Full Width [alt+shift+f] Shortcuts [alt+shift+k]
Sign Up [alt+shift+s] Log In [alt+shift+l]
1

Logo: Squawkbox

from tonsky.me [alt+shift+b] in programming

Squawkbox is a boiler fleet monitor.
14th Jul 2026

Stay updated

Get a weekly newsletter with the top 5 articles worth reading every week.

More from tonsky.me

Every Frame Perfect

A while ago I was reading about Wayland and this quote stuck with me: A stated goal of Wayland is “every frame is perfect”. And I think this is a goal we should all aspire to. Wayland is talking about the technical side of things (modern GPU stacks are very complex and Wayland is trying to take control back) but it could be applied to UI too. The rule of thumb is: If I take a screenshot of your app at any moment, it must make sense Why care about every frame? It builds trust. Users can’t see the code, so UI is the only way for them to judge the quality of the app. If UI looks good, that means developers had time to polish it, which means that they probably spent a comparable amount of time to iron out the code. It’s a heuristic, but a reasonable one. Now, what does it mean in practice? I can think of a few things: No white flashes between screens. No partially loaded content. No relayout while content loads. Internally consistent. If one part of the UI says “1 update available”, another part should not say “Checking for updates...” Precise animations. Animations often end up being forgotten. A UI might look great in both start and end states but very janky in between. Like this: If you feel like there are weird things going on there, there are! Look at slowed down version: Now let’s apply our rule and take screenshots in the middle of the animation. This doesn’t look right: Neither does this: Both of these frames are not perfect. Let’s look at another example. Safari: Placeholder text here moves from the center but cursor animates from the left position: Not the end of the world by any means, but it does create a feeling that these two components are not in sync with each other. Next thought: maybe they weren’t designed together? If so, then they might not work well together. That’s how trust is lost. This desynchronization can lead to a lot of confusion. For example, in Photos, when switching between Crop and Adjust mode, picture snaps into place immediately but the crop border is animated: This creates a false feeling that something subtly changes when you switch between modes. And you know what? I don’t want my UI to give me false feelings. I want it to be a precise instrument, not an animated toy. Sometimes animations are supposed to help you understand a transition, so it’s doubly sad when they make it harder. Follow the magnifying glass: Same with Youtube. They had the simplest task in the world: move a rectangle from one position to another! Yet they decided to do something very strange: Can you explain this? Does it make sense? Probably a technical limitation of the DOM architecture they decided earlier on. I call these situations “The technology has outsmarted the programmer”. But no matter the reason, the result is an imperfect frame. Sometimes animations are left out as an afterthought. Whatever happens, happens. Then we get this: The details are fascinating to watch: So yeah. Please pay attention not only to the start and end states, but also to everything in between. Every frame matters. I’ll leave you with this unprovoked zoom animation from Preview app. Take care!

