by the corresponding pieces.

If an enemy piece lands on a square on the second rank, the White Swans from the first rank can capture that piece by a Pawn like move. If this happens, the White Swan will also have to be flipped over and replaced by the piece/pieces they represent, counting as a single move.

These paragraphs don't seem needed anymore; they're natural consequences of the main rule, right?

How are the pieces introduced to the board? Players alternate turns placing their pieces?

You don't need to separate the pawn and king placements; since they're forced, just place them all at once.

I would create a new page for the modern elephant.

I'd be fine with dropping the Elephant Link entry from this page, if others agree that applying the term elephant to the alfil is outdated (we can always keep a note in the text here and in the new FA page to help direct folks).

This has ended up without a description as I submitted the form in a rush due to some apparently bugged aspects of both logging in and the Submission form (I might describe those further in another comment); the metadata editing form I now have access to was very useful for setting this to be a Piececlopedia page and correctly assigning attribution, but it seems (and I think this has been noted before) it lacks a field for adjusting the Description (as opposed from the, distinct, What's New text); is there any way for me to do this?

Page descriptions can be edited from the editors' Edit Links page ([links]). You should generally modify the Primary link; non-Primary ones are used to display e.g. alternative names in index pages, but are excluded in searches with primarylinksonly=on.

The rules are not clear, and in a similar fashion to mathematichess: you assume too much of the reader to understand what you mean instead of what you write.

As in mathematichess, it would probably help to group the rules into logical clusters, instead of the more-narrative style that's here now.

On to specifics:

1. (Not a rule question) Fisher Random was "unsuccessful" at randomizing chess?? Arimaa is random!?
2. Using "Black Swan" as both piece and event is just asking for confusion; the thematic gain is not worth it IMO.
3. When multiple pieces are revealed, the owner gets to choose their placement?
4. Do swans flip/reveal immediately after any move, or do you get to choose??
   1. If the latter, do you have to on the 8th rank?
   2. If the latter, with multiple pieces revealing on 7th/8th rank, can you place a pawn on 8th? If so, does it promote immediately?

Since Black Swans are flipped over one by one, there is not the risk that the board might be jammed, considering that the occurence of Black Swans (events) is only 4 out of 20.

It might be extraordinarily unlikely, but I think it is possible. Might as well say what to do.

It's also, I think, miscategorized.

Not Usual Equipment, because you have to differentiate all the pawns.

Not dice: I don't think the use of dice in initial setup randomization counts (precedent?).

Not crossover: blending different chess variants doesn't count (or almost all of them would be in this category); it requires blending with some other, more different, game.

And what does "Price" mean?

@A.M.: The What's New page relies on page contents being updated, not just the index information. You can manually update the Last Modification date from the edit index information page.

This description doesn't fit:

smallish ads appear sneakily from bottom, going upward on my screen

and I've never seen such, and I expect they shouldn't be possible.

Here too I think it's important to at least say something about expected balance. Can ChessCraft provide automated playtesting?

The content of this item's Description should appear in the page content, maybe with some hints as to how such mates are accomplished.

Movement diagrams are fixed, but a separate issue: the 0th rank in the setup diagrams.

@FergusDuniho, I think I mentioned this in another comment (on a different page), but don't remember now where or whether a conclusion was reached.

I approve of the general idea. I think reviewing the votes for approvals will be important to avoid abuse, and IP address might not be sufficient. But sock-puppet control is hard; hopefully the nicheness of our site will make this less of a concern.

On 1, I'd suggest to keep more than just the last revision, but more-aggressive deletion (compared to the general revision deletion policy, whatever that turns out to be) would be fine. Furthermore, I think removing all the revisions upon approval would be reasonable.

Part of me would like a more nuanced notion of "member" and "contributor" for voting, but I don't have good ideas for it. I think about the Stack Exchange system of "reputation", but I don't think anything we track currently would efficiently represent familiarity with the site and/or chess variant expertise. And as long as editors continue to review approvals, we can handle things. (Should there be a sort of negative vote that members can contribute? "This rule is unclear, please revise before publishing"?) If we move forward with this, we might want to consider raising the publication bar if we see a lot of "low-confidence" votes, as an alternative to limiting voting ability.

The roll is obligatory, with three exceptions: if the move ends with a check, the piece reaches the last line as rook, bishop or knight, or if a pawn reaches the last line.

The first exclusion as written includes discovered check; is that intended?

