Anne Gibson

7 October 2023

Not-confidentially to the folks at deck.blue

For those of you not on blue sky (the social media app) or even some of you who are: there’s an app called deck.blue that serves up blue sky in a tweetdeck (multi-column) interface that I’m trying out because the blue sky app won’t hold the place that I left off and let me go back to it.

Turns out deck.blue won’t do that either (is there a tweetbot for bluesky yet?) but while discovering that, I also discovered it has, well, janky scroll animation. I’m not sure I want to call it scrolljacking since this whole app (and all the other social media apps) is steeped in so much scripting that they all seem to write their own scrolling, but I digress.

When I asked them to stop it they asked me to clarify and I took a video of my screen and, well, here we are.

Hi! Sorry I’m posting this on my blog but deck.blue won’t let me upload a .mov file or an .mp4 file of my screen so I’m posting here.

I wrote:

see caption text below image for screenshot content — Dear blue sky tweet deck thingie: STOP HELPING ME SCROLL the jumping around is making me irrationally angry

You replied:

So let me clarify a few things because I’m not sure we’re talking about the same thing and even if we are, I’ve done enough tech support to know me setting context will help you.

I’m on Safari 16.6 on a MacBook Pro running the latest Ventura update.

I’ve got active accounts on twitter, Facebook, LinkedIn, tumblr, three Mastodon instances, Blue Sky, tiktok, Instagram^[1]technically, it’s so damn annoying to use I avoid it and probably a few dozen come-and-gone social media accounts.

With the exception of tumblr^[2]because tumblr’s threading mechanism is much more thorough than anyone else’s, mostly, plus I’m not reading the news on there, and tiktok ^[3]which doesn’t have anything one could call an “order” it’s just chaos you can flip through I read all of them in chronological order. Not reverse chronological order, actual chronological order.

In Tweetbot, on twitter that was easy because when I closed the app, it remembered what tweet I was looking at, and when I reopened the app I was still looking at that tweet, just with a lot more new tweets loaded “above” the screen that I had to scroll up to.

Since Tweetbot’s demise, my usual routine is to open a site, scroll down my feed until I recognize something, then slowly scroll up reading each post in turn, until I reach the top again. That way the posts make sense and I get to see everyone I’m following. (I don’t follow a lot of people on any given service because I learned on twitter that I’ve only got the time to read about 200 people worth of posts a day.)

What’s the issue?

The problem with deck.blue isn’t scrolling down, it’s my slow scroll up. I’ve got all the most recent skeets loaded loaded in memory I presume, because i just scrolled past them to get to my starting point. And as I scroll up, things seem smooth for a bit. Then suddenly the feed will jump down like half the distance of the column.

Alt text for video: Video of the column scrolling smoothly, then jumping position, then scrolling smoothly, then jumping position again. No audio. Sorry this alt is so pants, I’m still trying to figure out how to get wordpress to cooperate some days. Similar apologize if the video is taller than your browser.

So why’s this an issue?

First, as Nielsen-Norman point out about scrolljacking:

When it comes to scrolling, users have strong mental models: they expect to scroll vertically, at a consistent rate that is related to how they are physically interacting with their input device.

In most computer operating systems, the default scroll speed can be manually adjusted to be more sensitive or less sensitive, but it is always consistent. Users expect a consistent scroll rate across all the applications on their devices.

When a website alters the default scroll functionality, it “hijacks” the user’s control over their device and can generate disorientation.

So right off the top this is a usability issue because it breaks the user’s mental model, and when my mental model is broken I get grouchy and annoyed.

Second, there are a lot of people who find it easier to read when they have a strong visual anchor that they can associate with the line of text they’re currently reading. For example, people with dyslexia often find it easier to read printed (paper) text when they have a reading ruler to help them focus on a single line of text at a time.

A reading ruler works by reducing visual distractions and increasing focus. People with dyslexia often experience visual stress when reading, which can make it difficult to concentrate on the text. A reading ruler can help to reduce this stress by blocking out surrounding text and highlighting only one line at a time.

