Simon blogs

[Proxima]

"odor cart":  sounds like a bad translation perhaps?  I believe with languages like Japanese if you attempt to only translate individual Kanji characters in isolation and ignoring contexts, you can sometimes end up with results that have almost nothing to do with the actual words or phrase in question.  Also, doesn't some Kanji characters have multiple very different meanings (almost like homophones but actually written exactly the same way as well?).  Do we know if "odor cart" isn't just creative cherry-picking on a particular reading of the Kanji in question, even though it isn't actually the intended/correct reading being used wrt the name of that particular Shogi piece?

It's just Simon poking fun at the weird way the Japanese language works. 香車 means "incense chariot", but 香 by itself can mean "smell".

[ccexplore]

Ah okay, so cherry-picking it is. Unclear from OP whether it was tongue-in-cheek vs serious.

[ccexplore]

Pieces should be colored by type.

I'm actually a little surprised by this.  Wouldn't having actually different shapes like Western chess pieces do, be even clearer for telling apart piece types?

Is the color by type idea mainly to compensate for having to recognize Kanjis?  The pictorial approach seems mostly sufficient to me without also having different colors, though maybe the pictures used could be slightly better in some cases.

[Simon]

You gave up your nap? Clearly you don't have enough pets.

Yeah, I should not sacrifice naps. Watching the chess championship was a rare exception. In general, I shun television and try to avoid whatever imposes similar artifical schedules on my life.

Pets cost a lot of time. >_>;; I have too many hobbies already and pets are like an extra hobby.

I admit that I have gedankenexperimented with an African porcupine as a pet. Must build lots of mutual trust before you can safely cuddle.

frustrations I had was that the packages in the repositories were frequently out of date.

Yes, I had the same frustration in Debian. It's slightly better in Ubuntu. But for two years now, I've been very happy with Arch's rolling releases to stay cutting-edge.

constant frustrations with NeoLemmix and Lix is updates.

Right, this is the main problem.

Probably, many Windows people shun updates because they're so hard unless automated.

I shun installing new software that I don't clearly need, but I want to update the software that I decide to install.

"odor cart":  sounds like a bad translation perhaps?
attempt to only translate individual Kanji characters in isolation
results that have almost nothing to do with the actual words

It's just Simon poking fun at the weird way the Japanese language works. 香車 means "incense chariot", but 香 by itself can mean "smell".

Right. "Odor cart" came from a friend, he translated the kanjis separately. Even with the correct translation "incense chariot", it's still a flowery name for my taste -- incense has little to do with war, the theme of chess-likes.

Well, same criticism applies to bishop from chess.

Both stabs were intended: The floweriness of the name, and how Japanese connects unrelated kanji to give every single one a different meaning.

Pieces should be colored by type.

Wouldn't having actually different shapes like Western chess pieces do, be even clearer for telling apart piece types?

Two Shogi rules:

• When you capture a piece, the captured piece goes to your personal reserve. You may drop these pieces under your control back onto the board.
  
• When your piece moves far enough ahead, you may promote the piece, i.e., replace it with a more powerful piece after its move. Already-promoted pieces cannot promote again. Promotion is permanent until the piece is captured; then, its unpromoted type goes to the opponent's reserve.
  

This restricts the physical design. Standard design here is two-faced pieces with a pointy side. To show control, orient the piece to point away from its controller. To show promotion, flip the piece bottom-side-up, revealing the kanjis for the promoted type on the original piece's bottom side.

Color-coding by player is not feasible with two-faced pieces because color cannot change during capture.

Shape-coding by type is hard because the piece must rest physically stable on the game board when flipped. Maybe design 3-D shapes that are merely ((2-D shape per type) x unit interval).

Or design a four-faced piece (e.g., some variant of a cube) to have symbols for all four states: player A or B, unpromoted or promoted. But hm, that object will probably show too many faces on its left/right outer sides by accident when it's not meticulously aligned straight...

Is the color by type idea mainly to compensate for having to recognize Kanjis?  The pictorial approach seems mostly sufficient to me without also having different colors, though maybe the pictures used could be slightly better in some cases.

I'm concerned about understanding the board at a glance. Chess, Shogi, Go, ..., they're all about pattern recognition. Anything that makes glancing hard strikes me as contrary to the game design.

Letters, and thus kanji, are harder to recognize than symbols or shapes.

Colors alone can already be problematic (color blindness, weak conventions for the meanings). Colors are best to further differentiate shapes. It's best when there is only a handful of different types to distinguish:



-- Simon

[nin10doadict]

Pets are expensive and time consuming, yes. They are nice but they are a big commitment.