13th Jun 2026 1 votes
Claude is an Electron App because we’ve lost native

In “Why is Claude an Electron App?” Drew Breunig wonders: Claude spent $20k on an agent swarm implementing (kinda) a C-compiler in Rust, but desktop Claude is an Electron app. If code is free, why aren’t all apps native? And then argues that the answer is that LLMs are not good enough yet. They can do 90% of the work, so there’s still a substantial amount of manual polish, and thus, increased costs. But I think that’s not the real reason. The real reason is: native has nothing to offer. API-wise, native apps lost to web apps a long time ago. Native APIs are terrible to use, and OS vendors use everything in their power to make you not want to develop native apps for their platform. That explains the rise of Electron before LLM times, but it’s also a problem that LLMs solve now: if that was a real barrier to developing native apps, it doesn’t exist anymore. Then there’re looks and consistency. Some time ago, maybe in the late 90s and 2000s, native was ahead. It used to look good, it was consistent, and it all actually worked: the more apps used native look and feel, the better user experience was across apps (which we used to call programs). These days, though, native is as bad as the web, if not worse. Consistency is basically out the window. Anything can look like anything, buttons have no borders, contrast doesn’t exist, and neither do conventions. Apple, for example, seems to place traffic lights and corner radius by vibes rather than by any measurable guidelines. Maybe the server should round the corners? Looks could be good, but they also can be bad, and then you are stuck with platform-consistent, but generally bad UI (Liquid Glass ahem). It changes too often, too: the app you made today will look out of place next year, when Apple decides to change look and feel yet again. There’s no native look anymore. Computer UIs also degrade over time Theoretically, native apps can integrate with OS on a deeper level. This sounds nice, but what does that mean in practice? There are almost no good interoperable file formats; everything is locked inside individual apps, most services moved to the web, and OSes dropped the ball for making a good shared baseline. You can integrate with OS-provided calendar, but you can’t do it with web calendar. Well, you can, of course, but it’s easier on the web; native doesn’t help with it at all. Web pages only lead to more web pages Finally, the last hope of people longing for native is performance. They feel that native apps will be faster. Well, they can, but it doesn’t mean they will. Web apps can be faster, too, but in practice, nobody cares. There’s no technical reason why Slack needs to load 80 MiB just to show 10 channel names and 3 messages on a screen. The web is not the problem here! It’s a choice to be bad. What makes you think it’ll be different once the company decides to move to native? Don’t get me wrong: writing this brings me no joy. I don’t think web is a solution either. I just remember good times when native did a better-than-average job, and we were all better for using it, and it saddens me that these times have passed. I just don’t think that kidding ourselves that the only problem with software is Electron and it all will be butterflies and unicorns once we rewrite Slack in SwiftUI is not productive. The real problem is a lack of care. And the slop; you can build it with any stack.