Now, I don’t have dyslexia, but I do find it much easier to read when I can reduce visual distractions and increase focus. I’m pretty sure everyone does. Over the (many) years I’ve worked on computers and the web, I’ve made it a habit to pick a specific location — usually the top of a window, but sometimes further down, with my pointer right next to the text — that marks the line I’m reading. Instead of reading down the page, I use the scrollbar to move the line of text I’m reading to my “reading ruler” point.

So when I’m reading something and then because of inconsistent scrolling speed a totally different set of content has moved to my “reading ruler” point, I get really damned frustrated. Now I have to scroll down, find what I was reading, move it back to my reading ruler point, and then continue reading — until it happens again. ^[4]This is also why sites that play hide-show games on scroll with their global headers can get in the fucking sea. Looking at you, Medium. STOP MOVING MY READING RULER YOU BASTARDS.

As a designer, then, I find this technical “glitch” to be bad design because it undermines the user’s mental model. As an accessibility professional, I have concerns that it increases difficulty and/or destroys the ability for someone with a reading disability like dyslexia to follow the content. As an armchair developer I have no idea why something would unevenly scroll when the content’s already loaded. ^[5]I suspect it has something to do with the skeet height? But have no idea why?

And as a user, well, I’m really hoping this is something you can fix, because I think it will give you a much more effective application. Bonus points if you can do the same thing as Tapbots and let me read in chronological order without having to find my place every time I launched the app or walked away for a few minutes.

Anyway, all of this isn’t to name-and-shame, it’s to say hey, here’s the problem I’m having, I can’t post my video on the app so please excuse me for using my blog, and maybe the people following along will learn some stuff about scrolljacking and accessibility and dyslexia while they’re here. Thanks for listening!

Footnotes[+]

Footnotes
↑1	technically, it’s so damn annoying to use I avoid it
↑2	because tumblr’s threading mechanism is much more thorough than anyone else’s, mostly, plus I’m not reading the news on there
↑3	which doesn’t have anything one could call an “order” it’s just chaos you can flip through
↑4	This is also why sites that play hide-show games on scroll with their global headers can get in the fucking sea. Looking at you, Medium. STOP MOVING MY READING RULER YOU BASTARDS.
↑5	I suspect it has something to do with the skeet height? But have no idea why?

30 September 202330 September 2023

The role of dogs in UX Design

It’s important to own a dog (or five) because dogs are Nature’s way of saying “Hey! It’s 1am. If you had logged out of work on time you wouldn’t need to go stand in the rain and cold now. So get moving when they tell you to next time!” ^[1]Cats may be just as useful. I’ve never had one. None of my cat-owning friends have been dragged out into the snow at 3am by a kitten that got too many naps while Mom worked and wants to take an … Continue reading

Terriers Myka and Zoë standing in the rain on watch for rabbits. — 1:20 am is the perfect time to hunt rabbits.

Footnotes[+]

Footnotes
↑1	Cats may be just as useful. I’ve never had one. None of my cat-owning friends have been dragged out into the snow at 3am by a kitten that got too many naps while Mom worked and wants to take an hour to find a spot to pee though.

22 June 2023

Design is a Job by Mike Montiero

First, a note: this review is of the first edition of the Design Is a Job, not the recently-released second edition. I’ve read a lot of A Book Apart books and I can tell you that their second editions have never produced a worse book than the first edition — and the first editions are always excellent. So go get the second edition if you can.

I bought this book when it came out because I saw Mike give a presentation about being a good designer, probably at An Event Apart. I then stuck it on a shelf. That was about 10 years ago, when I was still an arrogant snot-nosed designer raised in the Philly area who thought she knew it all. Considering Mike’s reputation for much of the same (including the Philly) someone should have shoved the book in my hands, chained me to a chair, and said “LEARN.”

Now I’m 15+ years into my career as a UX Designer and knowledgable enough to know that if I want to to keep being an arrogant know-it-all designer I have to do better at knowing it all, and I have to work on my delivery. There’s only so many times in your life that you can get away with having “Must work on communication skills” will show up on your review before everything else about your growth won’t matter. Fortunately, I learned that a few years ago, and now I’m actually listening to people and, damn, that makes work easier.

This book will teach you those lessons faster, if you’re willing to learn. If you’re willing to learn, this will teach you how to work with a lawyer, what belongs in your contracts, how and why to talk about money, and how to present. If you’re not willing to learn, this book will present reasons on why you should be willing to learn, which hopefully would have penetrated my younger thick skull.

