- Gestures in a lot of applications have made things more confusing by hiding functionality that you now need to stumble into to discover.

- Sound cues are used all over the place. Anyone who's ever worked in a kitchen hears the godforsaken ubereats alert sound in their nightmares.

- About ten minutes ago, I got startled by my phone deciding that the "you should stand up" vibration pattern should be three long BZZZZ-es... amplified by it sitting on my hollow-sounding printer.

- If another fucking god damn website asks me to chat with an AI agent in it's stupid little floating chat bubble, only appearing AFTER I interact with the page so it's allowed to also make an annoying "chirp!" sound, I WILL become a chicken farmer in some remote forest eating only twigs, berries, and improperly-raised chicken eggs.

All of these things annoy me, and actively make me hate computers. A silent glass brick can go in my pocket because I know it's not going to bother me or beg me to talk outloud to it. If it was some sensory-overload distraction machine (which, by default, it is) it would find itself over the side of a bridge rather quickly. It's getting in the way of my human experience! The one where I'm the human, not the computer!!

However once you realize that you can add new gestures without having to defend adding a physical or screen real estate button, it takes a lot of discipline to avoid adding more. I like to think that Steve would have told most of their people to fuck off and we’d have one or two new gestures now, instead of twice as many. They would have found some other way.

For me personally a similar thing was when Ableton Live transitioned from having a more "direct" interface to having popup menus for absolutely everything, and it took time for me to adapt to it for live performances. To be fair I never really adapted and just moved to something else.

Rather than coming up with creative solutions like they did before they just kept adding things to those popup menus. The app went from magic (by enabling me to perform live effortlessly) to frankly difficult (by having the interface become difficult to memorize and getting in my way).

Coincidentally was when they also started racking up bugs so much that they needed a couple years without new features just to clean up bugs.

If you’re doing something like a sound mixer you should be able to move more things to a second screen. Run the main app on your new tablet and the ancillary functions on your old one. Or a small monitor if it’s view only.

Mind you in the early days pages used to have hundreds if not thousands of text links all over the place, the only sites that do this now are the hardcore conspiracy sites where the author just adds several new links a day.

So in this dimension at least web UIs have changed for the better.

It doesn't matter that I can force my phone's vibration motor to only output an anemic "buhhhh..." no mater what coeffienct of bothersomeness some app sends to it. The person causing my phone to make that API call still expects the cacophony of pain to emit from it. We all become numb to how annoying this all is because it becomes the standard TO BE annoyed and distracted.

The uber eats sound is annoying because it conveys nothing except "whatever you're doing is unimportant!!!! PAY ATTENTION TO UBER!! UBER THINGS ARE HAPPENING!!!". There's a million other better ways to do that, so *I* find the information. *I* go to the stupid glass brick when *I* can take on a new order. But because we already set the expectation that the user is allowed to set off an alarm in any kitchen in the city for the low-low price of overpaying for food, the stupid glass brick tells ME when it's time to deal with it.

Spatial computing (like the example of a note taking app) now introduces all of the extra work of cleaning to a digital note. The computer wants me to sort my own notes now. It opens up the potential of being an e-slob for no reason other than my ability to make it as equally messy as my desk.

I don't know why we would expect this even-more sensory-focused model of computing to not also ratchet up the stress and dread of being alive.

I'm 27 going on 95 I guess, just send me to the old folks home now lol

> I WILL become a chicken farmer

Chicken don't hold a reputation of being calm and considerate about what noise at what level they'll emit during the day or the night.

If I can advise, silk worm farming could be it

If for real mental breakdown vineyard farming is where you should be.

You can also sip away your woe by rewarding yourself with a glass of wine at the end of the day.

But I couldn't disagree more with the premise. It complains that computers have been reduced from physical, tactile, hulking mainframes to neutered generic text interfaces, but I've watched the opposite happen over the past two decades.

My phone is physical -- I swipe, pinch, and tap. It buzzes and dings and flashes. I squeeze my AirPods, I pay by holding my wrist up to a sensor, I tilt my iPad to play video games and draw on it with a pencil.

Everything the article complains about, we've already solved. All of its suggestions, we already have. It wants "multi-modality" but we already have that too -- I can change the volume on my iPhone with physical buttons while I dictate. I can listen to music while I scroll.

Our interfaces haven't lost their senses. Our interfaces have more senses than they've ever had before.

