Journal

Starting point (2025-01-07)

Took some time to think about this by making notes in Things on my phone:

Points for duration
Popping bubbles as typing
Precision scrolling
Taps and double taps
Swiping left and right
Don’t fixate too much on it being realistic, just cycling through those ideas
Plus emotional instructions
Pauses/waits
Taps in other places on the screen
Game like interactions? Angry birds, fruit ninja, etc? Associated emotions or just random? > I think random right? What is the general affect of a phone user? Take notes. Occasional > smile but mostly dead looking or serious or furrowed?
Is there a game over?
Are there leaderboards?
Can you quit?
Keep it fucking simple?
Double tap bubble style
Some kind of minimalist abstract language of input

Made a p5 template project which is where we are now. I’ll start with that assumption and go from there.

Question: what to do about desktop? Have “it is as if you were on your computer” too which the same stuff but some different interactions? Yes makes sense. But separate projects I suppose. One after the other. Phone first because it’s funnier. Computer one is the opposite of Boss Mode which is fun.

Getting it together (2025-01-07)

While I have a couple of moments let me think some more about the interactions/structure here. I will do a few sketches tomorrow to figure out some of the shape as well.

Structure

• Intro screen with the title, and then maybe a button for instructions and a button for playing, and maybe a “high score” listed on that front page (assuming the simplest version where people just try to beat their own score rather than a leaderboards thing which sounds like probably the wrong effort/reward? Especially if it’s easily hackable and thus kinda pointess)
• Instructions screen that may or may not actually be needed? But could be about how you perform the different interactions like tapping circles, dragging other things, etc. The objectives of the game etc.
• Game screen itself which shows the stuff, has your points building up somewhere and otherwise relatively abstract shapes etc. used to define what you do (see below)
• Game over screen (how would you have a game over though? I guess some kind of timer that triggers at a LACK of interaction and the screen starts going red or whatever as if you’re dying, hahaha, and then the game ends after… whatever… five to ten seconds of non-interaction. Love it.) Meta note here: classic example of design happening/falling out of some more structured activity like designing screens. In talking about the game over screen you think about what game over is and design it there in that moment.

Interactions

I probably need to sit with my phone for a bit to truly taxonomize this, and also observe other people, but as a starting point (which if we’re being honest is probably going to end up being good enough as phones aren’t that complex and I’m not really trying to precisely reproduce recognizable apps or anything)

• Bubbles/circles that you tap to dismiss (sound effect?)
  • These can be used to simulate typing (if we cluster them down the bottom and have them kind of repeatedly respawn to allow for a rapid typing motion)
  • These can also be used for any taps we want on the main screen itself as if you’re selecting stuff in an app etc.
• Double-bubbles?
  • As in some notation that indicates you should double tap. Just thinking of something like liking an image on Instagram, though it’s true that’s the only double tap I can even immediately think of in that context? Would want some way to indicate it, like a double circle button and you pop the outer and then the inner
• Swipes
  • Most obviously left and right as in a matching app, need to think about look at how this work in terms of whether it’s like you’re “throwing” or more precisely dragging? Do I need to get into that level of sophistication? Should I distinguish between dragging and swiping at the basic level? Or allow both behaviours to be selected by the player on the same element?
  • One obvious UI element here is the way a slider works, press down to select the handle, drag it to the indicated position, release (much as in It is as if you were playing chess - mentioning that makes me note that there’s a failure state there - does the element just reset?)
  • But if it’s about a “throw” style of thing maybe a ball with an arrow and you’re meant to just physically flick it off the screen rather than drag it? Will this start to get confusion? This is what sketching it for eh
• Scrolls
  • To the extent that doom scrolling is a big thing it obviously gotta be in there, any scrolling, doom or nondoom or predoom
  • I think these can be pretty precise, so they work well with the slider model probably
• Pinches/two finger interactions
  • Pinch in and pinch outs
  • I guess two shapes that you grab and drag into a circle or out of a circle? This one sounds a little informationally dense and might be hard to represent? Unsure about that one…

Are those the main ones?

Sequencing

And then there’s the question of sequencing. I don’t like the idea of simulating a specific app experience, but I think it’s true that there are kind of higher level organizatinos that are worth preserving right? Like swipe down then double tap, swipe down then double tap (Instagram where you’re liking pictures), or swipe right, scroll down, swipe left, etc… (and note that these could potentially be paired with emotion/face notes… but it may be best not to… I can feel this very real tension around whether I’m implying a narrative/real use or not - note that this seems key to me and that as of right now my heart/mind lies with the idea of NOT trying to do any simulation)

What are some of the larger organizational units we might think about (more and more this is seeming like I need some sort of small field study - if only of myself - of using apps and seeing what the behaviour is over time?)

