📚 Kamal's Readwise Wiki
Concept

UI/UX and Visual Design Craft

The hands-on trade of interface design — typography, color, speed, microcopy, user research, and the emotional pull of products — drawn from the library's largest professional cluster.

designuxuitypographyfrontend

The single biggest professional cluster in this library is design craft: not design-as-theory but the day-to-day work of making interfaces that feel right. The highlights circle a handful of convictions — that design is how it works, not how it looks1; that speed and reliability are moral commitments, not optimizations2; that words are half the interface3; that research means finding the unheard voice rather than the loudest4; and that products ultimately win on feeling, not feature lists5. What follows organizes the flagged passages into the craft's working parts.

Design is how it works

The philosophical spine of the collection is Jony Ive and Steve Jobs at Apple. Jobs's line is quoted in full: "Most people make the mistake of thinking design is what it looks like… That's not what we think design is. It's not just what it looks like and feels like. Design is how it works."1 Ive's contribution is a suspicion of style itself — "style has a corrosive effect on design, making a product seem old before its time"1 — and an ambition for the work to vanish: his "ultimate goal is for his designs to disappear," happiest "when the user doesn't notice his work at all."1 The through-line is care over carelessness: "What I really despise is when I sense some carelessness in a product."1

That craftsmanship ethos recurs outside software. Sigma's minimalist BF camera is admired for embodying kodawari, "the uncompromising commitment to perfection" — seven hours of five-axis CNC machining per unibody, nine cameras a day6 — and for teaching through excellence rather than volume: "A single masterpiece can teach us more than any number of the mediocre products of a given period or school."6

Speed and reliability are the craft

Craig Mod supplies the library's strongest argument that performance is design. Speed is "a good proxy for general engineering quality"; if an app stumbles on simple tasks, "it can mean the engineers aren't obsessive detail sticklers."2 His ideal is software that unbloats over time — Sublime Text handling a 50,000-line file and only getting faster over a decade, growing "smooth over like a river stone."2 The counter-case is Google Maps "dying a tragic, public death by a thousand cuts of slowness."2 His companion essay reduces it to a formula, "Consistency + reliability = fluency,"7 and pleads for companies to elevate the "boring nuts and bolts… to first-class rank" rather than shipping splashy masthead features.7 On screen, he notes, "to feel tactile is to move without latency."2

Words are 50% of interfaces

Microcopy gets treated as a first-class discipline. "Words are 50% of interfaces,"3 and good UX writing role-plays the interface as a conversation: a product asks, the user acts.8 Three concrete rules recur:

  • Don't over-promise. Prefer "Send a code" to "Get a code" — you can't guarantee the SMS arrives.8
  • Reason before instruction. "To get your results, provide your email" sticks better than the reverse order, which forces a reread.8
  • Speak in the user's frame. "Until Jan 31, you can deposit $400 more" beats "Your deposit limit is $1,000 per calendar month."8

Mod extends this to error dialogs: macOS shifting "Don't Save" to "Delete" is a copy regression — "Pressing delete feels violent."2 Jason Fried's wordplay makes the same point about tone: reframe the FAQ as a "YesAQ" — the questions you say yes to.9

The micro-craft: type, color, and pixels

A large share of the flagged tweets are pure execution detail — the swipe-file layer of the trade.

Detail The craft note Source
Numerals that jump when animated font-variant-numeric: tabular-nums makes figures monospace so sliders don't jitter 10
Typography polish en-dash for ranges, m-dash to break clauses, no indent on first paragraph, indent or spacing (never both) 11
Bullet vs. middle dot knowing • from · is part of typographic literacy 11
Reference-grade UI Cron praised for color contrast, monospace figures, en spaces 12
Typeface pairing GT Sectra + Simplon Mono as a considered voice 13
Realistic buttons Campsite-style depth from just 1 stroke + inner shadow + drop shadow + gradient 14
Color, systematically "simple, scientific rules" — 10 practical tips because everything you design needs a color 15
Advanced CSS Linear's animated component reverse-engineered as 10 blend functions and "a couple lines of JS" 16

The tooling layer shows up too: Erik Kennedy's tip to keep an iOS template with pre-cut slots (status bar, title, search, tab bar) to speed layout17; Tailwind block libraries like oxbowui's 427 free components18; and, at the newest edge, Claude Code's frontend-design skill for "beautiful green field apps."19

Research: chase the product, not the data

The research highlights are unusually opinionated, and they lean anti-metric. Robin Rendle's formulation is the sharpest: "You can only build a great product if you care more for the vibes than for the data,"4 and research earns its keep by "finding the people who are not being heard, and adding their voices to the conversation."4 He is scathing about consensus-by-numbers — "wacko pseudo-scientific ways to prioritize tasks and even whackier ways of measuring the success of a product (NPS, questionnaires, etc.)" — arguing great design "can only thrive with a dictator/director" who can say ship.20