Hard disagree. It's incredibly distracting and the constant movement of text, the introduction and disappearance of images within the medium makes it incredibly difficult to concentrate on the message.

It screams 'look at me, I'm really smart with all these neat effects'. But you know what interface for articles like this has served us pretty well for > 1000 years? Just the words. Please, just display the words rather than this conceit.

— Apple HIG

In 1985, after a year of finding that pretty but unlabeled icons confused customers, the Apple human interface group took on the motto "A word is worth a thousand pictures.

https://www.asktog.com/columns/038MacUITrends.html

Linked from Daring Fireball back in the day.

https://daringfireball.net/linked/2010/05/12/tog-word

But in UIs you usually have to have some text equivalent somewhere, on hover or long-press or in a menu or just as text for screenreader users, so you don't generally get to avoid translation even if you take visible labels away.

You don't have to agree with the author, but I don't see any irony here!

I'm currently stuck on LTE due to a power outage. The page is horrible to try to read due to most of the images being either in the process of being loaded or not loaded at all.

Reader mode is a standard feature on all major browsers on both desktop and mobile. Given you're so vocal about how articles should work by just "displaying the words", I'd suggest that you acquaintance yourself with the one feature that does exactly that.

Thanks to reader mode, you get to concentrate on the message. And we get to keep our joy.

"Then came terminals and command lines. Physical knobs turned into typed commands—more powerful, but our digital world became less embodied. Then came terminals and command lines. Physical knobs turned into typed commands—more powerful, but our digital world became less embodied. Then came terminals and command lines. Physical knobs turned into typed commands—more powerful, but our digital world became less embodied. Then came terminals and command lines. Physical knobs turned into typed commands—more powerful, but our digital world became less embodied. Then came terminals and command lines. Physical knobs turned into typed commands—more powerful, but our digital world became less embodied. Then came terminals and command lines. Physical knobs turned into typed commands—more powerful, but our digital world became less embodied."

I stopped reading after that. There are also missing full stops, which means it's difficult to understand what's happening.

The style and emotional feeling of the page is the message. An article consisting of only words is not an "article like this", and if you are starting from that premise you already totally missed the author's point.

And for many people, that message is, "go away, this is not for you".

Which is a valid take if that's what the author intended, but generally speaking, when people take time to pen manifests, they expect them to be read and heeded.

I totally agree with OP that the 'flat' visual style is appalling. (And gray-on-gray text is an obscenity.)

And what do you mean, disconnected from physical reality? I listed the examples where I pay with my wrist, squeeze my earbuds, draw with a pencil. I also snap photos and take videos, record with voice memos, send a pin with my location. I track AirTags, I identify plants by pointing at them, I learn constellations by aiming at them. Computing is more connected to physical reality than ever.

Another good experience was shaking my phone (an Android Motorola) to turn the flash light on and off. Another great natural movement is taking my iPhone and transfer the playing music to an Homepod by tapping on the two together.

For almost everything else I loathe touch devices. While older ones may be clunky visually, they are far more ergonomic. Yes, my phone can do a lot of stuff, but the whole process for a specific one is always clunky and companies go out of their ways to block you from altering the UX.

It would be sad if in 10,000 years we evolved to lose our tactile senses.

I'll also add the buttons and switches and knobs are not all that tactile. They are a modern human creation.

> buttons and switches and knobs are not all that tactile

But they are, though. You can touch them without pressing, and, once you get used to them, they have different textures.

Edit: or even img2img of rough sketches.

This is about "senses" as in "the five senses".

Not "senses" as in "they've lost their senses/mind".

The author is making a pun on the latter, but the article is about the former.

"Flat" interfaces aren't bad because they lack an ineffable whimsy of embodied human experience, they're bad because they threw out the baby the bathwater, tossing decades of conventions and hard-learned accessibility lessons in the name of supporting a touchscreen.

Compared to 20 years ago, everyone is shipping half-website-half-desktop abominations (e.g. with Electron[0]) and reinventing UX wheels. Too many apps/sites impose "their own look" instead of following what the user has already learned. [1] Often users must guess whether certain things are even clickable, how a certain toggle looks when enabled, whether certain settings are a single-select option or a multi-select tickbox... And memorize those rules with per-app or per-website granularity.

> You can talk while clicking, listen while reading, look at an image while spinning a knob, gesture while talking.