• Instagram - scrolling down interspersed with (double? doesn’t have to be) tapping, maybe even commenting
• Matching - swiping left/right pairs with scrolling (through a profile? I literally don’t know how this software works - do you even scroll up/down in it??? I have asked Femke, but I can also just get one of these apps I guess – later she reports in that you go left/right for no-match/match, up fast for super-like, up slow for scrolling the profile)
• Reading the news/internet more generally? Scrolling, tapping, scrolling more… but a slower kind of scrolling I suppose; could arguably involve back swiping to change to previous page, but that’s seeming more simulation-y
• Settings? Where you’re kind of tapping through a hierarchical menu… e.g. just a sequence of taps and scrolls… I guess that’s basically just the previous one though so maybe that doesn’t matter
• Match-3 game: targeted small swipes all over the screen
• Angry Birds: targets small swipes at edge, followed by extensive waiting (too extensive really for the purpose)

Who are you trying to fool?

In amongst all this, a key question: WHO is meant to be fooled? Not the player, they’re meant to look like they’re on their phone which means mostly just a cursory glance from someone on the metro. But COULD mean they’re at home or at a party looking busy which might get more scrutiny. But still, it’s not like you ever look at anyone and think you know what they’re doing? Maybe matching swipes. Or games.

This is where the sort of metanarrative of the game comes in I guess. Much as in the previous two… the idea that it’s a tool with a purpose. In this case my working theory (which I really like) is that it’s a tool both to look like you’re on your phone (and so a NORMAL PERSON) but also to NOT be on your phone and thus not subject to the abject terrors and punishments of social media and the news and so on… but then of course to ACTUALLY STILL BE ON YOUR PHONE in terms of shutting out the world, hunching, wasting time… but then maybe arguably to ACTUALLY BE MEDITATING??? Ha ha… haaaa? What if…

The game-iness

A part of all this is that we have a game layer. You score points for

• Duration
• Successful interactions

And I suppose that’s all. And that’s plenty. Could be really juicy, could be restrained. Unclear for right now. Kind of funny the juicier it is, but the juicier it is the less I can buy into the meditation story… which I do actually quite enjoy?

Emotional guide; face notes

There’s the component of this that refers back to It is as if you were playing chess, which is guidance on how to compose yourself physically, and most obviously facially (though because I just wrote “physically” I’m realizing that postural changes work pretty well here too). I don’t think these need to be connected to the interactions - you can smirk, frown, raise an eyebrow, be dead-faced (the most common note, haha) to anything any content any interaction. But this is an important bit for the “urban camouflage” idea involved in this.

God, this is actually pretty good? I’m talking myself into this pretty hard right now. Ah, young love.

Early sketching (2025-01-08)

The resolution on this image sucks, but it’s still legible I think. Or not? Maybe click to view? Struggling with just how terrible it is…

Next step is some prototyping, indeed.

First prototype; Feedback; Suite talking; and beyond (2025-01-10)

Early days still but did produce a simple prototype of randomized circle-tapping yesterday (which would be the build associated with this journal’s commit). I was happier than expected with the visual presentation on my phone, though need to discourage/rule out landscape because it just doesn’t work and isn’t part of the way you generally think of/see people doing stuff on their phones. But yeah the opening salvo worked pretty well. The commit included a couple of things to think about:

• Consideration: the amount of time between taps
• Consideration: The size of the icon, the size of the hitbox(circle)
• Consideration: OOP versus functional programming as this scales up?
• Consideration: Juice (or lack therein) around the act of tapping (sound, animation, something else?)

(44c547a)

Feedback

That’s all mostly technical stuff but it’s true especially I need some kind of philosophical position especially on the feedback element (juice or not juice?). There clearly needs to be feedback but my inclination as of right now is to lean into the kind of “Zen” conceit involved here and to make the sounds simple chimes or other zenny sounds (gongs?). Some way of continuing the emotional tone of calm, sort of trying to catch onto the seeming calm of this kind of phone use, but making it actually calm, but making it still engaging? So prototyping some sound formats makes sense, starting with Zenning. The other more obvious would be like a “pop” and the satisfaction around that, but I think that may lead to a more brainless experience?

Some simple animation for the circles appearing and disappearing makes sense too - maybe a combo fade+size in, and then just a size-out? Will try something today as I want to prototype a bit.

Then there’s the expressions thing… how to communicate body posture, emotion, thoughts, faces. Where to put it. I can just copy the chess game but worth revisting. Spoke with Rilla this morning around using emoticons (as in ASCII emoji) as a way to suggest things, but I think I probably still prefer the austerity of texts for this element… there’s this balancing act/trick around trying to hit the engagement/dissociation/peace target. Which I like as a core challenge for this one actually. And for the whole potential suite.

Suite talking?

One of the things that has been rising in my mind through thinking about it, talkng to Femke (about matching apps), and talking to Csongor via the growing stuff repo, is this sense of a larger and more specialized suite of apps based on this core concept. It is as if you were on your phone is the generalized idea, but it’s clear that an interesting and worthwile thing to make would be replacement apps for all the classics (WhatsApp, Messages, TikTok, Instagram, Facebook, etc.). Each one is a minimalist set of interactions that replicate what it looks like to be on those apps, but subtracts the content. And you could have all of them to create this phone environment that’s purely about the (soothing?) motions of social media etc., but is really about calm and inner peace because you get the advantages of the Wall of your phone (nobody should bother you, you’re entitled to this space) without the anxieties/pressures of the actual content.