As an avid Tetris Attack player, I will say that color is the main way you will differentiate the blocks in your mind when you get good at the game. The fact that there are only 4 different shades of gray in the GameBoy version makes it much harder to play, as they had to try to differentiate the blocks more by redesigning the patterns on them.
I do wonder if play would suffer if we had different color blocks but they were all the same shape? Without any way to test it myself I can't say for sure, but I think it might.

I was unaware that captured pieces could be used by the player that took them in Shogi. Neat idea, but indeed restrictive to the piece design.

[ccexplore]

Even with the correct translation "incense chariot", it's still a flowery name for my taste -- incense has little to do with war, the theme of chess-likes.

Yeah, that was curious enough for me also, that I had to check out what Wikipedia has to say.  Turns out the "incense" part apparently originated from Buddhism at some point in the history of the game in Japan (emphasis below added by me):

According to Kōji Shimizu, chief researcher at the Archaeological Institute of Kashihara, Nara Prefecture, the names of the Heian shogi pieces keep those of chaturanga (general, elephant, horse, chariot and soldier), and add to them the five treasures of Buddhism (jade, gold, silver, katsura tree, and incense).

So I guess it was a historical/cultural artifact that we wound up with incense chariot. Gold, silver and jade are also used in some of the other Shogi pieces' Kanji names, but I guess at least we westerners would be more used to also reading them in some contexts more as rankings rather than the literal metals/minerals, due to our tradition of gold/silver/bronze representing #1-3.

---------

Also thanks for clarifying with those Shogi rules, that explains a lot and indeed makes it sensible to complement Shogi piece types with color in addition to the traditional Kanji.  I'll assume that traditionally the native Japanese figured if you are educated enough in Japan to know how to play Shogi, you are expected/assumed to have no problem proficiently recognize the comparatively few Kanjis used on the pieces (given IIRC from somewhere I read or heard, Japanese adults are expected to know something like some hundreds or thousands [!] of Kanji characters for proficiency in general reading comprehension).  Not saying this justifies not having better/additional means of differentiating types, merely that the natives probably didn't even think it was an issue.

[namida]

(given IIRC from somewhere I read or heard, Japanese adults are expected to know something like some hundreds or thousands [!] of Kanji characters for proficiency in general reading comprehension)

Knowledge of around 2000 kanji would be needed for a "normal" adult level of proficiency, I believe. Far more than that exist, but the obscure ones are generally written with furigana (ie: the reading of the kanji written above / alongside it in hiragana, which is another Japanese script that's more comparable to the English alphabet than it is to Chinese characters).

I've probably gotten a bit rusty nowdays, but at one point I was able to read about 400 and write about 150. Even that point was hard enough to read; although I'd guess it's probably not so bad when you're reading and writing them on a daily basis as a normal part of life.

[Simon]

Start by feature-bloating

The quote by Antoine de Saint-Exupery: Perfection [in design] is achieved not when there is nothing more to add, but when there is nothing left to take away.

The corollary is: To make a new program, system, game, programming language, ..., you must bloat the project with features first. Not merely brainstorm to think about features. You must implement the features at the cost of your time and resources, and then experience the features during normal use.

And then be courageous to remove some features again, and convince the userbase that the resulting project is better. Or start a project from scratch if the previous design is good enough to hold on its own. The decision to adapt or start from scratch is hard.



I've been far too hesitant applying this wisdom (bloat to try) to my own projects.

I'm not busy with Lix these weeks, pursuing life & other interests. But certainly, I will return, as I always have for ~13 years. I should work on the Lix physics changes that have been on the backburner for 1.5 years. Either 0.10 or 0.11 should be an unstable version where we test a ton, and where we change physics even with patch releases (0.10.5 -> 0.10.6). Then 0.12 will be the next stable that should ideally hold for at least 1 year, my rough expectation for stable physics.

Any details are best left for a strategic post in Lix in the future. The current physics, i.e., the 0.9 physics, will remain stable for several months.

-- Simon

[Simon]

Orange juice

Hotels serve breakfast, and in Germany, they offer a lovely variety of Brötchen (bread rolls), Aufschnitt (slices of cheese, salami, ham, ...), many cereals, coffee, tea, ... and juice. Nearly every time, there are two choices of juice.



One of the two juices is always orange juice. Everybody wants orange juice. It's pure sugar, some vitamins, fresh, pure in taste, kickstarts your economy circulatory system, goes with everything, it's perfect. It's exactly what you want in the morning to be fit for a taxing day.

