I’ve taught UX or some variant of it for a long time, over a decade. And sometimes students would ask about if I imagine myself teaching UX forever. And sometimes I would joke with them. Well,
Check in with me in ten years.
Right now I teach humans how to use computers to make computers more useful and understandable to other humans.
In ten years I might be teaching computers how to make what humans do more understandable to other computers.
Well if this isn’t just the cutest most wonderful little piece of reporting and storytelling I’ve seen in a good while. Tiana McGee, Daisy Chung, Dea Bankova, Julia Wolfe, Travis Hartman, Holly Murtha, and Claudia Parsons have teamed up to make something truly delightful.
If you’re curious, Neel Dhanesha covers the backstory.
I’ve been playing a lot of these sudoku puzzles over the past year, and I’ll have more to say about them at some point. But this one from Lorena is—there’s really no other word for it: cursed.
One of my goals for this year is to get a better handle on the financial picture of the business at Button School, so this week, I’ve been working with some spreadsheets and trying to figure out ways to understand and visualize the data better for myself.
I have transaction records for course sales in Airtable and Stripe, as well as bank records, all of which I can download as CSV documents.
At first I tried messing around with the data as spreadsheets, which meant trying to piece back together how pivot tables work, and lots of trial and error with the chart generation settings in Sheets and Numbers. I was able to get some workable charts breaking down revenue by month, for example, but I could really only reasonably do this for one source of records at a time, and started running up against a wall when I wanted to combine data sources and do more advanced filtering.
I had another problem: how many steps was I going to have to either write down or remember to perform in order to get from the raw data to a usable visualization?
This seemed like a situation that called for firing up my text editor and writing some code. But how to start?
This is the point where I often get a bit stuck trying to decide how to proceed, thinking about what would be the best way to build something like this. I tend to prefer light-weight solutions that will be easy to maintain over time, but that I also won’t mind throwing away and starting over from scratch. (For various reasons, I also tend to shy away from low- and no-code solutions, since—well—how is that going to be expressive and precise enough?)
In this case, luckily I remembered that the team at Observable had somewhat recently released Observable Framework, an open source static site generator that seemed like it should give me a lot of what I might otherwise look for from something like a Jupyter notebook, but with the benefit of allowing me to work in JavaScript (which I’m more comfortable with than Python) and first-class support for D3 (which produces visualizations friendlier to my soft-hearted designer sensibilities).
I fired up a new project locally with some suggested examples to start, looked through what was there, and then… crickets.
Reading through the docs and looking at the examples, I could see that there were some nice ways to import data from a local CSV file to work with, but being unfamiliar with the idioms and libraries available1, I found myself playing a bit of whack-a-mole guessing at what did what.
It’s 2024, and I decided to try turning to an LLM2. After giving the LLM some context about what I was attempting to do, and some back and forth sharing code snippets that I knew were working, as well as some intended example inputs and outputs, I managed to cobble things together.
Based on this experience, I have a few observations:
I think after this experience I can see more clearly why developers seem to be the most optimistic about the possibilities of LLM technology. That said, I am even more convinced that prompting an LLM to do something that you don’t know how to do yourself would likely result in a maddeningly confusing outcome with much less chance of success.
It may be a bit trite, but I hold to the point of view that technology, and especially computing technology, should augment, not substitute, human thought. I’m curious to see if the development and use of LLMs moves more in this direction.
But I’m not holding my breath.
I haven’t used D3 professionally since I worked with its predecessor, Protovis, more than a decade ago. ↩
I am deeply skeptical of the hype around these technologies, and concerned about their usage for all sorts of reasons, including their impact on labor and the environment—not to mention the speculative economic bubble surrounding them that so smacks of the height of blockchain boosterism, and the questionable legality of hoovering up a bunch of human-created content without permission to feed these machines “data”. However, as a teacher I am very aware that my students make use of them within their studies as well as their work, and I would feel remiss if I didn’t try to understand how they may be useful. If nothing else, if or when I field questions about them, I would prefer my criticisms to be grounded in some practical first-hand experience. If you have made or would make a different choice about this than I have, I certainly don’t blame you. ↩
Every year, I watch more movies than I can possibly remember, which, maybe you can relate? 2024 was a tough year with a bit of an upswing to it for me. Overall, I don’t think it was the best year for cinema, but there were still some standouts for me.
Give June Squib an award! Thelma is Mission Impossible if Ethan Hunt were a nonagenarian widow, holding things together, and holding space for her grandson struggling to launch in his early 20s. Thelma still proudly lives alone, and the action kicks off when she’s scammed out of some money, and her family starts thinking about curbing her independence.