Making this a much bigger project is pretty interesting. I think it still makes sense to start with the general case, though, and then perhaps to expand from there. Because I suspect the larger/broken down Suite would have to be a little more professional, maybe even on the App store instead of “just” a website. Though maybe the website works for sheer accessibility as the browser really isn’t bad in terms of the screen real estate available and you could make a little fake home screen and so on for it.

Anyway, I think this is the right starting approach for right now.

And beyond

Rilla asked the question “when are you going to make It is as if you were being a person?” and that’s genuinely pretty inspiring. I can imagine an album or a podcast (which? podcast is funnier) which has tracks/episodes that are literally vocal instructions for “how to be” in a specific context. Could consult with Jorge and Jadé about language, pacing, etc. Maybe make background music in the PO or something fun, and just put them out there. Like 20 minute tracks (akin to meditations) where you don’t have to deal with the pressure of figuring out how to be normal/unobtrusive.

Plays into neurodiversity stuff in pretty obvious ways, but I also think it can play strangely into entry points to meditation, freedom from ourselves. Could also imagine a novel where all these things exist, but let’s not go there please. PLEASE.

So that’s something else to consider building up some experitise and scripts for, but it’s a totally separate project, but might be fun to think about an episode/track for A MAZE. A general purpose track would be really funny too. Just “around the house” would be funny. There’s a sort of bizarre “life coaching” angle in it too. It’s definitely funny.

To p5 or not to p5? Thumb radius? The real thing? Zen? (2025-01-13)

To p5 or not to p5?

The most boring thing first. In running tests it seems like at least in the ways I’ve been able to investigate, p5 is giving me a pretty massive lag on touch events, enough that it feels bad. I can’t seem to find a way to make this not happen - mostly it seems like any touch fires two events, and the second one is actually the one that takes effect and is delayed enough that it feels laggy. It would be great if I could figure this out as I’m pretty comfortable in p5, but if I can’t?

I played around a teeny bit with PixiJS to write this little thing

// Create our application instance
(async () => {
    const app = new PIXI.Application();
    await app.init({
        width: window.innerWidth,
        height: window.innerHeight,
        backgroundColor: 0x2c3e50,
        antialias: true
    })
    document.body.appendChild(app.canvas);

   const gr  = new PIXI.Graphics()
    .circle(200, 200, 30)
    .fill(0xffffff)
    .on("pointerdown", (event) => {
        app.stage.removeChild(gr);
    });

    gr.eventMode = "static";
    
    app.stage.addChild(gr)
   
})();

I ran this in their sandbox and learned a few basic Pixi things to get a circle you can tap and it goes away. It goes away instantly. However, it’s also true that I can get “instant” performance in p5 so long as there’s no conditional checking if the circle is clicked… which obviously I need, but… why the fuck would a very minimal if-statement cause a substantial delay in processing exactly?

Well okay while I was writing that and becoming so incredulous that this kind of problem could conceivably exist I dug further into the double-event problem and wrote code to ignore the “mousedown” event that’s triggered (LATE) on a touch, listening only to the “touchstart”. That turns out to work it seems - I get a responsive feel for the taps (though I need to test directly on my phone).

So for now I suppose I won’t abandon p5 just to keep prototyping in a world that I know, but there are probably a bunch of other questions that are going to arrive around potential physics and feel that I dunno if I can solve. Swipes are a huge one (though I can use the swiping library I think to address some of this stuff? Hammer was it? Something else? Swipe.js?). Still really unclear on how to handle specifically the “swipe right” kind of feeling most of all in terms of symbols… could I have literally an arrow that says… swipe right? swipe left? scroll down? scroll up? stop? etc… more directed, less zen? Maybe that’s the thing that gets me to a less zen place which I think might have been a distraction? (Though I quite like the zen version of this? As opposed to the “fitting in as a human” version… hmm it’s still unclear…)

Maybe you could even incorporate the instruction iself into the text about what you do with your face? “Swipe right and wince”?

Thumb radius?

Exciting I know but as I was interacting with the early prototype on the metro I found that the circles were spawning too high sometimes for me to reach them with solo thumb, which I think it’s a pretty default design issue that you are not supposed to have in all these apps? So I should probably think about a spawn zone that’s lowed down to deal with that so that this thing can be successfully one handed.

And thinking of that made me think of

The real thing?

What if I approve the design of this game in terms of “secretly” switching between modes which reflect the interaction patterns of different apps, e.g. a TikTok mode, a YouTube mode, an Instagram mode, a Messages mode, a reading the news/website mode? That would help sustain more “sensible” patterns. I wouldn’t even need to tell the user this is happening up top, and it would then feed more or less directly into work on the App Suite version of this project, for which this thing is in some ways a prototype/general case?