Those are all things people do after "make computer do what I want" has become automatic.

Now when--for example--trying to find the 21st item they just added inside a list that is vertically limited to 20 and the custom grey-on-grey scrollbar is always hidden unless you've currently hovering a mouse exactly in the right 5-pixel-wide strip between two columns of the interface.

[0] A sample listing of software readers may be familiar with: https://www.electronjs.org/apps

[1] That may be due to deliberate "remember us" branding, whatever was fastest-to-ship, because things to look new to get somebody a promotion, because they want to create a switching-cost so current users feel bad trying to use a competitor's product... Or because someone like the blog-poster has misguidedly tried to make a "richer experience."

I remember 20-25 years ago mostly using Windows widgets to make Enterprise apps. It was fast to make, fast to run, and users back then knew how to use. Didn't look the best, but at least was consistent.

The next 5 years we sort of tried to do our best, but most things were still sort of standard-ish.

Then for about 5 years or things like Bootstrap, Material, etc, dominated. It was nothing special, but at least consistent between apps.

But in the last 10 years pretty much every company I worked on had a custom UI built ENTIRELY from scratch by a designer and a small army of developers to implement it. It looks "the same but different" in an uncanny valley way.

I honestly feel like this is the worst possible use of frontend developers, period. Not only from a financial perspective but also from an end result.

But hey jobs are generated so what do I know...

Before software only needs to be useful. Now the C Suite thinks it needs to be engaging and isolating like a casino.

A similar effect I believe is what’s been largely responsible for the healthy botique indieware scene on macOS too.

What I’m talking about are apps built not because there weren’t serviceable options in their categories prior, but because there weren’t any that made an effort to be at unity with the larger GNOME desktop. Apps like Errands[0], Folio[1], Shortwave[2], and Newsflash[3].

[0]: https://apps.gnome.org/en/List/ [1]: https://flathub.org/apps/com.toolstack.Folio [2]: https://apps.gnome.org/Shortwave/ [3]: https://apps.gnome.org/en/NewsFlash/

Older UIs separated computer output from where user input is expected. Compare Windows 7 or 2000 to Windows 10 or 11 and the older Windows versions win. It's the same with Android.

UX designers follow fashion as much as the garment manufacturers on the catwalks of Milan, even if the fashion is uncomfortable.

Just because something is new, that doesn't mean it is better.

Good designers still exist but they are simply crowded out.

The same is happening today with the AI generated apps. Most front ends now in another 10 years will be filled with AI generated apps. But good design and applications will be around but they will be crowded out.

And you see this in almost other industries as well. For example, architecture has simply gotten worse. A building from today looks much, much worse than let's say a building from even 300 years back.

So we will simply have worse software and worse performing software which breaks down all the time in the near future and we will all suffer but there is no solution out of this.

Things don't always get better.

You don’t even have to go back that far. They still knew how to build decent buildings just 80-90 years ago. I think it all kinda changed after the Second World War. Maybe they needed to conserve money and build as much as possible?

The real challenge is that UI designs are ultimately constrained by their hardware. This means major interface innovations often limit where the software can actually be used.

Take tablet-optimized apps, for instance. They can fully embrace touch interaction, but this leaves desktop-only users completely out of the loop.

So unfortunately, truly revolutionary interfaces tend to require equally revolutionary hardware to match .

https://wattenberger.com/thoughts/our-interfaces-have-lost-t...

What are those floating letters? Does the keyboard have 3 rows of keys, or 4? What's going on near where the esc key should be? Why does the screen look like the back of a park bench?

  > The real challenge is that UI designs are ultimately constrained by their hardware.

The problem is what you mean by "UI"

UI means "User Interface". It does not mean "Software defined User Interface".

  User interfaces are composed of one or more layers, including a human–machine interface (HMI) that typically interfaces machines with physical input hardware (such as keyboards, mice, or game pads) and output hardware (such as computer monitors, speakers, and printers). 

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

The prime examples given were about mixing and matching capabilities that most hardware already has. Most computers and tablets already have a microphone and some kind of tactile input (touch or keyboard).

So, I wouldn’t say that you’re wrong in tying UI innovations to hardware, but it feels like perhaps you didn’t read the whole article. We can innovate by remixing existing functionality without having to wait on entirely new paradigms being adopted and universally available.