She’s having none of it, and what unfolds is a charming and wonderfully-paced thriller about two people in their golden years proving they’ve still got it.
Stop me if you’ve heard this one before. A teenager gets high with her friends in the woods, and is visited by her 40-ish-yo self, who is Aubrey Plaza.
Yes, with its concept alone, this film is silly and hilarious. What surprised me was how much heart it has, and how deeply and slyly the film turns your expectations against you to deliver a series of gut punches.
To top it off, My Old Ass features some of the most beautiful sun-kissed end-of-adolescence photography I’ve seen in a good long while.
I’m a big old sucker for animation as a form of moving paintings, and whew does this movie deliver. Ostensibly a film about an abandoned android adapting to survive in the wilderness, Wild Robot gives us a much-needed story about the joys and heartbreaks and sacrifices of parenthood.
Oh, and the voice cast is stacked.
Since I stopped teaching full time a couple of years ago, I’ve become an unabashed fan of the work coming out of Dropout. These days I watch most of their game shows, as well as their brilliant productions of standup comedy that often verges on performance art. However, I found my way in through their series of live and recorded games of Dungeons & Dragons called Dimension 20, in particular the Fantasy High series.
Now, Andrea Long Chu has written a fun deep dive into the history and world of the game, where she reflects on her observations as a now practicing dungeon master, which is the person who facilitates these games:
But conventional wisdom holds that talking about one’s campaign is about as interesting as talking about one’s dreams. Most D&D campaigns are private affairs that leave few traces outside of scribbled notes and fond memories. No description can capture the rush of the game or the general anarchy wrought by the whims of the dice and the spontaneity of the players. For this reason, it is difficult to develop a proper aesthetic account of D&D. It is less like reviewing a book and more like reviewing a book club.
More of my work these days is focused on designing courses than on teaching, which means I’m not always as involved with students personally, and Andrea’s perspective holds something that’s true for me about courses too. In my experience teaching a great course is much more akin to running a book club than offering a book. It’s great to be able to offer a good book to folks, and I don’t kid myself: courses designed for learners to go through at their own pace are texts—even if they involve formats like video explainers or tutorial walk-throughs.
What I admire about Dimension 20’s approach to releasing well-produced live games, is that it feels like an in-between of a text and a club. They’ve prepared things with two audiences in mind: the players in the recording, who interact with the facilitator and drive the action and story forward through their characters; and the audience at home, watching all of this unfold. It’s less engaging than playing the game yourself, but comes closer to capturing the spirit and feeling of a live interactive space for what is essentially passive consumption.
I’ll be curious to try a format like this for classes in the future as an experiment, to run a live course with a small group of students who can ask real questions and run into real problems in the moment, allowing them to act as student surrogates for others who can enjoy the course on their own time from home, in a different way.
The question on my mind is: how do you bottle up the urgency of the present moment? How can you make that moment re-livable in a future “now” with its own kind of urgency and potency?
Must-read from Caroline Sinders:
I unsubscribed from 16 online services and documented the ways companies made it, knowingly (or unknowingly), challenging.
A wonderfully nerdy dive into a set of questions about Japanese web design I’ve often asked myself with little satisfactory information to form answers with.
Chat interfaces are having a moment, and Amelia Wattenberger has put together a fantastic essay shedding light onto many well-known and long-standing issues with chat compared to more typical graphical user interfaces:
Good tools make it clear how they should be used. And more importantly, how they should not be used. If we think about a good pair of gloves, it’s immediately obvious how we should use them. They’re hand-shaped! We put them on our hands. And the specific material tells us more: metal mesh gloves are for preventing physical harm, rubber gloves are for preventing chemical harm, and leather gloves are for looking cool on a motorcycle.
Compare that to looking at a typical chat interface. The only clue we receive is that we should type characters into the textbox. The interface looks the same as a Google search box, a login form, and a credit card field.
Of course, users can learn over time what prompts work well and which don’t, but the burden to learn what works still lies with every single user. When it could instead be baked into the interface.
Kirk McElhearn explores and offers thoughtful ciriticism of Apple’s new app for classical music. It’s a great case for deeper information architecture work.
Luke Wroblewski reflects on an oldie/goodie:
So how can you tell when there’s the right amount of visual hierarchy in a UI design? Just squint. This will blur the design just enough to quickly identify if the important elements stand out. I call this the squint test.
Accessibility in design is a form of empathy: trying to reach beyond your own personal perspective to try to understand other people who, in this case, very literally don’t see the world the same way you do.
Blah blah blah web components. Something something relative color syntax. All I really care about is that they fixed the font-optical-sizing: auto bug. It’s been driving me bonkers for months!