That would allow me to tackle the whole thing mode by mode rather than as a whole mega thing.

So that’s the next kind of plan; break it into modes, have the modes switch at random points of time, but don’t tell the user so they don’t need to worry as much about being “in character” for the mode. This is sort of the “least stressful” version?

Zen?

I need to think more about the Zen thing. Versus “just” the social camouflage thing.

Swipe prototyping; visual prototyping; the main thing (2025-01-16)

Swipe prototyping

I did get a basic swiping thing going with a visual indicator today. Just a little pip that runs along inside a bar based on swipe velocity. It’s not much, but it’s felt helpful to have something a bit more clearly responsive and it felt like it got a bit at the question of what kind of visual representation to have… which in turn led to doing some visual prototyping because I’m still not really sold on the approach…

Visual prototyping

The main thing

In putting together the above visual prototypes (which I did not labour over, but which were helpful as a way to break free from the code) I kept running up against the classic stupid question: what is this thing about. As with any portion of the project, it’s really hard to make serious decisions about anything without having the underlying principles in place.

There are a few competing ideas going on here, not all necessarily on the same axes…

• Is this about social camouflage? Like the speculative design idea of a near future (present present?) where people just want to look like they’re nice normal folks using their phones, but they’re exhausted by the actual act of absorbing all the information, dealing with their brand, keeping up with the online world
• Is this about a zen experience with your phone, leveraging the familiar (soothing?) motions of using a phone, but reframing them to be about abstract ritualistic movements that can be meditative and freeing?
• To what extent is this about fidelity to specific applications? As in, should this recognizably feel like “I’m on TikTok” or “I’m browsing Pinterest”, or should it be a layer of abstraction away from that, more at the level of “I’m scrolling, I’m swiping, I’m tapping” in various combinations?
• How precise are the interactions in the experience? Is it “scrolling” or is it “scrolling to exactly this position”? Is it “typing” or is it “typing this word”?
• How gamified is the experience? I’ve thought in terms of points initially, how does this fit in with the ideas I’m listing out here? What do points say about the experience? Am I still interested in the idea you can lose by not interacting often enough? (Would this be something than Zen mode turns off?) Does the gamified part of this (points, losing on time) end up being a “commentary” on social media in a stupid way or a good way?

I’ve thought about the Social Camouflage thing as primary - helping people fit in while avoiding social media exhaustion- What about passive experiences of a cellphone? When people are watching YouTube, they’re just watching YouTube; do we have a “Watch this rectangle” activity? I mean, that’s very funny, but goes against the idea you can “lose” I had earlier. Or does the timer not expire for watching the rectangle? Watch this progress bar, with no timer for losing.

• but it seems like Zen can be in there too “just” by adding a Zen Mode to the experience either at the beginning or even as a toggle always at the top of the screen?

In writing this out I think a more abstract understanding is better. So no need in this version to think hard about what the interaction sequences (and specific screen locations) are for, say, TikTok.

I’m concerned by the tension between relief/social camouflage and the gamified version where it seems like it would be very stressful. Why would you play this as a game specifically? There’s some way in which that ends up feeling too close to the real experience? Am I just concerned that without some gamified element people won’t see the point in interacting longer term? But if it’s an application, then it’s not really intended to be played in that way - it’s a tool not a game?

This is making me think we kill the game-y elements.

So we’d be landing on something like:

• Social camouflage is the main perspective. You can look like you’re on your phone without being on your phone, while actually being on your phone.
• There is a Zen Mode, where everything just makes gong sounds. Kind of a joke, and kind of not a joke. Could include some “wisdom” in the instructions.
• It is abstract, so just a sequence of classic interactions (maybe grouped into things like scrolling and tapping, scrolling and scrolling, swiping and scrolling, but maybe not even that); not intended to mimic specific apps (but there’s a whole separate project where we do exactly that)
• It is not a game so there are no points, it’s just there as an app when you need it.

Nexts (2025-01-23)

I’ve done more prototyping at this point such that we’ve got passably okay version of generally tapping locations, swiping left or right, keyboard input.

The big interactions missing are double tap (maybe I seek to solve that today in the abstract) and scrolling (which introduces the fear of working with swipes and pans somehow, but maybe it’ll be fiiiine)

Once I get some kind of version of those two things it would be time to think more structurally about how all this works in terms of a flow through the game, notably changing modes, how I’m going to deal with the question of visual representations, how I’m going to deal with colour (or not), how I’m going to deal with sound, …

Uh quite a lot of stuff. My main feeling, though, is that the project is on a good trajectory that I can handle. And that it remains a pretty fun feeling idea.

Refactor? Visuals (2025-01-23)

Refactor?

As I started trying to “easily” “breezily” implement double taps it felt like the moment that everything would break apart (mostly my mind) in terms of the structures at play. It spiralled me out into thinking (fantasizing??) about OOP and inheritance and a beautiful clean cool implementation of All This. So I stopped what I was doing and turned to visuals as a way to do something a bit less dizzying.