The second exclusion sounds like it's there to prevent rolling a pawn and it being stuck; but maybe it fits better to force a roll, and if it lands pawn then the player immediately promotes (chooses a face)?

I've removed the Incomplete Information category. I don't think we include randomness in that definition, just deterministic information that's hidden (?).

This is an interesting theme.

Have you done any testing for balance?

1. Only one ghost can be in hand: do newly captured pieces disappear, or can you decide whether to replace the ghost in hand?
2. Ghosts are captured as normal?
3. Dropping a ghost from hand use a turn?
4. The five turn wait isn't clear to me. A freshly captured piece has to wait 5 turns before returning, but what do you mean by the last ghost part?

H.G. is an editor, see Who is Behind the Chess Variant Pages?. (The two junior editors aren't active AFAIK.) But again, as he has said in a number of comments, his focus is more programmatical, including the IDs and now Jocly.

There is just one review queue, which can be viewed by anyone. (Glancing now, it looks like the oldest page needing review (not "Uncreated" or having red text suggesting an editor comment without response) is from Oct 1, so just over three months right now.) Work other than reviewing on the other hand is up to editors' discretion, but my last post mostly summarizes that difference in focuses.

Edit: The 'Man and Beast 09' has already been released! Kind of weird.

M&B was released in 2008, and broken by a website change circa 2019. As I alluded in the previous comment, I am fixing a 5-year-old mistake on our part.

@FergusDuniho

This page has a broken diagram designer image; the URL

https://www.chessvariants.com/play/pbm/drawdiagram.php?
code=5--------------6-------------
2%7B%253%7D%7B%252%7D%7B%253%7D2------------
2%7B%252%7D%7B%251%7D%7B%251%7D%7B%252%7D2-----------
2%7B4%7D%7B%251%7D%7B1BI%7D%7B%251%7D%7B%253%7D2-----------
2%7B%252%7D%7B%251%7D%7B%251%7D%7B%252%7D2-2%7B%253%7D2------
2%7B%253%7D%7B%252%7D%7B%253%7D2-1%7B%252%7D2%7B%252%7D1------
6-%7B%253%7D2%7B%251%7D2%7B%253%7D------
5-2%7B%251%7D2%7B%251%7D2-----------
1%7B%252%7D2%7B_AS_RO%7D2%7B%252%7D1-----
2%7B%253%7D2-2%7B%251%7D2%7B%251%7D2-----
1%7B%252%7D2%7B%252%7D1-%7B%253%7D2%7B%251%7D2%7B%253%7D-----
%7B%253%7D1%7B%253%7D.%7B%253%7D1%7B%253%7D-1%7B%252%7D2%7B%252%7D1-----
2.2.2-2%7B%253%7D2-----
1%7B%252%7D%7B%253%7D%7B%251%7D%7B_JG_.ROQ%7D%7B%251%7D%7B%253%7D%7B%252%7D1-----------
2.2.2------------
%7B%253%7D1%7B%253%7D.%7B%253%7D1%7B%253%7D-------------
1%7B%252%7D2%7B%252%7D1--------------
2%7B%253%7D2-----
&cols=19&nocoordinates=on&set=alfaerie-many&shape=hhex&board=210.102.021.

(newlines added to prevent a long horizontal scroll) produces the error message

The color number 4 has not been assigned to a color. Make sure you assign color values to every color used.

Presumably this is something like {4}, but then this is another regression in the Designer, and we should seek out other pages that might have the issue. But there are lots of other examples on this page with circled numbers, maybe it's just missing a %? (Have I understood this correctly, that {%4} should produce a circled numeral 4, and that the brackets as well as the percent sign is being URL encoded, leading to the horrendous strings in the source here, like %7B%254%7D?

@FergusDuniho

I thought this was next on my list to fix the diagrams for (from a change in the diagram designer), but I'd already done the movement marker conversion. However, now I see that

1. two of the piece images don't show up, using the combination codes B!R and R!B. In the setup diagram these pieces use B.R and B.R instead, and they show up. Was there yet another change to the diagram designer for which we need to scan the database for broken images, or was this an old mistake?

2. in the setup diagram, the 0th rank displays offset (on my laptop); any idea what's wrong there?

Fergus, please change the font color for last actions being Notification comments to green (or omit them in favor of the last action being the edit itself).

I updated the royal push text to avoid the interpretation I had before. There are still two things left to explain about it:

1. "lower" pieces: can a king push a queen, or does this just mean royals vs non?
2. Can a queen push a piece two squares, if there is a joined territory?