From Maxwell Neely-Cohen’s Letter from the Editor:
The growth at all costs mindset endemic to much of the web’s infrastructure has inevitably led to bait-filled search engine results and thin-skinned billionaires buying entire social media platforms to promote their own egos.
Meanwhile, literary magazines, including many of our favorites, are shuttering at a terrifying rate, often due to the capriciousness of their wealthy patrons or corporate parents. Book publishing is beset with megamergers, dreams of monopoly, the threat of private equity vultures, all while cultural and critical institutions that still overwhelmingly support only a small cohort of the connected.
As friends and colleagues Maris Kreizman and Lincoln Michel have observed, there are seemingly endless attempts at publishing startups, techbro pleas for “enhanced” books, and hastily written AI stories spamming the submission accounts of magazines, all promotions of the worst possible forms of exploring the relationship between digital technology and literature.
Kyle Chayka, a writer we love, recently remarked “can we please build a more artisanal internet / media landscape” (in response to a depressing analysis of the future state of the media business in light of AI text generation hype, from Ryan Broderick, another writer we love).
We hope the html review can be a small part of that future landscape. One of many alternatives, along with our growing list of friends, compatriots, and idols.
Always happy to see folks keeping the web weird.
Molly White is on the case.
If anyone out there is looking for a great RSS service, I can very happily recommend Feedbin. I’ve been a satisfied customer for 10 years, never had any problems. I mostly catch up on feeds from Reeder on my phone. After messaging, it’s my most used app—a steal for only $5. I also use NetNewsWire on my laptop sometimes, and can definitely recommend it if you’re looking for a free option.
Congratulations to Betina Hsieh, with some thoughtful reflections as she prepares to take on a new role as professor of teacher education at the University of Washington:
I now see myself as a researcher, something I might not have said 10 years ago. I have been able to explore research on teachers and teacher educators in ways that have moved me and have helped me learn so much about myself, about teacher candidates and teachers, about teaching (my own and that of others), and about the ways structures and systems can often act to perpetuate the push up and push out of so many incredibly talented people from classrooms.
I am working on trusting myself and trusting my community, trusting the faith they have in me and that their love and respect are well-placed. I am working on the grace and humility necessary to respond (rather than react) when I am called-in and pushed to grow. I am working on trusting that the right opportunities open up at the right times, and it’s not for me to decide that I’m not good enough.
Sarah Scire in conversation with Holden Foreman:
Foreman makes the case that accessibility in journalism is important for everyone: Making news products more accessible, after all, often means making them more user-friendly and efficient. He hopes to discover and standardize ways of making the Post’s journalism accessible to as many people as possible.
More dedicated roles like this please!
A love letter from Srini Kadamati.
Great write-up from Paige Nuzzolillo.
Megan Erin Miller provides a great overview:
Service designers have to be nimble and able to see the big picture and zoom into the details. As a service is made of up a wide variety of touchpoints (points of interaction between the customer and business — think: website, app, phone call, billing statement, store, shipping notifications, etc.), by nature, we have to be able to zoom into the touchpoint level. But in order to engage the business in orchestrating a multi-touchpoint relationship with the customer (the service experience), we have to be able to engage in strategic discussions and planning to set the conditions for success for the “encounters” with the service.
I feel like something like this could also be helpful for information architecture…
How did I not know about these? Sometimes I could really use a poopnado or mer-unicorn.
Bartosz Ciechanowski:
In this article, I’ll focus on the delicate interplay between many of the forces that act on a bicycle and its parts when riding. We’ll witness how forces applied through tires make a bicycle accelerate, brake, and turn, and we’ll also investigate how the wheels and the frame handle those different forces without breaking.
These are absolute catnip for me. My other favorites in this series: Mechanical Watch & Cameras and Lenses.
Sheon Han covers the history of blind programmers buiding assistive technologies for blind and partially sighted users:
Nearly three decades have passed since JAWS for Windows was released, during which possibly tens of thousands of blind and partially sighted programmers entered software development. Just as it was in Henter’s time, it’s a field that is relatively inclusive for people who are blind, as the accessibility barriers are lower than in many hands-on jobs. These days, this is in no small part thanks to JAWS, a piece of software pioneered by a blind programmer.
From Devon Govett and the design system team at Adobe:
This includes a full suite of fully featured components and hooks including calendars, date and time fields, and range pickers, all with a focus on internationalization and accessibility. It also includes @internationalized/date, a brand new framework-agnostic library for locale-aware date and time manipulation.
If you’ve ever worked on the design or implementation of components for choosing dates, I’m sure you’ll appreciate the incredible amount of work and attention to detail this represents.