3rd Mar 2026 1 votes
It’s hard to justify Tahoe icons

I was reading Macintosh Human Interface Guidelines from 1992 and found this nice illustration: accompanied by explanation: Fast forward to 2025. Apple releases macOS Tahoe. Main attraction? Adding unpleasant, distracting, illegible, messy, cluttered, confusing, frustrating icons (their words, not mine!) to every menu item: Sequoia → Tahoe It’s bad. But why exactly is it bad? Let’s delve into it! Disclaimer: screenshots are a mix from macOS 26.1 and 26.2, taken from stock Apple apps only that come pre-installed with the system. No system settings were modified. Icons should differentiate The main function of an icon is to help you find what you are looking for faster. Perhaps counter-intuitively, adding an icon to everything is exactly the wrong thing to do. To stand out, things need to be different. But if everything has an icon, nothing stands out. The same applies to color: black-and-white icons look clean, but they don’t help you find things faster! Microsoft used to know this: Look how much faster you can find Save or Share in the right variant: It also looks cleaner. Less cluttered. A colored version would be even better (clearer separation of text from icon, faster to find): I know you won’t like how it looks. I don’t like it either. These icons are hard to work with. You’ll have to actually design for color to look nice. But the principle stands: it is way easier to use. Consistency between apps If you want icons to work, they need to be consistent. I need to be able to learn what to look for. For example, I see a “Cut” command and next to it. Okay, I think. Next time I’m looking for “Cut,” I might save some time and start looking for instead. How is Tahoe doing on that front? I present to you: Fifty Shades of “New”: I even collected them all together, so the absurdity of the situation is more obvious. Granted, some of them are different operations, so they have different icons. I guess creating a smart folder is different from creating a journal entry. But this? Or this: Or this: There is no excuse. Same deal with open: Save: Yes. One of them is a checkmark. And they can’t even agree on the direction of an arrow! Close: Find (which is sometimes called Search, and sometimes Filter): Delete (from Cut-Copy-Paste-Delete fame): Minimize window. These are not some obscure, unique operations. These are OS basics, these are foundational. Every app has them, and they are always in the same place. They shouldn’t look different! Consistency inside the same app Icons are also used in toolbars. Conceptually, operations in a toolbar are identical to operations called through the menu, and thus should use the same icons. That’s the simplest case to implement: inside the same app, often on the same screen. How hard can it be to stay consistent? Preview: Photos: same and mismatch, but reversed ¯\_(ツ)_/¯ Maps and others often use different symbols for zoom: Icon reuse Another cardinal sin is to use the same icon for different actions. Imagine: I have learned that means “New”: Then I open an app and see. “Cool”, I think, “I already know what it means”: Gotcha! You’d think: okay, means quick look: Sometimes, sure. Some other times, means “Show completed”: Sometimes is “Import”: Sometimes is “Updates”: Same as with consistency, icon reuse doesn’t only happen between apps. Sometimes you see in a toolbar: Then go to the menu in the same app and see means something else: Sometimes identical icons meet in the same menu. Sometimes next to each other. Sometimes they put an entire barrage of identical icons in a row: This doesn’t help anyone. No user will find a menu item faster or will understand the function better if all icons are the same. The worst case of icon reuse so far has been the Photos app: It feels like the person tasked with choosing a unique icon for every menu item just ran out of ideas. Understandable. Too much nuance When looking at icons, we usually allow for slight differences in execution. That lets us, for example, understand that these technically different road signs mean the same thing: Same applies for icons: if you draw an arrow going out of the box in one place and also an arrow and the box but at a slightly different angle, or with different stroke width, or make one filled, we will understand them as meaning the same thing. Like, is supposed to mean something else from ? Come on! Or two-letter As that only slightly differ in the font size: A pencil is “Rename” but a slightly thicker pencil is “Highlight”? Arrows that use different diagonals? Three dots occupying ⅔ of space vs three dots occupying everything. Seriously? Slightly darker dots? The sheet of paper that changes meaning depending on if its corner is folded or if there are lines inside? But the final boss are arrows. They are all different: Supposedly, a user must become an expert at noticing how squished the circle is, if it starts top to right or bottom to right, and how far the arrow’s end goes. Do I care? Honestly, no. I could’ve given it a shot, maybe, if Apple applied these consistently. But Apple considers and to mean the same thing in one place, and expects me to notice minute details like this in another? Sorry, I can’t trust you. Not after everything I’ve seen. Detalization Icons are supposed to be easily recognizable from a distance. Every