But the point remains. At some stage I either need to move to a “better framework” for this project (plain JS? Pixi? Phaser??) or I need to build some of that better framework myself. Neither option is super appealing in this exact moment, but the stage to do it is probably closer to now than to later.

Ugh. But also, kind of fun once you get into it? I feel like I’m oscillating on it, but probably it’s “easy” (ahahahahaha) to work with p5 and OOP/inheritance. Plus I’ve just kind of wanted to build something out entirely in p5 and I guess this could be it?

Visuals

Anyway I also did some more visual stuff to think about the front end of this thing.

More ideas about scrolling. The main preoccupations here are going to be:

• The question of whether there’s a target and whether you can overshoot it with a too-vigorous swipe-scroll
• The question of confusion around the scoll down/up relative to actual swipe motion and how you indicate/don’t indicate that thing

Doublt tapping seem relatively fine. If we’re going symbols then I think the double circle makes plenty of sense in terms of a target and a simple indication of whether the double is complete.

New on the scene, and plausible with the current plan to remove the “game” stuff around scoring and passivity. The idea that one of your tasks is just to look at something for a set amount of time. (Note that this isn’t actually incompatible with points now that I think about it, if we explicitly introduce the timer for how long to be passive for.)

The question of what the instructions should look like. Pretty trivial tbh but I thought I should do it. Probably a black background lends them more credibility. What about a white box outline? Probably too confusing if we have a “watch this square” setup somewhere in there. Putting them at the top, where we’re less likely to have interactions (because of thumb radius) seems smart.

And then two kind of vomit-inducing colour palettes. I don’t like either of them but I thought I should at least gesture toward it. The muted greyscale that references/is from the Chess edition makes sense to me. Abstraction, not fun, not unfun, just there. Dark mode/light mode? Perhaps yes.

I OOPed in the code pool; I wanna be an FSM (2025-01-28)

I OOPed in the code pool

Wow, Pippin. Just wow.

Well I did convert the “code base” to OOP over the last couple of sessions and it was predictably pretty satisfying.

The unsurprising result is that it makes the project feel kind of manageable in a way it wasn’t feeling before when I was kludging away at the functions level. It no longer seems like I’ll go further and further insane as I try to add additional ideas into the code because the place to intervene is much clearer and I can rely on inheritance to keep things (relatively) calm.

I wanna be an FSM

The more surprising result, which I mentioned in commit 4d55b31 is that it feels much more like there’s an alignment between implementation and the spirit of the project. Some of that is probably ultra ultra inside baseball in the sense that it’s just something I care about… the idea that the code reflects the result in some fun way, whether metaphorical or literal or whateveral.

But I think there could be more to it. The code is now broken down mostly into

• States, which are activities like “Web Browsing” or “Messaging”, say
• Interactions, which are atomic actions taken like “Tapping in a random location” or “Swiping left”

With the idea that a State is just a particular kind of sequencing (statistical? more calculated than that? a pattern? a grammer? Tracery?) of a specific set of interactions. A finite state machine OR WHATEVER. Amirite? But probably I am right and that is it and I should actually look at FSM implementation ideas.

But the appealing thing here is the connections I think get drawn between (let’s say I’m right and it’s) FSMs and the human desire to know what to do. It kind of explicitly is an answer to the very human cry “I want to be a Finite State Machine”? Input to output baby.

So I’m saying that in switching programming paradigms I’m finding a better way of thinking about/expressing a part of what this whole thing is about, which is the desire we maybe have (sometimes?) to be more like computers doing computer things in computer ways. (And perhaps to some extent a ton of videogames are exactly that in a more obfuscated way?). Anyway, I’d like to pursue this further as I go, but here’s at least a marker of something found.

Reimplementing the wheel?; Structures; A pause explodes my brain (2025-01-30)

Reimplementing the wheel?

Yesterday I spent time making the scrolling interaction more like what I think of when I do it on my phone, which is crucially that I will semi often flick scroll (swipe into a scroll) and then catch the page to stop it (a press). I felt like that scrolling behaviour needed to be in there, so I made the effort to do so. And as part of all that introduced the pan/drag based scrolling as well.

So I’ve ended up with a kind of complete representation of scrolling at this point. Including different “length” scrollbars (visually the same but different relative scrolling speed - needs work). All of which does for sure have me wondering if I’m just reimplementing stuff that already exists if I just used some framework?

In a way, too, I think my intuitions here are related to the sense that you could implement a lot of (all of?) the features I have so far by literally having… you guessed it, a scrollable HTML element. That you would… scroll… in the way that things scroll… on mobile… i.e. get all the behaviours for “free”.