I’ve been teaching design systems and accessibility courses lately, and there’s an amazing amount of overlap these days. This work from Adobe is a great example of why: building out flexible components which can be reused throughout an organization or even beyond suddenly changes the impact of accessibility efforts for those components.
David C. Brock celebrates the legacy of the earliest attempt to commercialize a personal computer with a graphical user interface:
The people who developed the Alto came to Xerox PARC from universities, industrial labs, and commercial ventures, bringing with them diverse experiences and skills. But these engineers and programmers largely shared the same point of view. They conceived and developed the Alto in a remarkable burst of creativity, used it to develop diverse and pathbreaking software, and then moved out of Xerox, taking their achievements, design knowledge, and experiences into the wider world, where they and others built on the foundation they had established.
The photos accompanying this article are a real treat. If you’re interested in going back just a bit further, Doug Engelbart’s “Mother of All Demos” presentation in 1968 was a major source of inspiration for PARC.
Dan Klammer has done the legwork of putting together a comprehensive and highly usable set of system fonts available on all the major platforms, organized by classification. Using system fonts means no font downloading, which means faster web sites and web apps, and no re-rendering when fonts load.
Accessibility is unfortunately still an afterthought on many projects. User interaction and accessibility requirements are poorly documented, at best. Or forgotten, when handing over designs to developer teams. And fixing it later costs a LOT more than building it right to begin with. Great documentation helps teams implement accessibility requirements the right way. I will tell you why, what and how designers can document different aspects of accessibility and user interactions requirements, to build better more inclusive products.
Highly recommended.
If you prefer, Stéphanie has also given this as a talk a few times. I prefer the recording of her talk at Inclusive Design 24.
I had the pleasure of meeting Jukay and his team at Pursuit in 2017 before I left New York, and I was very impressed with the work they were doing to lower access barriers for professional tech education. Here Jukay discusses some of how they make that work:
One shining solution is called outcomes-based financing, a model where workforce development programs tie payment to a trainee’s employment trajectory, rather than charging fees up-front like colleges and universities.
Here’s how it works: thanks to funding from social impact investors through “job bonds,” participants don’t incur any cost during their training – not a single dime up-front. Then, after a graduate secures a high-paying job, they pay a percentage of their earnings for a set amount of time to help repay the bond’s debt service. If someone doesn’t immediately get a job or if their salary is below a certain level, they pay nothing. If they stop working for any reason, their payments pause, too.
There are a lot of people who talk big about “the future of education.” I’m not sure whether the ideas Jukay presents here would make sense outside of the space he’s been working in, but I do think it’s worth paying more attention to folks like him who have been doing the legwork to figure things out.
At some point, I personally hope the community college systems in the U.S. will catch up in terms of serving these needs outside of private foundations and companies.
I love reading about nerdy design processes like this one from Alex Hollander:
In other words, Wikipedia — a major, legacy website (top 10 ranked, for 10+ years) — had an interface that hadn’t been changed for 15 years. And then one day the Chief Product Officer came to our team (1 product manager, ~6 engineers, 1 quality assurance person, ½ a scrum master, ½ a data analyst, ½ a community liaison, and myself), and tasked us with making significant improvements. It might honestly be a once-in-the-history-of-the-internet kind of situation. Exciting, but rather difficult.
This is a great example of the power of storytelling, process, and emphasis on data and results in talking about the effectiveness of design work. If you’re working on a portfolio right now, I encourage you to take some notes on this one.
Bite-size selfie video tips from user experience researchers. As I describe it, I realize it doesn’t sound like it would work nearly as well as it does!
Great work from Will Soward documenting considerations for neurodiverse learners based on his work on Learning Management Systems.
Michelle Ngo recently shared this short documentary on designing inclusively that she worked on. Tight and well-edited, with so many great stories. I especially loved the fashion designs for kids, and the efforts to bring young people into the process of designing for their needs. Great stuff!
Mandy advocates for digging deeper when facing burnout:
So how do we approach that repair? I think we have to first name the experience. Acknowledging your burnout (or that of your teammates) is one step, but the next is to unpack what that means. Maybe (probably) you’ve been working for too hard for way too long. Maybe it’s a loss of faith in the work itself—that something that had seemed important or valuable to you no longer does. Maybe you’ve been doing the same work for a while now, and have grown bored with it. Maybe you feel isolated or unsafe or like you can’t influence decisions. Maybe the way you think about your work has changed, and you don’t know where this job fits in anymore. Maybe it’s all of these things. The point is that each of these subjects—overwork, loss of faith, boredom, isolation, lack of autonomy, etc.—are all more readily addressable than burnout is on its own.