icon designer knows: small details are no-go. You can have them sometimes, maybe, for aesthetic purposes, but you can’t rely on them. And icons in Tahoe menus are tiny. Most of them fit in a 12×12 pixel square (actual resolution is 24×24 because of Retina), and because many of them are not square, one dimension is usually even less than 12. It’s not a lot of space to work with! Even Windows 95 had 16×16 icons. If we take the typical DPI of that era at 72 dots per inch, we get a physical icon size of 0.22 inches (5.6 mm). On a modern MacBook Pro with 254 DPI, Tahoe’s 24×24 icons are 0.09 inches (2.4 mm). Sure, 24 is bigger than 16, but in reality, these icons’ area is 4 times as small! Simulated physical size comparison between 16×16 at 72 DPI (left) and 24×24 at 254 DPI (right) So when I see this: I struggle. I can tell they are different. But I definitely struggle to tell what’s being drawn. Even zoomed in 20×, it’s still a mess: Or here. These are three different icons: Am I supposed to tell plus sign from sparkle here? Some of these lines are half the pixel thicker than the other lines, and that’s supposed to be the main point: Is this supposed to be an arrow? A paintbrush? Look, a tiny camera. It even got an even tinier viewfinder, which you can almost see if you zoom in 20×: Or here. There is a box, inside that box is a circle, and inside it is a tiny letter. i with a total height of 2 pixels: Don’t see it? I don’t. But it’s there... And this is a window! It even has traffic lights! How adorable: Remember: these are retina pixels, ¼ of a real pixel. Steve Jobs himself claimed they were invisible. It turns out there’s a magic number right around 300 pixels per inch, that when you hold something around to 10 to 12 inches away from your eyes, is the limit of the human retina to differentiate the pixels. And yet, Tahoe icons rely on you being able to see them. Pixel grid When you have so little space to work with, every pixel matters. You can make a good icon, but you have to choose your pixels very carefully. For Tahoe icons, Apple decided to use vector fonts instead of good old-fashioned bitmaps. It saves Apple resources—draw once, use everywhere. Any size, any display resolution, any font width. But there’re downsides: fonts are hard to position vertically, their size doesn’t map directly to pixels, stroke width doesn’t map 1-to-1 to pixel grid, etc. So, they work everywhere, but they also look blurry and mediocre everywhere: Tahoe icon (left) and its pixel-aligned version (right). They certainly start to work better once you give them more pixels. iPad OS 26 vs macOS 26 or make graphics simpler. But the combination of small details and tiny icon size is deadly. So, until Apple releases MacBooks with 380+ DPI, unfortunately, we still have to care about the pixel grid. Confusing metaphors Icons might serve another function: to help users understand the meaning of the command. For example, once you know the context (move window), these icons explain what’s going on faster than words: But for this to work, the user must understand what’s drawn on the icon. It must be a familiar object with a clear translation to computer action (like Trash can → Delete), a widely used symbol, or an easy-to-understand diagram. HIG: A rookie mistake would be to misrepresent the object. For example, this is how selection looks like: But its icon looks like this: Honestly, I’ve been writing this essay for a week, and I still have zero ideas why it looks like that. There’s an object that looks like this, but it’s a text block in Freeform/Preview: It’s called character.textbox in SF Symbols: Why did it become a metaphor for “Select all”? My best guess is it’s a mistake. Another place uses text selection from iOS as a metaphor. On a Mac! Some concepts have obvious or well-established metaphors. In that case, it’s a mistake not to use them. For example, bookmarks: . Apple, for some reason, went with a book: Sometimes you already have an interface element and can use it for an icon. However, try not to confuse your users. Dots in a rectangle look like password input, not permissions: Icon here says “Check” but the action is “Uncheck”. Terrible mistake: icon doesn’t help, it actively confuses the user. It’s also tempting to construct a two-level icon: an object and some sort of indicator. Like, a checkbox and a cross, meaning “Delete checkbox”: Or a user and a checkmark, like “Check the user”: Unfortunately, constructs like this rarely work. Users don’t build sentences from building blocks you provide; they have no desire to solve these puzzles. Finding metaphors is hard. Nouns are easier than verbs, and menu items are mostly verbs. How does open look? Like an arrow pointing to the top right? Why? I’m not saying there’s an obvious metaphor for “Open” Apple missed. There isn’t. But that’s the point: if you can’t find a good metaphor, using no icon is better than using a bad, confusing, or nonsensical icon. There’s a game I like to play to test the quality of the metaphor. Remove the labels and try to guess the meaning. Give it a try: It’s delusional to think that there’s a good icon for every action if you think hard enough. There isn’t. It’s a lost battle from the start. No amount of money or “management decisions” is going to change that. The problems are 100% self-inflicted. All this being said, I