Why not do that? I don’t have an incredible argument, but I feel cautious that approaching this as a sort of “weirdly implemented webpage” might lead to gaps in my ability to control how things look and flow into each other… that it might turn out to be secretly awkward to do well (god knows the way I’m doing it may too of course). I think there’s maybe some philosophy involved in this too; like if I had a truly scrolling element then it’s somehow against the principle that the player only performs the “act” and doesn’t actually get the result. But I don’t know if that truly holds up or not, given I have a UI element that shows scrolling and, for all intents and purposes, is scrolling? So I dunno. I dunno.

Anyway, for now the point is to use p5 to make a project and I’m happy to use the canvas and have a sort of app-y feel to the whole thing, rather than tangle with HTML which, let’s face it, is a nightmare much of the time… positioning and all that.. ughhhhhhhh. Just give me a rectangle baby.

Structures

Continuing with pretty meta stuff, I did some refactoring that meant that my Browsing state went from controlling stuff to just being a configuration object specifying the touch events (e.g. swipe, tap, pan) and interactions (e.g. scroll, tap) that the state should run through. Which felt really liberating in that moment - the vision of all the specific states/use-cases/activities as configuration of an underlying structure.

But does that have significant implications for how things unfold down the line? Does it design/engineer me into a corner? Or a channel? Or whatever? Does it close off design space at the code level?

There are various things where I’d want flexibility per activity, but as I’m writing this I feel like it’s mostly data driven at that point? That any activity is just a sequence of interactions fundamentally? So maybe it’s fine? I DON’T KNOW.

A pause explodes my brain

The other very tiny thing I did today was add a pause between interactions. So you finish a scroll. It disappears. There’s a beat. Then the next thing shows up - another scroll or a tap depending RNG. That was fundamentally just a needed thing and I implemented the dumbest version (a delay with a small random variance in timing) just to have it there because it felt badly off without it.

But in doing so it felt like it highlighted/invited other elements I now need to address.

Acknowledgement

The interaction seemed to want some kind of UI acknowledgement of task completion, I guess in a way that previews and justifies the element disappearing. So maybe the element turns green, or flashes, or whatever smiles at you. Maybe there’s a sound. Maybe a check-mark shows up. The specifics of this depend a bit on the personality of the overall project actually. There’s a big difference between “turns green” and “Clippy shows up and congratulates you” for instance. So it’s a thinker.

Does this look right?

And then of course there’s the question of how long these delays should be. Because those delays are in part about staging out the user’s interactions. You don’t want them just scrolling non stop every 500 milliseconds because that wouldn’t “look right” right?

(Although I think there’s a big question floating around that I haven’t tackled which is the question of HOW RIGHT this should look - is someone glancing your way? Are they studying you to make sure you’re REALLY ON YOUR PHONE? Or is the metric that the player should feel like they are REALLY ON THEIR PHONE? That they can imagine some kind of task flow where they would be doing what they are doing?)

Acts, actions, and interactions

And ALSO in adding the pause I found myself wondering about the place in the flow of all this of the “acting” (the instructions for facial expressions and other stuff, breathing patterns). I wondered in the commit whether there was a case to be made for having those Acts be part of the larger flow, so not on screen at the same time as the interaction but rather as an action of their own.

There are ways that makes a kind of very soothing sense to me… a flow of actual interactions and more emotional/psychological/postural actions. There’s a part of me that things that having actions and interactions on screen at the same time would be a kind of cognitive overload risk when I’m trying to keep it as simple and robotic as possible.

The counterargument to that is that there might be a case for performing an interaction WITH a specific affect, and if they’re sequential I don’t see how that would work. Something like “perform the next interaction with an exasperated face”? Oh… actually I really kind of like that. “React to the previous interaction with disbelief”? Hmmm… that versus “Perform this interaction with an exasperated face” …

Hmm I’m convincing myself of the serial flow. I will at the very least try it out next time.

Dating; acting; testing (2025-02-10)

Well I “missed” a week of development there because I was preparing a talk I gave on Thursday at the ARTSLab at the University of New Mexico for the Gale Memorial Lecture Series. Fancy me. I did talk about, or at least mention, this game so there was a teeny tiny sense in which I did stuff. The idea of this game got some good laughs and some interest from one audience member about helping people with addiction to phones etc. which was kind of interesting - I told them about the zen gongs version and how that had made me think about it as a potential meditation aid. But that’s ended up feeling more like a “fork” of the overall project. But funny to think about forking it.

Dating

I’ve already marvelled about it in the commit messages today, but it was really pleasing to implement a Dating activity with a minimum of confusion and difficulty thanks to the framework I’d set up for the Browsing activity. I was able to just add a set of classes that represented the approapriate swipes and then it kind of just worked without too much difficulty - it was working within say 20 minutes of starting in on it I think.

As of now the Dating activity is “just” swiping left or right or up with different weightings. So it’s mostly left (no match, 20 of them in the array), a bit of right (match, 3 in the array) and a tiny bit up (super match, 1 in the array). That is a breakdown of 83%/12.5%/4.5%. Probably too weighted on a superlike.