Mike’s examples are clear and concise, truthful while still keeping everyone’s privacy appropriately, and constructive. Also, he’s funny.

Talking to teens about sex is a lot like talking to designers about contracts. “We’re being careful. We’re in love. We trust each other. They have an agile process. He promised there wouldn’t be any backend development.

More than anything else, what Mike brings to this book is his desire to see you succeed as a designer, whether you’re working inside a big corporation or as a consultant or at an agency or on your own. He emphasizes treating our peers in the industry with respect, so that we can all raise the quality of design in the industry and, presumably succeed even more.

If you are any kind of designer, content strategist, product manager, or even engineer, read this book. You’ll find ways to improve your working relationships with others as well as ways to produce and execute better ideas, wherever you are.

2 October 2022

You Should Write a Book by Katel LeDû and Lisa Maria Marquis

When I first started working in User Experience, my mentors explained to me that part of my responsibility as a designer was to pass on to others what I had learned. My team exemplified that ethic: they were running the IA Summit that year, most of them had given at least one presentation about the things they’d learned at work, and one was writing a book.

“Someday, that will be you,” they said. “You should write a book.”

I was shocked years later to find out that not all UX teams or leaders believed it was our responsibility to pass on what we’d learned. Still, I knew enough people who, like me, felt that passing on our knowledge was one of the responsibilities of a good designer, so I found myself writing a column on a design-related site, then launching my own design-related site with some friends, then giving talks about the things I’m passionate about. But finding time to write a book? Hah.

Now I’m in a situation where the time has been thrust upon me. I don’t particularly like public speaking, and I’d much rather not attend a mess of conferences in our not-quite-post-COVID 19 world. Writing a book seems like a reasonable choice. Still, I am not usually a person to take big risks.

You Should Write a Book by Katel LeDû and Lisa Maria Marquis is a critical book for people like me, who don’t want to move forward without some guideposts and who don’t want to bug their friends for hints or tips.

It does not go into depth about every single step. (There are lots of resources that will provide that for you, and they are dependent on the choices you make on how to write, polish, and publish your book.)

It does give you the map of the process, the understanding of why each of the process steps is there, what differentiates one choice for another, and a pretty deep appendix of next-step sources.

It is also intentionally written to encourage both non-marginalized and marginalized writers. The authors acknowledge in the beginning that they are “college-educated, able-bodied, cis white American women whose direct book-publishing expertise derives mostly from a single organization: A Book Apart. While we hope our perspective is useful, it is also bound to be limited in certain respects.” These authors essentially say that the process for publishing a book is the same for both marginalized and non-marginalized authors, except that that marginalized authors have to do it on hard mode.

This book will not tell you how to choose whether to self-publish or not. It will tell you about self-publishing and what the differences between it and traditional publishing are. It won’t tell you where to find editing or marketing services. It will tell you what they are and why you want them. It won’t tell you what to write about. It won’t tell you what anything costs — those numbers go out of date too quickly anyway. It will gently laugh at you if you think that you can retire on the money from one book.

Because the authors are deeply and intimately familiar with the publishing house A Book Apart, where they both work, they are able to not only give you insights into ABA’s publishing methods but also give you insights from ABA’s authors, who they’ve spoken to about the publishing experience.

In short, all models of the publishing industry are wrong, and this one is useful.

29 August 202129 August 2021

The Clock of the Long Now, by Stewart Brand

I’m thinking about writing a book, and the ^[1]nonfiction book I’m thinking about writing is about building design systems and making sure they’re accessible. ^[2]The list of fiction books in progress is embarrassing.

It’s a big undertaking. ^[3]Building a design system, I mean, not the book. Although I think the book will be a big undertaking too — it’s just that books tend to be finite creatures that you can finish, and design … Continue reading

It’s an undertaking that I’ve been involved with for three different companies now — as a consumer of the system at Vanguard, and as a designer at both Boomi and Vertex.

But really, I wouldn’t have made the jump from consumer to designer successfully at all if Jeremy Keith hadn’t given a talk in 2018 called The Way of the Web. ^[4]It is not one of his best-known talks and for that matter if I hadn’t found these sketchnotes I’d really be wondering if I made it up right now.