Nadyne Richmond makes a great point about keeping decision-making in focus when planning, conducting, and evaluating user research:
Defining the decisions that will be made as a result of the research can be a powerful tool when a team is using research for the wrong reasons. I’ve seen teams become so risk-adverse that they want to test every single thing. I’ve also seen teams become so beholden to a process that they don’t know when or even how to deviate from it. In those cases, defining the decision that will be driven and why that decision matters illuminates the more important problem that needs to be addressed within the team.
This subset of css that these tools use is based on how websites look today. Or on how they looked in the last ten years — which is the same thing. Websites haven’t changed. This means that with these tools we can design websites that look like websites-as-we-know-them, but it is much harder, if not impossible, to come up with new ideas. If we accept that these are the tools we will use from now on, we accept that the web looks like it looks today. And behaves like it behaves today.
I’m not sure I agree with this. As a teacher, I find it empowering for students that they can use tools like Figma with a fuller sense of confidence about the implementability of their designs. I think it’s equally arguable that the convergence of our tools with how we design is a sign of the maturity of the field.
If anything, I find myself wanting design tools to adopt more of the layout technologies available in rendering engines, not less—especially for things like grid-based layouts and tabular data, which most interface tools handle like a minor nightmare.
Maybe I lack the vision or imagination to dream up web designs that expand far beyond what we’ve acheived so far. That said, it’s not hard to find cases where design tools work so much faster to try out ideas that would feel slow or impossible to realize quickly in code:
I want to see our design, production, and content tools improve, but I don’t think building in CSS-like features is inherently holding us back.
I’m trying my hand at self-directed tutorial videos, starting with a series on using Figma to design user interfaces.
Julia Evans posted some comics about CSS last year:
I’ve been writing a tiny bit more CSS recently, and I’ve decided to actually take some time to learn CSS instead of just flailing around and deciding “oh no, this is impossible”.
CSS feels a little like systems programming / Linux to me – there are a lot of counterintuitive facts that you need to learn to be effective with it, but I think once you learn those facts it gets a lot easier.
Julia’s takes are always genial and practical. Have you ever seen a more appealing breakdown of CSS selectors? No, no you have not.
Ink & Switch offer an overview on the history, current state, and possible new approaches to providing end-users with more power over computation through various approaches of extensibility. From the introduction:
This article tackles the question of why this is so. We’ll start by describing three qualities we think are important for end-user programming: embodiment, living systems, and in-place toolchains. We’ll survey the prior art and try to illuminate what has made this problem so immensely difficult. Then we will document the experiments we’ve done at the Ink & Switch research lab in adding automation and customization capabilities to a digital sketchbook application.
We believe in a computing future where programming is within the grasp of everyone and hope this article can inspire and challenge our industry.
Today another class graduated, and I feel a mix of proud, tired, nostalgic, hopeful, and sad. This time, however, I find myself turning my attention to teachers I work with, more than to students.
This week I learned that two of my fellow teachers are stepping away from teaching to work on curriculum. I am happy for them, and excited to see what they will do, but I pause to reflect on my own past.
After I had been teaching full time for a few years, I moved into a non-teaching role, managing and coaching teachers. I jumped at it despite being unprepared: my school trained me in management (which I enjoyed) and teacher coaching (which I was very skeptical about). I was tired of the grind of teaching, and moving into a management role seemed like a good way to support teachers.
Ultimately school management burned me out in a way that teaching never has. I eventually left that school to take an industry job, only to return to teaching after a year. These have easily been the most productive and effective teaching years of my career—thanks to my time spent observing and coaching other teachers; training sessions I partcipated in to prepare for managing teachers; and collaborations with co-teachers and TAs who have taught, challenged, and inspired me.
At my current school, most of the teachers stay for only a year or two. One promising teacher I worked with burned out after six months and moved into an administrative role; most others have left the school and returned to industry practice. The past year I’ve noticed less turnover of teachers. I think many have simply stayed to avoid searching for a job during a pandemic. (The irony that we prepare students for a job search during the same pandemic is not lost on me.) I find myself wishing that the teachers who have stayed may discover that the job gets easier and more rewarding, and stick with it.
I’m hopeful that my colleagues who are moving to work on curriculum will feel challenged, will learn and grow in new and unexpected ways, and that they will find approaches that help our school’s teachers and students.
But in my heart I hope that they will find their way back to teaching.
Josh W. Comeau delivers what it says on the tin:
This tutorial is meant to be accessible to developers of all experience levels. It can be thought of as “CSS transitions 101”. That said, I’ve sprinkled in some interesting and obscure tidbits — no matter your experience level, I bet you’ll learn something!
I am a total sucker for this kind of thing.