gotta give Apple credit where credit is due. When they are good at choosing metaphors, they are good: Symmetrical actions A special case of a confusing metaphor is using different metaphors for actions that are direct opposites of one another. Like Undo/Redo, Open/Close, Left/Right. It’s good when their icons use the same metaphor: Because it saves you time and cognitive resources. Learn one, get another one for free. Because of that, it’s a mistake not to use common metaphors for related actions: Or here: Another mistake is to create symmetry where there is none. “Back” and “See all”? Some menus in Tahoe make both mistakes. E.g. lack of symmetry between Show/Hide and false symmetry between completed/subtasks: Import not mirrored by Export but by Share: Text in icons HIG again: Authors of HIG are arguing against including text as a part of an icon. So something like this: or this: would not fly in 1992. I agree, but Tahoe has more serious problems: icons consisting only of text. Like this: It’s unclear where “metaphorical, abstract icon text that is not supposed to be read literally” ends and actual text starts. They use the same font, the same color, so how am I supposed to differentiate? Icons just get in a way: A...Complete? AaFont? What does it mean? I can maybe understand and . Dots are supposed to represent something. I can imagine thinking that led to . But ? No decorations. No effects. Just plain Abc. Really? Text transformations One might think that using icons to illustrate text transformations is a better idea. Like, you look at this: or this: or this: and just from the icon alone understand what will happen with the text. Icon illustrates the action. Also, BIU are well-established in word processing, so all upside? Not exactly. The problem is the same—text icon looks like text, not icon. Plus, these icons are excessive. What’s the point of taking the first letter and repeating it? The word “Bold” already starts with a letter “B”, it reads just as easily, so why double it? Look at it again: It’s also repeated once more as a shortcut... There is a better way to design this menu: And it was known to Apple for at least 33 years. System elements in icons Operating system, of course, uses some visual elements for its own purposes. Like window controls, resize handles, cursors, shortcuts, etc. It would be a mistake to use those in icons. Unfortunately, Apple fell into this trap, too. They reused arrows. Key shortcuts: HIG has an entire section on ellipsis specifically and how dangerous it is to use it anywhere else in the menu. And this exact problem is in Tahoe, too. Icons break scanning Without icons, you can just scan the menu from top to bottom, reading only the first letters. Because they all align: macOS Sequoia In Tahoe, though, some menu items have icons, some don’t, and they are aligned differently: Some items can have both checkmarks and icons, or have only one of them, or have neither, so we get situations like this: Ugh. Special mention This menu deserves its own category: Same icon for different actions. Missing the obvious metaphor. Somehow making the first one slightly smaller than the second and third. Congratulations! It got it all. Is HIG still relevant? I’ve been mentioning HIG a lot, and you might be wondering: is an interface manual from 1992 still relevant today? Haven’t computers changed so much that entirely new principles, designs, and idioms apply? Yes and no. Of course, advice on how to adapt your icons to black-and-white displays is obsolete. But the principles—as long as they are good principles—still apply, because they are based on how humans work, not how computers work. Humans don’t get a new release every year. Our memory doesn’t double. Our eyesight doesn’t become sharper. Attention works the same way it always has. Visual recognition, motor skills—all of this is exactly as it was in 1992. So yeah, until we get a direct chip-to-brain interface, HIG will stay relevant. Conclusion In my opinion, Apple took on an impossible task: to add an icon to every menu item. There are just not enough good metaphors to do something like that. But even if there were, the premise itself is questionable: if everything has an icon, it doesn’t mean users will find what they are looking for faster. And even if the premise was solid, I still wish I could say: they did the best they could, given the goal. But that’s not true either: they did a poor job consistently applying the metaphors and designing the icons themselves. I hope this article would be helpful in avoiding common mistakes in icon design, which Apple managed to collect all in one OS release. I love computers, I love interfaces, I love visual communication. It makes me sad seeing perfectly good knowledge already accessible 30 years ago being completely ignored or thrown away today. On the upside: it’s not that hard anymore to design better than Apple! Let’s drink to that. Happy New year! From SF Symbols: a smiley face calling somebody on the phone Notes During review of this post I was made familiar with Jim Nielsen’s article, which hits a lot of the same points as I do. I take that as a sign there’s some common truth behind our reasoning. Also note: Safari → File menu got worse since 26.0. Used to have only 4 icons, now it’s 18! Thanks Kevin, Ryan, and Nicki for reading drafts of this post.