In that talk, Keith summarized the concept of pace layers, presented by Stewart Brand. It surmised that culture builds from a bottom layer of nature (which changes very slowly) through multiple layers until it reaches fashion (which changes almost too rapidly to track), and that by understanding the relationships between the layers, we can better understand how and where to effect change.

Keith’s suggestion in his talk was that a similarly-structured pace layer could be built from TCP up through Javascript and other scripting libraries, from the slowest changing layers of the internet through the fastest.

At the time I was prepping for my own talk so the idea kind of grabbed on to the back of my brain… and there it started gnawing on things.

Building a design system is often challenging and frustrating because some things about it (such as, for example, the organization’s voice and tone, or the corporate font) change rarely, but have huge impacts. Others, such as the fashion of flat design or the implementation of specific interactions, change frequently but only impact small subsets of the full design system. As a designer, I’ve found that understanding the difference between scheduling a change to the corporate font and scheduling a change to the 4th carousel owned by the 2nd subdivision of importance in the organization is the key to ensuring that the design system is given the budget to grow — and the key to ensuring that the design and development partners don’t want to string me up.

But to explain all of that, first I had to be able to explain pace layers. And while I was pretty sure I did understand them, if one wants to start writing a book it’s a really good idea to read the source material before running one’s mouth.

Then I bought the wrong book.

See, my brain is a muddled mess of memories from various conferences and web events. Learning about pace layers is definitely one that sticks out. Another that sticks out is discussion of the Long Now Foundation.

This foundation is dedicated to taking the long view on time. The very long view. Like, a clock that runs for ten thousand years long. People who place bets that take a minimum of two years to prove out true long. Storing a sample of all the world’s languages in nickel long.

And while The Clock of the Long Now does, indeed, discuss pace layers in chapter 7, the bulk of the book is about taking the long view on what we put into the world. In the case of web design and development this can mean everything from choosing HTML as your primary markup language (and not coding everything in JavaScript) to deciding which companies and projects you’ll still be proud to be associated with 20+ years into your career.

(I also remember a lot of conversations about how future forward design is progressive enhancement from older technology, and a lot of men getting their pictures taken in a space helmet, but once again the terminology to find proof of these things eludes me tonight.)

It turned that reading The Clock of the Long Now was not a mistake on my part (more of a lucky accident) because it’s helped me understand and maybe even figure out how to explain that one of the biggest challenges with a design system is that it’s a system. You’re designing a system that, if successful, cuts design and development time for the whole organization. But that means you’re also designing a system that will require time to develop, time to adopt, time to grow, time to stress-test. It is not a short-term two-quarter project with a payout at the end.

In other words, it works in long time, but in corporate-driven show-me-the-returns time, it may be hard for our Product people to buy into.

One of the amusing things about the book is that the edition I read was published right after the turn of the millennium (or the year 2000 for you pedants, don’t @ me about when centuries start), which is itself a millennia in web years. The Long Now Foundation still exists, which is a bit of a relief because basing my ideas on design systems on a long vision that subsequently died in the 20 years since the book was written would’ve been embarrassing.

Brand did seem to go out of his way to not make predictions about what the future would bring, outside of the fact that it 10,000 years from now life will likely be very different from today. Still, there was more than one point where something said would pull me out of the book. Brand seemed quite optimistic, that the web would have the power to prevent people in power from spreading lies, for example, at which point I almost spat my drink across the room.

The book’s thesis, that we should take a long view if we want to work together to solve problems, is still valid 20 years later, and has helped me better shape the kinds of discussion points I’ll need to get the funding necessary for a quality design system. (I’m pretty sure Brand didn’t see that coming, but then, he probably wouldn’t be surprised either.)

So it’s a good, relatively short read, not directly about the web but definitely about the underpinnings of being a developer or a designer, and I recommend it to anyone interested in the longer view.

Footnotes[+]

Footnotes
↑1	nonfiction
↑2	The list of fiction books in progress is embarrassing.
↑3	Building a design system, I mean, not the book. Although I think the book will be a big undertaking too — it’s just that books tend to be finite creatures that you can finish, and design systems are not so much.
↑4	It is not one of his best-known talks and for that matter if I hadn’t found these sketchnotes I’d really be wondering if I made it up right now.