Simon blogs

[mobius]

Gin Rummy (and many variants of Rummy) was common here in U.S. My favorite to play was so-called "Contract
Rummy". Works similar to 'plain' Rummy except as follows;

1) there are a set number of rounds, the version I played always had 5.
2) each hand has a specific assortment of hands you need to make before you can discard, to go out (get rid of all cards, thus ending the round). 2 sets or 2 runs or 1 run and 1 set etc. I can't remember the exact rules for each round anymore except that it got  more difficult as the game progressed.
3) I don't remember if this was singular to this particular game or not but a player could "buy" a (1) card from another player by taking another card from the pile, but only 1 per turn or something like that.
4) Every round played like normal rummy (players lay their cards down whenever and first person to throw their last card away ends the round). But the final round functioned like Gin Rummy; in order to lay out you have to have all the cards in your hand fit the contract, thus the first person to lay out also ends the round and everyone else is stuck with all the cards in their hand.

The rest I think functions mostly like Rummy. After the final round the player with most points wins.

There's also this game called Rummikub; which I never played but looks like a mix between Rummy and dominoes... no idea how it works.
https://en.wikipedia.org/wiki/Rummikub

here's another Yatzhee like game I've not really played but heard of:

https://en.wikipedia.org/wiki/Farkle

[Dominator_101]

I've mostly played what I think is typically called Rummy 500, which rather than solely holding cards in your hand you instead play your sets visibly, and can play cards on the other player's sets as well. It also involves being able to pick up one or many cards from the discard pile. I enjoy the additional strategical elements that it involves. In general I prefer this version to Gin, but still enjoy both.

Rummikub is pretty fun. It kind of plays like Rummy (you're trying to make sets of at least three matching/three in a row), but one of the things I like most about it is that you can mix and match tiles that are on the board as long as everything ends up a valid set. So for example if I have a a red five, blue five, and green two, and the board has a green 3-4-5 set, I can play my 2 at the front to make it 2-3-4, then yoink the 5 to make a 5-5-5 set as well. Sometimes I'd end up rearranging half the board just to get one or two tiles out of my hand.

[Simon]

Right, the appeal of Gin is that it's a potentially reasonable 2-player game. I've tried it, and it felt like it was still lots of luck, but possibly there is indeed more depth.

The other Rummies are better with more players than with 2.

Before, to me, Rummikub had always looked like an arbitrary Rummy-style ruleset that merely sold well because of its nice tiles. But you sound as if the ruleset holds up by itself, and it should be playable with two standard decks with standard jokers. I'll see when I get the chance to try it.

Thanks for the ideas!

-- Simon

[Simon]

Esoterics with Simon

I've learned the mate with bishop and knight against a naked king.

There's an interesting limit to how well you can study it purely with a computer. The defending computer wants to prolong the mating sequence. Its favorite way to prolong the B+N endgame is to flee with its king into the wrong corner, from where it takes the most moves to mate. But from that wrong corner, the remaining moves are easy: The attacker can apply a largely fixed pattern.

The computer would pose a bigger problem to the human by avoiding starting points to common human patterns, even if it allowed shorter mates that an engine might find.

Edit Simon 2023-07-08: I've split all C++ posts into C++ with Simon.

-- Simon

[Simon]

User Intercardface

No company makes sufficiently good playing cards. Some designs come closer to the ideal, but none check all the boxes. It shouldn't be so hard, but apparently it is.



The four suits should be in four clearly different colors. More generally, in a 2-axis deck of suits × ranks, the n suits should be in n colors, and the color distribution should solve the Christmas tree problem.

We can certainly argue if spades or clubs should be green, or whether the diamons should be blue or orange. It's hard. If you pick orange, like in the Skat Turnierbild (left image above), your orange is either too similar to the red hearts, or too light, failing to contrast from the white background. If you pick blue, as in the Copag 4-color deck (right picture above, with the four aces), you're prone to choosing too similar a hue for the blue diamonds and green clubs.

Nobody wants a 1-color 4-suit deck with black hearts and diamonds. Why should anybody want a 2-color 4-suit deck? This is such a basic UI principle: Make different things look different.



Well, some solitaire games have rules for packing red suits only on black suits, hmm ... the answer must then be to still have 4 colors, but choose more similar colors for hearts and diamonds, e.g., red and orange, and two darker colors for spades and clubs, e.g., black and a moderately dark green.

Above picture is my unpublished Turnierbild Dark, a 4-color dark-mode mod of a card deck for PySolFC. I wanted a 4-color deck and chose the Skat Turnierbild colors, mainly because I'm used to them, and also because orange diamonds and red hearts are reasonably close. But the spades and hearts look too similar in lightness for me. With this deck, it's harder than necessary to play games with alternating red-black packing. It's better for games that pack in suit or regardless of suit.

Hard to say how to improve it, given that I also want the dark background. Maybe the background should have one hue for clubs and spades, and a different hue for hearts and diamonds? We're getting really experimental here! But it's UI with Simon, what else do you wish for.