That being said I think it misses what made the old physical interfaces so appealing and useful. It's not that there's something inherently superior about multimodality; it's that physical interfaces are permanent, with defined edges and definite shape. Unlike screens you know exactly what's where, building muscle memory every time you use it. There are no hidden menus or moving parts.

Multimodality - such as being able to see the position of a slider at a glance, or feel its position by touch - is useful because it reinforces the absolute existence of a control and its state across multiple senses. Interfaces using voice and gestures like suggested are the exact opposite of that, because each point of interaction becomes even more disconnected and vague.

Some things are best if they stay simple, efficient, reliable stable, and quiet. Not needy, demanding, high-maintenance, attempting to ensnare us through as many of our senses as they can get their claws on.

Some things are an experience, other things should just be quietly useful. Do we ask ourselves which we should be, before adding another colorful icon, with a red dot in the corner, with a number inside the red dot, to the poor user's screen?

And I hate haptic feedback. I keep my phone on silent 24/7 just to not feel my phone creepily zapping my fingers, and for some reason silent mode is the only way I can accomplish that.

When Steve Jobs was saying DropBox is not a product but a feature, it surely didn't reassure DropBox that they could keep making simple and transparent services that the platform owner will gently play nice with.

https://worrydream.com/ABriefRantOnTheFutureOfInteractionDes...

Flat design is much more of a blank canvas than old desktop UI toolkits were, but companies are loathe to invest in "Juice" because it requires a level of attention to detail that works against fast iteration.

The whole article felt disconnected from reality to me, and this might be the core part that underlies it all.

The author seems to be hostile enough to smartphones that they don't really see much more physical using them has got. In particular in the last decade or so we got rid of so much of frequent polling to see if something needs attention or not, and notification management has become leaps and bounds better, at least on android.

But the author's focus seems to be more on social media ?

And they also idealize what people were doing before smartphones.

People scrolling Twitter rage thread today would probably not have been sketching birds in their backyard. Some perhaps, most surely not. TV was already widely popular for a reason.

I was on my roof in the dark at 1:30 in the morning in the cold and wind. I'm tired, can't really see much, but still need to actively work with the camera's controls. Thankfully, the X-T5 is covered in physical dials, switches and buttons. Without looking at the camera's screen, I can quickly change shooting modes and the majority of the settings I care about and be confident that I changed the right things.

The same cannot be said about a large number of modern cameras, which opt instead for a more digital approach.

In terms of modern "computing" devices, my cameras are an absolute joy to use compared to most of my other hardware.

So much so that I've recently been finding myself looking to recreate this tactile experience on my general purpose computers. I've been looking at weird bespoke dials, switches and various input hardware to make processing the photos (among other tasks) feel more tactile.

> The same cannot be said about a large number of modern cameras, which opt instead for a more digital approach.

I feel you. I can only imagine the horror of looking at a small but very bright touch screen in the dark, eyes adjusting, etc.

> So much so that I've recently been finding myself looking to recreate this tactile experience on my general purpose computers. I've been looking at weird bespoke dials, switches and various input hardware to make processing the photos (among other tasks) feel more tactile.

A physical knob on my desk to precisely control the volume of the speakers is a very handy one for me. Don't know how to live without that. Especially because one big yank on the knob mutes unsuspected annoying sounds.

Yes, flat design is too flat, and AI chat is too devoid of friction.

But mobile and tablets are better at certain things [1], and we shouldn't get rid of that either.

I saw somewhere (Bret Victor?) that tools have two parts: the part that fits the problem, and the part that fits the human. The example was a hammer; the head fit the problem (the nail), and the handle fit the human (the hand).

Notably, the two parts must fit their respective things, but they also have to work together.

That is what we should be doing: creating harmonious tools that fit the problem and the human. What that looks like will be different for every tool.

Our interfaces currently have two problems:

* Because they can have any appearance, appearance gets more attention than being a good tool. Example: flat design (good appearance) overriding skeuomorphic design (human fit).

* No one wants to redesign everything, so we all reuse the same base stuff (Electron, Qt, etc.) even if the result won't fit (one or both ends) or harmonize.

I would love to fix both of those problems, but because people are lazy, it essentially means creating a GUI framework that is flexible enough to fit almost any problem and any human (accessibility included) while making sure that flexibility does not destroy harmony.

While I am working on that, it is a tall order, and I am almost certain I will not succeed.