The practitioner books ground this in method. Just Enough Research defines research as "simply systematic inquiry" and insists "where humans are concerned, context is everything," with ethnography's core question being "What do people do and why do they do it?"21 Build Better Products supplies the machinery: the User Lifecycle Funnel ("in reality it's more of a sieve"), measurable business goals, and the discipline of interviewing to listen, not sell — "you're going to learn new things, not just confirm what you think you already know."22 Both converge on the field's most durable maxim, phrased by Carter Bryden via DHH: "When a user brings you a problem, they're always right. When they bring you a solution, they're usually wrong."23 DHH's own version: "Work on behalf of users, not by their request… dig for the problems."24 And the technique that unlocks it, from Teresa Torres — ask for stories, not direct questions.25

mindmap
  root((Design craft))
    Substance
      Design is how it works
      Make design disappear
      Speed = fluency
    Words
      Interface as conversation
      Don't over-promise
      Reason before instruction
    Micro-craft
      Tabular numerals
      Type & color rules
      Depth from shadows
    Research
      Problems not solutions
      Ask for stories
      Vibes over data
    Emotion
      Feelings not features
      Tribal markers
      Design for screenshots

Feedback, critique, and defending the work

Two books zero in on the interpersonal craft of shipping design through an organization. Discussing Design separates real critique — "analysis that uses critical thinking to determine whether a design is expected to achieve its desired objectives" — from mere reaction or direction-giving, and describes the designer's mind as "a toggle, switching between creative thinking… and analytical thinking."26 Articulating Design Decisions reframes stakeholder conflict as usually not conflict at all: "the majority of issues or concerns that our stakeholders bring… are often just a matter of misunderstanding or miscommunication," and warns that "when we get defensive, we fail to focus on the real issues."27

Emotional products eat rational products

The library's most-highlighted product essay argues that features lose to feeling. "They don't just solve problems, they create feelings," and the winning emotional loop runs "feel inadequate → discover thing that cool people do → join thing → feel transformed → evangelize thing" — versus a rational loop that "starts with a spreadsheet & often end[s] with… a slightly better spreadsheet."5 Aesthetics become "ideology made visible"; design choices are "tribal markers" (brutalist = technical, pastel gradients = progressive, dark mode = serious).5 The operative instructions: "design for screenshots" — every screen is marketing5 — and build "the product that makes people feel the most interesting version of themselves."5

The behavioral engine behind this is Hooked: Fogg's model that any behavior needs motivation, ability, and a trigger28; the primacy of internal, emotional triggers ("boredom, loneliness, frustration… prompt an almost instantaneous and often mindless action")28; and simplicity defined as removing steps until you reach "the simplest possible process," since "any technology that significantly reduces the steps to complete a task will enjoy high adoption."28 Alan Cooper's "postures" add nuance — Figma as a sovereign app should be tuned "for intermediate users, rather than first timers."29

The bias toward addition — and the case for subtraction

A recurring craft instinct is that good design usually means removing things, and that our instincts fight it. "People are biased towards solving problems through addition rather than subtraction… because adding something makes you feel like you are advancing, while taking something away makes you feel like you are retreating."30 The cautionary tale is Starbucks' app enabling "over 170,000 ways to customize" a drink; the model is the pedal-less Strider Bike.30 Jobs's own version of subtraction was Apple's four-quadrant product grid — two notebooks, two desktops — replacing a bloated pipeline.1

Doing beats knowing — and room for fun

Two smaller notes round out the ethos. When a six-week Basecamp onboarding overhaul lifted conversion ~30% with no clear cause, Fried's response was liberating: "The point wasn't to know, it was to do. And it was done."31 And craft leaves room for delight — Sticker Mule's mule that "hides while you type your password" is celebrated as a business making "room for fun."32 Even reading behavior is worth designing for: Nieman Lab's five modes — "Scan, Read, Read (long), Idle, and Shallow" — remind designers that attention arrives in different shapes.33

  1. Jony Ive.md ↩↩↩↩↩↩

  2. Fast Software, the Best Software.md ↩↩↩↩↩↩

  3. Tweets From Obvious.md ↩↩

  4. Chase the Product, Not the Data.md ↩↩↩

  5. Emotional Products Eat Rational Products.md ↩↩↩↩↩

  6. Sigma BF Less Camera, More Vision.md ↩↩

  7. Brilliant Hardware in the Valley of the Software Slump.md ↩↩

  8. How to Improve Your Microcopy UX Writing Tips for Non-Ux Writers.md ↩↩↩↩

  9. Tweets From Jason Fried.md ↩

  10. Tweets From Tyler Van Der Hoeven.md ↩

  11. Tweets From Raphael Schaad.md ↩↩

  12. Tweets From Alfred Lua.md ↩

  13. Tweets From Dixon & Moe.md ↩

  14. Tweets From Ludvig Rask.md ↩

  15. Tweets From Dan Liu ✦ lu.ma.md ↩

  16. Tweets From Akella.md ↩

  17. Tweets From Erik D. Kennedy.md ↩

  18. Tweets From Michael Andreuzza.md ↩

  19. Tweets From Cat Wu.md ↩

  20. Design Ain’t a Democracy.md ↩

  21. Just Enough Research.md ↩

  22. Build Better Products.md ↩

  23. Tweets From Carter Bryden.md ↩

  24. Tweets From DHH.md ↩

  25. Tweets From Lenny Rachitsky 🤗.md ↩

  26. Discussing Design.md ↩

  27. Articulating Design Decisions.md ↩

  28. Hooked.md ↩↩↩

  29. Tweets From Akshay Verma.md ↩

  30. Take Something Away.md ↩↩

  31. We Increased Conversion ~30% and We Don't Know Exactly How.md ↩

  32. Tweets From Sam McAllister.md ↩

  33. Tweets From Nieman Lab.md ↩