Are you sure you mean the What's New text, and not the Item Description? The former only shows up a few places, most notably the What's New page, while the latter shows up in most index listings, content headers, etc. You can modify What's New text in the index information and while updating the text, but for now updating the Description has to be done by an editor.

I agree with most of this, but should spend some more time looking through the lists.

I don't like "stem" as the theme name, being too jargonic. I think two separate ones is fine, or even put math as a sub-theme of science. (This one will be huge depending on what exact criteria we place; board geometries are geometry, all 3d variants are math, ...)

I tend to think of "wargame" rather differently than "rpg". Wargame to me is more like Joe's variants, with lots of pieces moving in formations, while the descriptions here (esp. hit points, leveling up) sound more distictly like rpgs.

If you were to limit the revision deletion to groups of revisions by a single author within a certain timeframe (a week? a month?), would it lose most of the size savings? It would be nice, IMO, to keep a longer-term history of an article, while squashing a flurry of small changes into one larger one.

But also, saving the diffs instead, why would that break things? The deletion of a revision would just result in the squashing of the two diffs, right?

I had assumed the newly merged territory (d5+d6) would count the three black kings at d4, e5, and d7.

Also, the new territory created at e5 would give the white player even more points.

e5 isn't a territory at all, let alone for white.

There were two recent requests to be added to this page:

• https://www.chessvariants.com/index/listcomments.php?id=52460
• https://www.chessvariants.com/index/listcomments.php?id=52461

But this page is a hard-coded html file last updated in 2002. I suggest we add the themes here as Tags instead, as appropriate.

It is, but the link text is "3 to the 5", which is odd since the short description doesn't seem to match either. But I presume Gilman was the one who set these...

I can't think of any outstanding questions now. But the rules are disjuncted and conversational, making them harder to parse (for me). I was planning on taking a pass at editing when I next have the time.

Odd, I edited the content through the editors script, got no error, can see that a revision has been created, but the content is still empty and the revision shows no diff with the original.

We should be careful not to explode the list of categories, and keep tags in mind. (It would be nice to be able to sort the tag listing by count.)

The Topic Index has links to usual-equipment listings that need to be updated.

Can you provide the scoring for the example?

In particular, are c5 and d6 "joined"? It's not clear from the description what exactly that means, and what "control" means.

The game end also seems a little fuzzy; won't someone just refuse to move as soon as they're ahead at the beginning of their turn?

I do like the simplifications you've made so far.

Fergus, such a change should really be done after some time to discuss all the ramifications, and making all the code changes at once.

Diff-setup, diff-move, diff-capture all only really make sense in the context of usual-ish equipment, I think, along the lines of what you mentioned for setup.

I think Usual-Other should just become Usual; it's not necessarily "other" in the (vague) way the Other category is, it's just playable with the usual equipment without having fallen into one or more of the other Usual* categories.

Modest should be its own category. They will be a subset of Usual, at least until someone gives a convincing argument about a variant that should violate that pattern. I don't think it should be removed: it has useful signal, even if fuzzy, and many of the other categories are similarly fuzzy anyway (such is just the nature of categorizing [chess variants]).

I tried to publish this, but got

Modify Item
SQLSTATE[01000]: Warning: 1265 Data truncated for column 'Categories' at row 1

presumably something to do with the category changes?

I think the Usual Equipment categories don't belong here?

I agree (though I like 0,1,a,b,c,d,e for replacing "6" with something more special), but I can also see wanting something less symmetrical.

However, I'm still concerned about the coordinates inside each face in Bob's approach, and whether it's even well-defined. The four-tuples solves that.

Two of the games here are best seen as playing in full 3d space, just restricted to the outermost cells of a cube (Gilman's Empty Cube Chess calls it as such, while Judith's Cube-Surface Board Chesses doesn't seem to say it, but the merging of adjacent squares I think makes it so). The rest are actually on the surface (though several are incompletely described, and at least one is really just a circular chess board).

In the latter category, bishops aren't colorbound, and their passage through corners needs a decision. Playing on the surface also means all the pieces are more long-range. A rook can keep an opposing king confined to one face, patrolling the bordering cells from four other faces all at once from any such square (and sliding around to the other side if the king gets too close). Two rooks then force mate, guarding each other along that perimeter and performing the usual net on the king's face, just from its boundary.

The (solid) tesseract can be described as the set of points (x,y,z,w) with every coordinate between 0 and 1. The 2d faces then are characterized by setting exactly two coordinates to either 0 or 1 (and letting the other two coordinates range between 0 and 1).