The other juice option varies between hotels. Sometimes, it's apple juice. Sometimes, it's multivitamin juice. Apple juice tastes pale in comparison, and multivitamin juice is weird. People rarely want either such juice on its own. They accept it grudgingly when the dispenser for orange juice is already empty. It's the consolation prize for when you're too late at the breakfast and the clerk hasn't yet refilled the orange juice.

Now assume you're a hotel administrator. Your job is to order food and goods to keep the place running. Each noon, you observe how much orange juice and how much apple juice has been consumed that morning. You see that they're about 50:50. Will you conclude that they're equally wanted? Or will you conclude that your clerks are too lazy to check and refill the orange juice dispenser frequently enough?

Nice psychological fallacy. I'm not sure if this already has a name: To wrongly deduce from demand of the second-best good that the second-best good is worth more.



Writing prompts

There are these fancy forum games or level review threads: You pick a topic, say something nice about it, and say something bad about it. I feel like such a writing prompt has a chance to, either on the nice section or on the bad section, generate reviews that are too short and obvious. That's not what the prompt's designer intended; they intended neutral reviews.

The prompt should rather be to pick something you love, and write a one-sided negative review on it, review all its terrible features, wrong design choices, ... With lots of detail. Or pick something you don't like or don't care for, and praise its strong points. (If the topic is prefixed, such as the next level to be reviewed, you must honestly decide whether you like that level or not, and skew your review the opposite direction.)

For clarity, because such a writing prompt is uncommon, the first sentence should explain that the review is deliberately skewed against the writer's overall opinion. It's also good to have several people write on the same prompt, to produce at least one review of each skew.

Maybe I should review some computer/board games like this. Or write nice things about some rant topic? Look how convenient all the white webpages are: I can install a single browser plugin that inverts all colors, and it'll turn every website black. Were many websites nastily dark already and easy on the eyes, then I'd have far more hassle adjusting the plugin to each site. And if web designers decide to decorate their website with token smiling people that do nothing related at all, the color inverter will give them cyan skin, they'll look like Frankenstein's monster. Allows me to focus on the facts in the text.

Very hard to not fall into irony. <_< Irony shouldn't be the point of the prompt.

-- Simon

[Simon]

Game mechanic ideas

These are very loose ideas.

Three piles. The game consists only of playing cards. Each player has their cards distributed among 3 personal face-down piles. Each turn, you may pick up exactly one of your piles, and do something with the cards: Maybe they're action cards, maybe they're worth points only in certain combinations, maybe you can redistribute them among the piles, ...

The catch is that you cannot pick the same pile twice in a row. If you put all your good cards in one pile, you can only use them in even-numbered turns, and your opponents will know exactly that you can't do anything at all in odd-numbered turns. If you distribute your good cards only among two of three piles, you can play good stuff every turn, but your opponents will know exactly when the stuff is coming. To keep your options open, you'd have to use all three piles.



Neighbors in a circle. This needs at least 5 people, and ideally, exactly 5. They're sitting in a circle. You're in a weak alliance with your immediate neighbors: Maybe for each pair (one pair is you + left neighbor, another is you + right neighbor), each member of the pair scores based on something that the pair constructed during the game. Thus, you'll be happy if your neighbor does well, but you won't help them unconditionally.

The catch is that you can attack opponents who sit exactly 2 seats away from you. If you're player C during player order ...-A-B-C-D-E-F..., then players A and E aren't your neighbors, therefore you have an honest interest in harming them.

Assume you (C) want to attack A. B will discourage you from attacking A because B is A's neighbor and has a moderate interest in A doing well. B will suggest that you attack E instead. Then D will be unhappy and encourage you in your original plan to attack A. Then B and D will get into an argument. When D's turn comes, he will be happy to attack B, which makes E and F very happy, etc.

It is desirable to have an odd number of people in the circle: Otherwise, there are two disjoint sets of people who get into arguments only with other members of the same set.

-- Simon

[kieranmillar]

Three piles. The game consists only of playing cards. Each player has their cards distributed among 3 personal face-down piles. Each turn, you may pick up exactly one of your piles, and do something with the cards: Maybe they're action cards, maybe they're worth points only in certain combinations, maybe you can redistribute them among the piles, ...

The catch is that you cannot pick the same pile twice in a row. If you put all your good cards in one pile, you can only use them in even-numbered turns, and your opponents will know exactly that you can't do anything at all in odd-numbered turns. If you distribute your good cards only among two of three piles, you can play good stuff every turn, but your opponents will know exactly when the stuff is coming. To keep your options open, you'd have to use all three piles.