This has done the usual thing of opening up questions, particularly around:

• representing weightings/probabilities (the current approach of proportional representation in an array feels more and more clunky and I probably need something more tuned to specifying probabilities directly?)
• the question of how specific to the activity I actually want to be: does it matter how many lefts/rights/ups there are? I mean, yes it does, but to what extent
• and continuing from that, what about reading someone’s profile which is a little sub-action sequence… does that matter? It would matter in the version where I’m “replacing the app” in a later iteration of this project; but does it matter for this first one, or are we just trying roughly to get the rough ideas of the interactions rather than have them seem accurately simulated in terms of sequencing etc.?

Acting

I added acting back in. It’s currently serial, and it’s currently that you have some probability of getting an Act versus an Interaction. It works in the sense that it shows up and so on, but I need to start taking a closer look at how it feels, building to…

Testing

How do I test this thing? Most obviously I just test it on my own for as long as I’m able to identify things that aren’t satisfying, but at some point I probably need to show it to someone else.

There’s the question of framing too, which may come out of any testing I do.

Let me pick up my phone now and do some “dating”…

…

It felt… pretty great at a base level? I continue to feel like the core project here makes sense. I don’t even totally know why, but just being told to take the actions associated with a more specific kind of real world activity and having them be recognized is in itself satisfying.

In doing it there are some questions/notes though…

• The timing between acts and interactions felt completely janky… the acts felt very divorced from the flow of the whole thing, there was for sure an argument/feeling for them to be visible at the same time, maybe as two independent systems, maybe actually connected to one another
• The question of the relevance of the text was more important than I would have predicted… the things I had in there like eyebrow scratches and three nods didn’t feel all that well adapted to the activity of swiping for dates…
• But that in turn raises the weird question of just how specific you want to get there… the weird question I suppose of just how tuned to specific use contexts our body language is with our phones, which I did NOT expect
• But to the extent the Acts are meant to support that broader feeling of “being on your phone” successfully there’s an odd balance involved in how much you do or don’t understand/know “what you’re doing” on your phone…
• Maybe some of it is “just” the specificity of dating as an activity though? it’s not a potentially generic as browsing the web or text messaging…
• The strong emotional valence involved almost feels kind of tricky and that bleeds out into questions about the emotions involved in the other things… texting about the death of your dog versus texting about which showing of a movie to go to… different expressions and you wouldn’t(?) just immediately shift one one state to the next and back again emotionally?

Hmmm, so there’s some kind of challenging stuff in there.

• Specialist acts for different activities or even drilling down to iterations?
• Pairing acts and interactions?
• Grammars of acts and interactions?

😮‍💨

What would it take for this game to be done? (2025-02-14)

Not that it’s “taking too long” but it might be a good time to consolidate what the different arms of this are that require more work. Obviously it’s not ready now, but what does it need? In no particular order…

• Zen Mode - in or out? I think it’s pretty funny so probably in; would need to UI design for that, which means it needs to go somewhere (top menu I suppose?) and it could go there with an info button too I suppose
• How many activities is enough? - well that’s a good question, Jim. At least a sweep of the most obvious phone activities? So I’d be missing Video Watching most obviously right now? Browsing, photo stuff, video, dating, messaging… seems like enough. So let’s say “add video and see” (TikTok versus YouTube though?)
• How often do you switch activities? - I suppose just often enough for it not to be excruciating? Some of that is about the support the Acts do/don’t offer to things. It’s clearly not as engaging as Work or Love because it’s taking away the sense of progress. How much weight can the Acts carry, how much should they do?
• Writing Acts, Targeted Acts? - Right now the list of acts is incredibly short and not even very appropriate; these need to be written; should they be written with Tracery in hand?; should they be more concrete/specific to emotional circumstances?; should they be tuned into specific activities? (As I write this I think “probably”)
• Interaction Sequencing - Right now interactions have probabilities of showing up but they’re not actually sequenced in a way that would build a sense of a well-known interaction (e.g. scrolling Instragram is like: look look doubletap scroll down look scroll down look look look scroll down scroll down look doubletap – or something); having a kind of “grammar” of interactions sounds pretty cool but also really tiring, but also maybe what I should be doing… at worst could just be selecting from an array of atomic interactions that deal with “one media unit” or something? Regular expressions? How deep does one go into this exactly?
• Symbols, words, content - As of right now I’ve leant toward a kind of symbolic UI idea where a scroll is indicated by an arrow and a scroll bar. One could also imagine a purely language oriented one “tap” “swipe left” etc. (though you would lose some positional specifics?). One could also imagine a kind of simulated low level UI where scrolling actually scrolls stuff. Though then how do you tell someone to scroll in the first place? Scrolling is confusing because of the up/down direction versus movement thing. It’s the main thing that makes me annoyed by the symbols I have. But if I imagine an actual “here’s instagram” thing I wonder just how legible that is and how to issue instructions and indicate stopping points etc.?
• That is, abstraction - A lot of this is circling around levels of abstraction. How abstract should this be to get at its core idea? Do I want the person to visualize what they “would be doing” or not? Should there even be legible activities in this one, or just a gigantic mish-mash or interactions? Should I be separating out the ideas of a generic “on my phone” from the more specific “on TikTok” setup I’ve also been imagining? Are these ideas kind of colliding here when they shouldn’t be?