[1]: https://news.ycombinator.com/item?id=43350339

You want an AI that argues with you?

Kneading dough sucks if you do it a lot. It's monotonous and tiring. That's why frequent bakers use mixers with bread hooks.

Instruments are designed to be as friction-free as possible, given physical constraints. Friction makes expression more difficult. A violin is the most expressive because it has the least "friction" of valves and hammers and buttons getting in the way.

Sketching similarly is low-friction. That's why it's so much easier than oil painting. You can express yourself hundreds of times more easily, which is why oil painters start with tons of preparatory sketches.

I fundamentally disagree with the premise that friction is desirable. It's not.

I agree that too much friction is terrible.

But what I was saying, and the article seems to be implying, is that too little friction is terrible too.

Using Stable Diffusion is lower friction than sketching or painting. But the latter two are better.

The difference is that there is friction that leads to a good outcome, and friction that does not. Mixers with bread hooks are eliminating bad friction, whereas Stable Diffusion is removing good friction.

And in fact, there's actually more friction when using Stable Diffusion if you have an end in mind; trying to get it to output what you want is high in bad friction.

I just don't buy it.

> Using Stable Diffusion is lower friction than sketching or painting. But the latter two are better.

No, the latter two aren't "better". All three are totally different tools for achieving different purposes. I'm going to use Stable Diffusion to raise engagement on my blog with a hero image and a relevant thumbnail, I'm going to sketch to explore visual ideas and improve my skill of seeing, and I'm going to oil paint to carefully craft something designed to hopefully hang on someone's wall for a long time. (Well, not me because I don't know how to oil paint, but you get the idea.) I'm certainly not going to oil-paint something to illustrate my blog. Oil painting isn't "better".

And when I use ChatGPT to ask questions about math or physics or history or culture, the last thing I want to do is to make the process more difficult. I already spend enough time typing a prompt the AI can clearly understand. There's no way in which it would be made better with "good friction".

I mean, I literally don't know what you mean by "good friction". I don't think I've ever encountered it in my life. Life in general is challenging enough without having to add more challenge for no reason.

Friction with growth is good friction, so long as the friction is minimized for the amount of growth.

And by "growth," I mean anything that helps people to "level up," such as learning, gaining a skill, becoming more Christ-like, whatever.

Growth cannot happen without friction. Your use of AI is stunting whatever growth you could have gained from those processes.

"So what?" you may say. However, someone who applies friction to growth consistently in their blog/code/whatever will find that growth compounds like interest, and though they may be less "productive," their productivity will be better in the long run because they will have the skills to go beyond anything you could ever dream of doing.

Since you use the example of a blog, just writing blog entries has plenty of friction inherent in thinking and researching and writing.

I don't want more friction in generating hero images. That's not something I want to level up in. AI is not stunting anything, because if it weren't for AI, my blog wouldn't have images at all, or they'd be stock images that were even worse. But images help your posts reach an audience, so they're necessary. So Stable Diffusion is great.

I'm not living my life to build all the skills. Skills are a means, not an end. If I can choose between spending quality time with friends vs. building skills illustrating hero images by hand, I'm going to choose the quality time with my friends, because I already don't have enough of that.

Also, the very first example of supposedly "good friction" was kneading dough. That's not leveling you up each time you knead. Just use a dough hook if you've got one.

Anyway, as someone with a blog, people actually have conplimented me for not using images unless they support the point in the post. I see plenty of negative comments about blog posts with header images.

Ha, quite right, that's what happens when you get the bad friction of spreading an exchange out an hour or two... ;)

I think I understand what you're saying, but I don't think it's what TFA was calling friction, and I don't think it should be called friction at all. I think you're just talking about investing in building skills if they will be beneficial in the long run, and I don't think anyone will argue with that. But calling that "friction" is confusing.

And that's why your original comment that "AI chat is too devoid of friction" or later saying "too little friction is terrible too" still makes no sense to me. You should never add friction, or complain that something's friction is too low. AI chat is a tool like a dough hook. It gives you the answers you need faster than driving to a library or trawling through Google links. Adding friction to ChatGPT makes as much sense as saying you should use a card catalog rather than a library terminal to find a book's shelf. There's enough friction in life already. Don't add more.

A violin is most expressive because it is the one that uses friction (rubbing a bow on strings).

Also, moving the bow just right is harder than slamming keys.