Index size. The index is what's in the corner: A number/letter and one suit symbol. Regular index (left picture) is too small, even while the card is in your hand. Jumbo index (right picture) is much better. Maybe it's a tad too big already, and the middle is ideal? Hard to tell. We want to fan it in hand and also see it clearly on the table.

If I'd have to choose right now, I'd pick that middle size. It's sometimes called Blackjack index, but that term is hard to websearch.



This is Magnum index, offering a logical extreme to the size problem. The entire card is nothing but indices. In hand, you see the good Blackjack-sized index. On the table, you see the big centerpiece.

It's an interesting idea, I've never tried it, and I'm neutral on it. But I've never seen it with four-corner indices (i.e., all four corners feature a Blackjack-sized index), which brings us to ...

Four indices. American cards put indices in two corners only. Does no American ever fan cards in hand in the wrong direction? European cards always have indices in all four corners. You can fan them both ways. I'd estimate that 20 percent of Europeans fan cards the uncommon way, i.e., the direction in which they wouldn't see indices on American cards.

Always put four indices! This is really easy to accomplish and everybody should always get it right. One counterargument to four indices is clutter, but the solution to clutter is a good box, not to restrict your fanning-in-hand direction. Which brings us to ...



The box. Surround the interior (everything that is not a corner index) with a box. Here, the eight of spades has no box, but the ace of hearts has this box. Boxes are common for court cards in American-style decks, but you should do it on spot cards, too. Do it everywhere!

Some Bridge decks have a yellowish hue in the box. That makes it even better to distinguish the in-box design from the out-of-box corner indices. I like it.

The box is the correct solution to clutter. Then you can have big indices in all four corners and it will still look good.

Card width: Practically always you want narrow cards; a common size is Bridge-sized. Poker-sized cards are wider and harder to fan.

The community cards in Texas Hold'em might look better on the table with Poker-sized cards, but that alone is a weak reason for such an unwieldy card format. If you want readability on the table, you want clearly readable, big, box-separated indices. Card width is secondary, but if you get big indices with your wide cards, that may be a good reason. For Texas Hold'em, consider the 4-color Copag, see the image of the four aces at top of this post.

The game of Tarot makes you fan 18 cards in your hand, and, as a result, Tarot cards are even taller than Bridge-sized; they're about 1:2. Not sure about that size for general use.



This is one of the best designs of physical playing cards: Desjgn Classic Circle.

• Four colors, even though clubs and diamonds are slightly similar,
  
• a box with light yellow background,
  
• Jumbo index, nice and big, maybe 5 % too big,
  
• Bridge-sized, not too wide, not too narrow.
  

Downsides:

• Cards have only 2 indices, not 4 indices in all 4 corners.
  
• The deck is surprisingly expensive.
  

-- Simon

[Silken Healer]

Thank you for sharing PySolFC! Like you, I am into card game programs a lot and this is the best one by far! It brings back the good card games on Windows and not the bloated versions we have with adverts like now. Before this, the one I played was called "xmsol" if you want to check it out https://sourceforge.net/projects/pysolfc/ but I think PySolFC makes it obsolete.

Do you mind publishing that the full dark-mode pack? In my personal preference, your design is probably my favorite playing card design ever both from the perspective of talking about the card designs for the topic and just what I genuinely what to use in playing PySolFC on my own lol.

Sorry I can't contribute to the blog much in the terms of the actual linear algebra part as I have not learned linear algebra in school yet.

[Simon]

The only missing piece to make it releasable is anti-aliasing and darkening the queen of spades. I've already reviewed how to install custom cardsets in PySolFC 2.21.0. It looks like I can get the set to run in a current version, nice.

I'll see how far I get tonight with the queen, and will post the dark cards here in 0-2 days.

-- Simon

[Silken Healer]

Thanks

[Dominator_101]

Four indices. American cards put indices in two corners only. Does no American ever fan cards in hand in the wrong direction?

To me this just seems like a self-fulfilling prophecy: why would anyone here learn to fan their cards the other way if you couldn't actually see what the cards were?

It actually took me a minute to figure out what fanning them the other way meant, at first I was wondering if you were all psychos that held the cards from the top or something. Didn't realize it meant having the left card on top rather than the right. Don't think I've ever even considered holding them that way. Just thinking about it that was feels like it'd be clunkier just to hold/fan, but that's likely just a muscle memory thing.

[Simon]

I've released my Turnierbild Dark for PySolFC.

figure out what fanning them the other way meant
it meant having the left card on top rather than the right.

Correct, with the uncommon way of fanning, I mean: The fan in hand (regardless of left or right hand) has the leftmost card fully exposed to the holder [_]]]]] instead of having the rightmost card exposed [[[[[_]. Having the leftmost card fully exposed hides incides on 2-index playing cards.