Pippin lies down on a sofa and holds his head in his hands.

…

As I lie here, I think that it might have been overstepping for this project to think so closely into activities. And you know, I was happy about it at the time, but I think I even made a note that this would be useful for later, rather than for this exact project? I think it’s more important for the “on your phone” project that it’s just… “on your phone”ness rather than looking like you’re doing something specific. That there’s maybe even a value to it being kind of inscrutibly just “on your phone” and not “oh he’s on tiktok, oh he’s on instagram”. To which end you’d actually just want a totally randomized set of actions you go through, and some kind of pretty generic Acts that you act out at the same time that match up with someone who is just “on their phone”…

I think I buy that? I still want to make the Suit version where it’s all a more specific exploration of specific interaction sets, but I don’t think I want to confused them right now.

So if I revisit what needs doing it’s more like:

• Are there any other interactions needed to “complete the set”?
  • Possibly just the “look at this box” interaction, I think that one is valuable because looking at your screen while inactive is a thing we do.
• Write proper Acts that convey “I’m on my phone”, think about what that means
  • Kind of blankness mostly? I can look at people on public transport on the way home today and reflect on what I look like too. Half smiles, silent laughs, blinking, raise eyebrows, pressed lips, subtle stuff…
• Write a much simpler chaining together of different interactions; just any interactions in any order (see what it’s like)
• Consider abstracting the swipe/pan a bit more? Pan as just a “drag this to here” and that can also be achieved by a swipe? (Is there a separate swiping motion we would want?)
• Deciding on the place of language for saying “swipe” versus just an image of an arrow, versus a pip that shows it to you?
  • Hmm so there’s still just a little visual design hitch here to think through

More visual design (2025-02-14)

Spent a bit of time in response to the issues identified doing some more visual prodding to see where I could get to and I think I like the direction. The tutorial thing is reassuring - can make it disappear, but initially having the language to set the behaviour is helpful I think.

And then there’s the idea of how to position any other UI elements, like a Zen Mode toggle and an info button?

Leading to the idea of just dealing with that on a title page to keep the play screen relative clean?

As I do all this, that red arrow in the mockups is making me want a splash of colour in amongst it. Maybe the interaction areas/items could be pink or something?

Anyway, it was pretty soothing to do this because I feel like I’m creeping back to something simpler and less directly connected to actual UI. More abstraction, not less.

Progress; Zen (2025-02-20)

Progress

Hwell… I implemented those mockups pretty literally over the last few days and the game has kind of jumped from feeling a bit off to pretty much feeling like what I wanted? It suddenly feels close to done – though naturally that doesn’t mean it is close to done.

Crucially I think it seems to work. I’ve had Jim look at it for the “does it make any sense” angle and he got it. As mentioned in a commit somewhere, I’ve had Felix play it and got the excellent feedback that he didn’t want to stop, helping to emphasize the instinct that there’s just a bizarre moreishness to these sorts of basic interactions (which I’m hoping will also play into the Zen mode when I get to it). I also showed it to a couple of other people from the perspective of observing and they confirmed that yes, I looked like I was on my phone, haha.

There are still plenty of things to tidy up around the language, some tweaks to layouts and more. And then there’s Zen mode, a menu, and such. But yeah it’s reached that point where it feels inevitable that I’ll release it as a real thingamie.

Zen

The next big question for the game is how to create the Zen mode. This is more of a design brainstormy moment but the point is to use the same structure of the game (like, exactly the same) but to have to it communicate and ideally be a meditative experience. So:

• Sound effects become zen sounds like gongs for taps and… other stuff…
• Should there be background music? Dare I make something on the OPZ? Or something generative with chords? Or not really, let people plug in their own music?
• Should the colours change? Is there a “zen palette”? And relatedly is there a “better” approach to the palette of the non-zen version? I keep thinking about a sort of rainbow hue shifting thing, but it that just basically pretty dumb? Might look nice? I can try it
• We need an approach to the act texts which my first thought was just like a guided breathing thing, then I wondered about a guided meditation sort of set of things around observing/feeling/etc., and another possibility is a single specific quote or whatever that just stays there and you can meditate on the one thing throughout. I don’t dislike any of those ideas, but I feel like I should probably just stick to a single one. A quote may be hard to fit? There’s good old “If death alone is certain, and the time of death is uncertain, what then should I do?”
• Should there be a timer to explicitly pair this with the idea that it’s a session of meditation? Or again outsource that? (Could be a recommendation in the explainer text?)

So those are I think the main concerns. None of them are actually all that high challenge, just a thing to do. I guess I try out the different potential textual approaches specifically.

I haven’t written a Why for this project. Let me do that.