I want some kind of magical piece of string which I can touch to both devices as a way of saying:

"you two, communicate now"

And then later, to break the spell, I'll just touch the ends of that string together.

I don't want to have to dig through settings, I want to manipulate physical objects around me.

However, I think the post is striking a chord because it's pointing to a deeper truth: after 70 years, we are still only scratching the surface of all the ways that humans and computers can potentially interact with each other.

TBH, I'm not especially against the idea, except that, if you make something look like a real-world object, it's important to make it behave like one.

There's a hell of a lot of digital interfaces (not just touchscreen stuff -digital dials and switches can also have the issue), that look like they should behave a certain way, but don't actually do it.

It is also hard to read.

https://worrydream.com/ABriefRantOnTheFutureOfInteractionDes...

I guess one on the reasons why he's building Dynamic Land.

> but we should have made typing feel like painting.

Maybe painting should should feel like painting and typing should feel like typing? I don't know about others but when I type, I just want to type, as efficiently and quickly as possible. I definitely don't want typing to feel like painting.

By the way, loading 92 MB of images to make me read 6 KB of text is brutal!

That's what I get for wanting to read the article before the comments here. Waited minutes only to be greeted with mediocre AI-generated images too, to add insult to injury.

The problem I see here is the more senses you engage - really, the more coordination you demand of the user, the harder it is to make the interface feel intuitive.

[1] https://worrydream.com/ABriefRantOnTheFutureOfInteractionDes...

Great watch.

Inventing On Principle: https://youtu.be/PUv66718DII

A couple other Bret Victor resources I gleaned from other comments:

Humane Representation Of Thought: https://youtu.be/agOdP2Bmieg

Dynamicland: https://dynamicland.org/

When it was purely utilitarian only a minority of people could use it.

When it became tailored to a wider audience it was bland and uninspired.

The further it goes towards art\skeumorphism the more niche the userbase.

The further it goes towards corporate capture the angrier the userbase.

I think we shouldn't pretend like there's a big obvious solution that everyone else has ignored.

Shit I remember reading the Windows 8 designer blogs. Microsoft did serious testing to come up with the most hated interface in their history. From an academic perspective they did a lot of things correctly. Still shit. In fact Microsoft have form here, I remember being one of the few people who enjoyed the original Xbox controller, which was fairly rapidly redesigned due to market angst.

And of course: https://dynamicland.org/

The first few images are impressively well curated and consistent though. I think the "rotoscoping"/"compositing" helps quite a bit to make it feel more cohesive and intentional - certainly it's a step above the usual "blog post header image" slop.

There are definitely issues that pop up but I find it weird people are calling this "slop". "Slop" is that bullshit people are posting where they take the first output and don't pay attention to details. Clearly this person played around with the parameters a lot. The problem with AI Art is taking the art out of the art (and the data theft). But if you're using AI, iterating, focusing on details, then I'm not sure how it is different than any other medium where you do the same things. Art takes time and makes you feel. Considering how it added to the message here and how clear it is that this person put time into the details, I find it hard to call it "slop" just because AI was used.

[0] https://wattenberger.com/thoughts/our-interfaces-have-lost-t...

[1] https://wattenberger.com/thoughts/our-interfaces-have-lost-t...

[2] https://wattenberger.com/thoughts/our-interfaces-have-lost-t...

[3] https://wattenberger.com/thoughts/our-interfaces-have-lost-t...

The claim that current technology is a regression compared to past technology is hard-to-swallow, though. I have a family friend who took CS classes in the early 70s. Punching out 0s and 1s, waiting a day to have a chance to run the program, etc. I do not get the impression that she views this as the pinnacle of enjoyable HCI.

My mom also has a funny story about visiting this friend during the chaos of finals. The friend showed my mom the computer area. My mom vividly remembers seeing students frantically trying to get their final programs done and punch card papers being scattered EVERYWHERE.

Back to the main point, I don't see a lot of extensive, enjoyable physical interaction in this past paradigm. Punching holes in paper would probably get tedious. Carrying the stack of papers over to the mainframe operators would also get annoying. And then you read out the results of your program one page at a time on physical paper. Sure, it's more physical, but is this really more enjoyable in the long run?

So exactly what point in the past is the author reminiscing about?

> We made painting feel like typing, but we should have made typing feel like painting.

I think this quote is worth ruminating for a few while. It reminds me of some Bret Victor talks.