There is a nice rabbit hole here with handedness (I don't think it matters much), muscle memory (matters a lot) and style of game: Trick-taking games and many Rummy variants require you to pull cards from the middle of 10+ card fans. But 5-card-draw Poker fans have fewer cards and change rarely.

I agree that it's a chicken-and-egg problem. Still, it's inexpensive to put 4 indices, and it can be largely clutter-free. Given how many Europeans fan cards the uncommon way, why curb your market share?

you were all psychos that held the cards from the top or something.

This is wonderfully absurd. Thanks for sharing!

-- Simon

[Simon]

I filed a bug against PySolFC (Archway doesn't let you pick arbitrary suit from arch). It dovetailed into UI design. Equally fun as playing!

I've always noticed this in Lix and NL: I play for a while, then get the urge to continue development. The vision of playing an improved version, or allowing others to play an improved version, can be more entertaining than playing the existing version. As a countermeasure, I had scheduled singleplayer livestreams to finally play through Lix seriously. And it's also very enjoyable, I should do that more again. After lemforum, there is NepsterLix, ClamLix, and Rubix's set.

I used to joke how my ideal board gaming night would go: Before the night, choose the game, email everybody a copy of the rules, FAQs, errarta, ..., and expect everybody to arrive well-prepared. Then, on the night, don't play the game, but hold a conference about bugs and corner cases in the rules. The winner is who found the best bugs, or the most bugs, or whose bugs eventually make the maintainers board game authors release an updated rulebook or errata.

-- Simon

[Simon]

Docking Hell

Gimp (the drawing program) has dockable subwindows. Those dockable windows are a serious contender for the prize for the most unintuitive UI. They offer the following nasty trap:

How do you re-dock a dialog that you undocked by accident?

this is maddening. i can't get any help elsewhere so hopefully someone here can get me out of this hell.

Source

Yes! Hell it is!

try as I might I cannot manage to redock it
It just refuses to be fixed like the dialogue in "Layers/Brushes" on the right.
P.S. My awkward workaroung so far is to close gimp, delete the ".gimp-2.8" folder and restart gimp again; and loose the previous settings Sad

Source

Resign, close the program, and nuke your settings.

That's a whole level worse than anything we complain about in our own software. Sure, we had the occasional NeoLemmix user cancel the replay by closing NL, then reopening NL. But never anybody had to erase the settings or reinstall the program into a fresh directory. Maybe Lix's resolution options have more lock-out potential than anything in NL. I don't even know if it's possible; it might be possible to willfully enter garbage and still somehow avoid the automatic fallback of 640x480.

Anyway, back to Gimp. When your settings directory on hard disk is more discoverable than your drag-and-drop GUI ... you've written quite some drag-and-drop UI.

You can right-click the subwindow and you'll find a checkbox option "Lock tab to dock". It's unchecked. Can you check it? Yes. Does that re-dock the subwindow? No, that would be too easy. This option prevents accidental undocking, but that's only useful before you've undocked your tab. Now it's too late, and the option won't do anything until you've re-docked, which you don't know how to do.

Solution: Drag-and-drop the tab, not the window. The window (which your system's window manager manages), called, e.g., Tool Options: You can drag this around, too, for sure, but it will never dock. You must drag the single tab from inside the window, and the tab is also called Tool Options.

When you drag the tab (not the window), suddenly, Gimp plays ball, rolls out the red carpet, and offers you nice grid previews for docking your fine dialog. Now it looks easy and well-implemented. It feels so nice that, certainly, nobody on the internet will now understand how this comfortable UI can be terrible.

-- Simon

[Dullstar]

One thing I really appreciate about your approach to Lix is that you will look at something that people find unintuitive, and say that the fact that it's unintuitive is a problem. There's definitely a lot of people who write off those sorts of concerns as "It's working as intended." [often implying: "It's not broken or badly designed; you're just stupid."]

[Simon]

Thanks! That means a lot to me.

Right, the designer's task is to make the user guess right. It's hard to influence the guess directly; the users bring many expectations from other applications than ours. If it's working as intended, but the users constantly guess wrong, then the design has a problem, even if it's intended.

Another view, more tongue-in-cheek: Design is de-signing, i.e., removing signs, or at least removing the need for signs. If you can guess how a program works without reading manuals/signs/..., it's well de-signed.

It's rare that an application is so useful that it can force arbitrary UI on the user. git comes to mind. But in git, the mental model of the commit graph should come first; the UI is merely a front-end to modify the graph. (Edit: Now I wrote this and will scare WillLem off git forever? There are nice GUI clients for git that do everything except rare advanced things.)

-- Simon

[WillLem]

Now I wrote this and will scare WillLem off git forever?

I use both Github and Bitbucket/Sourcetree, both are fine. Bitbucket and Sourcetree are nicer. Github is often remarkably unintuitive, but I think that most programmers secretly enjoy that

For example, people who do most things via the command line tend to make it very clear that that's their preference. They want the kudos! Fair enough.

Give me a nice, clean, friendly GUI with big buttons and sliders any day.