5th Jan 2026 1 votes
Statistics made simple

I have a weird relationship with statistics: on one hand, I try not to look at it too often. Maybe once or twice a year. It’s because analytics is not actionable: what difference does it make if a thousand people saw my article or ten thousand? I mean, sure, you might try to guess people’s tastes and only write about what’s popular, but that will destroy your soul pretty quickly. On the other hand, I feel nervous when something is not accounted for, recorded, or saved for future reference. I might not need it now, but what if ten years later I change my mind? Seeing your readers also helps to know you are not writing into the void. So I really don’t need much, something very basic: the number of readers per day/per article, maybe, would be enough. Final piece of the puzzle: I self-host my web projects, and I use an old-fashioned web server instead of delegating that task to Nginx. Static sites are popular and for a good reason: they are fast, lightweight, and fulfil their function. I, on the other hand, might have an unfinished gestalt or two: I want to feel the full power of the computer when serving my web pages, to be able to do fun stuff that is beyond static pages. I need that freedom that comes with a full programming language at your disposal. I want to program my own web server (in Clojure, sorry everybody else). Existing options All this led me on a quest for a statistics solution that would uniquely fit my needs. Google Analytics was out: bloated, not privacy-friendly, terrible UX, Google is evil, etc. What is going on? Some other JS solution might’ve been possible, but still questionable: SaaS? Paid? Will they be around in 10 years? Self-host? Are their cookies GDPR-compliant? How to count RSS feeds? Nginx has access logs, so I tried server-side statistics that feed off those (namely, Goatcounter). Easy to set up, but then I needed to create domains for them, manage accounts, monitor the process, and it wasn’t even performant enough on my server/request volume! My solution So I ended up building my own. You are welcome to join, if your constraints are similar to mine. This is how it looks: It’s pretty basic, but does a few things that were important to me. Setup Extremely easy to set up. And I mean it as a feature. Just add our middleware to your Ring stack and get everything automatically: collecting and reporting. (def app (-> routes ... (ring.middleware.params/wrap-params) (ring.middleware.cookies/wrap-cookies) ... (clj-simple-stats.core/wrap-stats))) ;; <-- just add this It’s zero setup in the best sense: nothing to configure, nothing to monitor, minimal dependency. It starts to work immediately and doesn’t ask anything from you, ever. See, you already have your web server, why not reuse all the setup you did for it anyway? Request types We distinguish between request types. In my case, I am only interested in live people, so I count them separately from RSS feed requests, favicon requests, redirects, wrong URLs, and bots. Bots are particularly active these days. Gotta get that AI training data from somewhere. RSS feeds are live people in a sense, so extra work was done to count them properly. Same reader requesting feed.xml 100 times in a day will only count as one request. Hosted RSS readers often report user count in User-Agent, like this: Feedly/1.0 (+http://www.feedly.com/fetcher.html; 457 subscribers; like FeedFetcher-Google) Mozilla/5.0 (compatible; BazQux/2.4; +https://bazqux.com/fetcher; 6 subscribers) Feedbin feed-id:1373711 - 142 subscribers My personal respect and thank you to everybody on this list. I see you. Graphs Visualization is important, and so is choosing the correct graph type. This is wrong: Continuous line suggests interpolation. It reads like between 1 visit at 5am and 11 visits at 6am there were points with 2, 3, 5, 9 visits in between. Maybe 5.5 visits even! That is not the case. This is how a semantically correct version of that graph should look: Some attention was also paid to having reasonable labels on axes. You won’t see something like 117, 234, 10875. We always choose round numbers appropriate to the scale: 100, 200, 500, 1K etc. Goes without saying that all graphs have the same vertical scale and syncrhonized horizontal scroll. Insights We don’t offer much (as I don’t need much), but you can narrow reports down by page, query, referrer, user agent, and any date slice. Not implemented (yet) It would be nice to have some insights into “What was this spike caused by?” Some basic breakdown by country would be nice. I do have IP addresses (for what they are worth), but I need a way to package GeoIP into some reasonable size (under 1 Mb, preferably; some loss of resolution is okay). Finally, one thing I am really interested in is “Who wrote about me?” I do have referrers, only question is how to separate signal from noise. Performance. DuckDB is a sport: it compresses data and runs column queries, so storing extra columns per row doesn’t affect query performance. Still, each dashboard hit is a query across the entire database, which at this moment (~3 years of data) sits around 600 MiB. I definitely need to look into building some pre-calculated aggregates. One day. How to get Head to github.com/tonsky/clj-simple-stats and follow the instructions: Let me know what you think! Is it usable to you? What could be improved? P.S. You can try the live example at tonsky.me/stats. The data was imported from Nginx access logs, which I turned on and off on a few occasions, so it’s a bit spotty. Still, it should give you a general idea.