If you ever get the chance, try the board game Mombasa, which uses this. You pick 3 cards from your hand to make up your resources to use that turn, and at the end of the turn, each one goes into a separate discard pile, you may then pick up only one of these discard piles to replenish your hand. So as turns progress the piles grow by necessity and the resources you want are split out between them.

Neighbors in a circle. This needs at least 5 people, and ideally, exactly 5. They're sitting in a circle. You're in a weak alliance with your immediate neighbors: Maybe for each pair (one pair is you + left neighbor, another is you + right neighbor), each member of the pair scores based on something that the pair constructed during the game. Thus, you'll be happy if your neighbor does well, but you won't help them unconditionally.

The catch is that you can attack opponents who sit exactly 2 seats away from you. If you're player C during player order ...-A-B-C-D-E-F..., then players A and E aren't your neighbors, therefore you have an honest interest in harming them.

Assume you (C) want to attack A. B will discourage you from attacking A because B is A's neighbor and has a moderate interest in A doing well. B will suggest that you attack E instead. Then D will be unhappy and encourage you in your original plan to attack A. Then B and D will get into an argument. When D's turn comes, he will be happy to attack B, which makes E and F very happy, etc.

It is desirable to have an odd number of people in the circle: Otherwise, there are two disjoint sets of people who get into arguments only with other members of the same set.

This mechanic of working with your direct neighbours appears in two board games I know of. The most well known is 7 Wonders, which is a card drafting game, everyone gets 7 cards, picks one, passes the hand clockwise, and keeps going until no cards remain, and you try to build the best results out of this. You can buy resources from your direct neighbours if you didn't get the ones you need, and there are some other things too that care only about your direct neighbours. Its well known because it scales from 3 - 7 players effortlessly without extending the length of the game.

Perhaps closer to your idea is a board game called Between Two Cities. Each player builds two cities, each in combination with a different neighbour. Each turn you pick two tiles and add one to one city and the other to the other city, and these tiles score under various rules. When all the tiles are placed, each city is scored and each player counts the worst scoring of their two cities they helped to build.

Neither of these games involve political attacks in the way your idea does, but the remaining mechanics might give good ideas for how to go about the rest of the game.

[Simon]

Thanks for the game recommendations!

I have played 7 Wonders, albeit only with 6 or more people. With so many people, somebody on the other side of the table will win seemingly at random because their neighbor screwed up. It'll be a better competitive game with 3 people. Even then it's still too little direct control for my taste, but that's personal.

Mombasa sounds like I'd like it casually.

Yeah, Between Two Cities comes closest, lacking the diplomacy of attacking distant players. I'm hesitant to play heavy diplomatic games anyway. As soon as games rely on diplomacy foremost, the exact rules of the core game are secondary. But I'm interested in game design by itself, I want a-ha moments from juicy rules and the resulting challenges.

Scoring games feel weird: You do many things that feel unrelated. I always feel like designers implement scoring and victory points when they're lazy and can't think of a coherent goal that's supported by the rest of the game.

-- Simon

[mobius]

here's an interesting puzzle which is clearly inspired by the game Mombasa. The goal is to get all seeds into the big bowl; you can watch and see how the rules work. See if you can solve it for yourself.

the game is "Jewels of the Oracle" and old PC point and click game. This particular puzzle was always one of my favorites from the game because it's fairly unique and somewhat challenging but pretty easy once you figure out all the rules.

https://youtu.be/Usz4PgEuwhA?t=91

[Simon]

Mombasa is not Mancala.

I remember seeing this singleplayer puzzle before, didn't remember the solution anymore and found it again. Very nice puzzle, thanks! Here is the raw description of the puzzle.

Singleplayer Mancala with 4x2 seeds.

• There are 5 bowls arranged in a circle. 4 bowls contain 2 seeds each. The 5th bowl is empty, it is the goal.
  
• You win once all 8 seeds are in the goal. Play consists of a series of moves until you either win or lose.
  
• A move consists of taking into hand all seeds from a non-empty non-goal bowl and sowing one of the taken seeds into each subsequent bowl (irrelevant whether or not it is the goal), going around the circle counterclockwise, until the hand is empty.
  
• You lose if the last seed of any move falls into an empty non-goal bowl.
  
• If the last seed of a move falls into the goal, you have free choice for the next move; otherwise, you must make the next move from the bowl that received the last seed.
  

-- Simon

[mobius]

Mombasa is not Mancala.

oops!

Come to think of it; I had thought Mombassa was a dance!

The description of Mombassa reminds me of a card game called Pit. Which is a turnless game that sort of resembles stock market transactions. THere are different types of cards; "wheat, gold, oil, sugar, etc." The goal is to trade cards until you're entire hand is composed of just one type.