Subdividing those faces into a 5x5 checkerboard, the centers of the squares are points with exactly two coordinates equal to 0 or 1 and the other two among {0.1, 0.3, 0.5, 0.7, 0.9}. Perhaps then a usable coordinate system just replaces those decimals by a,b,c,d,e respectively.

In that system, one rook path would start
01cc, 01dc, 01ec,
at which point the rook hits the edge point "011c" (not a playable square) and the two continuation paths must move off the extreme value of either of the first two coordinates:
, a11c, b11c, c11c, ...
, 0e1c, 0d1c, 0c1c, ...

Is there really anyone that uses the Utrecht Angel icon for a QN?

I'm not even familiar with any game that calls this piece an Angel.

https://www.chessvariants.com/link/angel-chess

(Huh, that page contains Ed's java applet of the game as a Related link, but not vice versa...??)

The ratings are averaged together (I believe only using the last rating per user), and that "average rating" can be used to sort in the database query. The average rating is also visible at the bottom of the comments section on each page, e.g. right now on this page it says:

Number of ratings: 3, Average rating: Average, Number of comments: 84

The ratings used to have an "Average" option; not sure when that got lost in the dropdown menu. But then Poor=1, Average=3?, Good=4, Excellent=5...was there another one for the remaining number of stars?

I think ratings (and the attached comments as a "Review") are a great more-granular-than-"favorite" scoring system, but aren't used enough (and apparently not clear enough). I would actually prefer more granularity, maybe a score out of 10; for one thing, there are more than 500 game pages with the Excellent average, so the database query doesn't show them all:
https://www.chessvariants.com/index/mainquery.php?type=Game&language=English&orderby=AvgRating&sortdescending=on

EDIT: there's a more recent rendition of the average ratings listing at
https://www.chessvariants.com/index/avgratings.php
and its comments mention the other missing rating, "BelowAverage"=2 stars.

The Dove can force mate with assistance from its king:
https://www.chessvariants.com/membergraphics/MSinteractive-diagrams/EGT.html?betza=fFvDbsNbAvHfG&name=Dove&img=bird

(The EGT appears to be broken right now, but that should work.)

@H.G., the EGT (linked in the Notes) is acting funny, producing a starting array that is horizontally symmetric.

You mean the ItemID, used e.g. in the URL of the page? That's rather harder to change, since it is used to connect the page content, Comments, Favorites, Links, Tags. I have done it before, manually. @Fergus, would utilizing the RelocatedPages table instead work for this?

Neat! I have no idea what these well-positioned but not much better than FIDE setups might look like. I wonder if over enough plays certain themes of setups would emerge.

Interesting! This will be somewhat unusual in the "Dice" category for not also being non-deterministic.

I wonder about the enforcement of orientation after the move. It enforces that a piece moving from one square in a particular direction will always turn into the same piece (OK, knight choice aside); is that convenience worth the additional rule? Maybe.

I doubt the setup diagram would be helpful, since the tops of the dice will just be the standard layout. But a few movement examples would be nice.

You had this marked with three "Usual Equipment" (sub)categories, which it certainly isn't?

Movement diagrams would be helpful for visitors less familiar with adjectives like "half" and "narrow".

@Jean-Louis, "query parameter" is being used here in a technical sense, as the part of a URL after the ?. We can certainly add a form as in other places on the site; but for now, you can try e.g.

https://www.chessvariants.com/index/favorites.php?sort=score&limit=10

@Fergus, somewhere in your changes you've hidden those hyphenated-instead-of-concatenated favorited items; but those favorites still exist!

I believe that's the current behavior for an incorrect password, but could be any other error in logging in...the actual login page doesn't seem to produce a useful message either. @Fergus?

I appear to have the movement diagrams here (and earlier M&B entries) fixed now. There was a further complication to my process on this one: some of the piece names had periods not at the beginning.

The diagram for the Fearless looks out of place, but in a way that I can't imagine was my mistake. Anybody with more patience than I: care to explain why it has the retreating two orthogonal squares?

OK, this page is done; please let me know if anything remains to fix here.

(There were two diagrams here too in which a piece name started with a period and the surrounding brackets had been URL-encoded, which made the first pass of the regex replacement convert too much. It was easy enough to spot and fix. I can't just begin the process by converting the URL-encoded %7B and %7D back to brackets, because then the main regex will stop too soon. Maybe I can adjust the main regex to exclude matches to %7B.?)

Your setup image needs to be uploaded here and then included in the page. Go to the Edit menu to upload, then copy the resulting URL into the page.

Why the weird "modern" naming (alternatives)? A few specifics:

Modern piece created by me, but there isn’t its full movement, so here it can look as existing one.

This doesn't seem to add much? What full movement, and what existing piece?

Modern piece independently created by several people, i. e. by me.

I think you're misusing "i.e." here; you mean "including"?

Fairy piece, there with Shogi & Chu Shogi origins, so all names are useful there. It moves only one space in any direction.

But it's also the restriction of the wazir and the man and ... to 1D. Again I don't think the comparison to higher dimensions helps here, just say "moves one square" and get to the king-swapping ability.

I would transpose the diagram, so that it fits better on a screen, except that I worry about mobile screens. Hmm.

I'm not very familiar with the other 1D games we have here, I think it would be worth contrasting them a bit.

I read this a while back, wasn't clear on the rules, but never commented to ask. Going over it again, I can't bring myself to try to figure it all out. While part of that is me being tired, I think another part of it will affect other readers. Please try to simplify the presentation of the rules, and reduce the use of all-caps. Possibly a shorter word than "superimposed", and ones that are more varied than "superimposed on" and "superimposed upon" would also help.

Playing on the 2d surface has the nicety of rook lines still actually restricting the enemy king into one side or the other. What does mating material look like here?

I'm not sure I understand the question.

For starters, can two rooks force mate against a lone king? If the corners weren't removed I think they could. Two queens probably suffice? Maybe it's best to see if this has been resolved on the surface of a 3D cube first...

But yes, the Bishops are not colorbound, strictly speaking. It's not possible, on a cube's corner (much less a tesseract's), to have a colorbound check pattern. I didn't realize that when I put four on each side, but then I decided that it wasn't that big of a deal; switching is a trick that requires rounding a corner.

I'd like to try to work out a bishop's path (and number of attacked squares) partly for this reason: it might be possible to effectively wrap around corners by going the long-long way around?...

Clockwise as seen by the player on the 2D board.

But I'm not sure that makes sense, in the same way that clockwise isn't unambiguous in 3D: it depends on from which side of the surface you're looking.

Can I assume the corner cells are deleted so you don't have to work out what to do with the diagonals there?

If a pawn or spear find themselves on an Open face, there are two (or all four, on faces 6 and 19!) directions that are "toward" the enemy Home face; how do they move then?

Playing on the 2d surface has the nicety of rook lines still actually restricting the enemy king into one side or the other. What does mating material look like here?

I tried to work out (but without paper) how many squares a rook attacks on an empty board. There are 12 faces that it reaches in each direction, but those overlap, I think four faces in common? So it should be 5*12*2-4-1=115 (that last being the rook's current cell)? What about the bishop, or nightrider?... Oh, I guess bishops aren't colorbound?

Is there a reasonable way to flatten this for displaying on a table/screen? (I suspect not, because of the forking of paths.)

Does the setup section's use of "clockwise" actually make sense?

Why alternate ordinary and berolina pawns? Doesn't that hurt pawn structures? (Does it just not matter on this wacky board?)

The counterpart BWND is obviously a highly-Br[w]ain'ed rodent.

That's so tortured a stretch, I'll help those unfamiliar: https://en.wikipedia.org/wiki/Pinky_and_the_Brain

I added another gorgon image.

Either that wasn't successful, or something funny is going on: I only see the figurine image.

I don't think these images add much to these pages. I'd rather see them in a separate article. To the points about different piece sets/themes, you could aggregate similarly themed examples into a "set" article.

Done.

@editors, I think the Primary Group ID for rules pages and for Courier preset landing pages is being generated differently, the former without hyphens and the latter with them in place of spaces. Reconciling that would make this work more automatically more often?

Yikes, I'm sorry you've had to go through so much additional effort Jean-Louis.

Are you familiar with the version history tool Fergus added? It's under the Edit menu, and lets you view, compare, and restore previous versions of the page. It's sometimes a little fickle, though that might've just been related to a server move?...

Anyway, keeping local copies is probably a good idea.

We should restart a discussion around the text entry editor we're using, CKEditor. Switching to a new text editor would be painful, but we might be at the point to do it, if we can find something suitable. Or would it be enough to switch to Markdown as the default format? Or have HTML format work without CKEditor (keeping that just for WYSIWYG format)?