15th Dec 2025 1 votes

More in programming

Attention is all you have

The Tetris effect is one of psychology’s most easy to reproduce experiments. Simply spend a bit of time playing the eponymous game every day for a few weeks. After a little while, you’ll start recognizing familiar Tetromino shapes in clouds, buildings, and everyday objects. You might even see them appear before your eyes when you start falling asleep. Tom Tang Attention hijacking There’s one lesson the Tetris effect teaches us: whatever you focus on long enough will end up shaping your thoughts. This can be a good thing since it’s how we learn new skills and discover new ideas. Sadly, less and less of our attention is focused intentionally. Instead of picking what we want to see we let other people decide what is supposed to be good for us. Do you want to watch a video? YouTube knows you like cooking and art streams. But why not also recommend a few clips about the stock market bubble, global warming, and the war in Iran. Doomscrolling will make you stay longer and click on a few more ads. Do you want to listen to music? Just open a Spotify playlist and let the algorithm figure out what you like. Please ignore the AI slop they will insert in between real songs to avoid paying royalties to real artists. Do you want to know how your colleagues are doing? Too bad, LinkedIn will bury any relevant career news between the opinion of complete strangers. It is surely just a coincidence that those strangers happen to be shilling whatever Microsoft is invested in at the moment. Do you want the opinion of strangers on a product? Well those Redditors you wanted to ask are probably just a bunch of LLMs talking to a bunch of Russian trolls now. I hope you didn’t value their opinion too much. If, like me and most people, you spend the major part of your day focused on your device, there’s no doubt it’s affecting you. And when you let someone else dictate what appears on your screen, it’s the same as giving them the key to your brain. New York Said Back to an intentional internet The internet wasn’t always like that. Before recommendation algorithms where a thing, you had to decide what you would be doing on the computer. You didn’t really have one big app that you could open and order it to entertain you. Instead, you had a few dozen of bookmarks to websites, each with a specific idea in mind. A site for video game news, that one website with lots of tutorials, a blog about anime that didn’t update often enough, a wiki about a TV show from the 90s… Of course awful things existed on the web. We had Encyclopedia Dramatica and Rotten.com, but you actually had to put the effort to go there if you wanted. Nobody was going to put pictures of dead kids and far-right propaganda as a suggestion after a pancake recipe or a cat video. The good thing is that this intentional internet is still around. It has just been a bit buried below the corporate web, but it’s not very hard to find. After all you’re on this blog, so you probably already have a good idea about it. The main difference between this time and now is you. When you want to get back to reading blogs, RSS feeds, and finish that tutorial instead of doomscrolling shorts, you have to get used to a slower internet. One where content is not infinite and doesn’t get updated every click. But like every habit, the only thing you have to do is to keep at it. And if you pay enough attention to it, something will click in your brain.

16 hours ago 2 votes
Trying the Software factory pattern.

One of the interesting challenges of the AI ecosystem in 2026 is that new, effective patterns emerge faster than I can adopt them. I’ll find a handful, get back to work, and realize a month later that I’d missed four or five more. The adoption cycle for Imprint this year has been something like: January: get every engineer onto Claude Code every single day March: ok, let’s also get everyone else onto Claude Code or Claude Cowork every single day April: local development is bottlenecked on checkout and worktree model, instead create ~10 local workspaces which each have an independent checkout of every repository, and operate at the workspace level, not at the repository level, so it can generate cross-repository pull requests across frontend, backend, infrastructure and data monorepos June: oh boy, agent-driven development is heavily constrained by lack of a common task management system with higher visibility and less permission complexity than Jira, so let’s migrate the entire company over to Linear and hard stop on Jira July: yikes, now we have visibility into all these tickets, many of them are trivial but managing them through local development isn’t scaling, let’s roll out an orchestrated harness which internally we call “Agent Fleet”, along the lines of Stripe’s Minions The most recent question for me has been figuring out how to adopt the software factory pattern. (After some light research, the specific AI-context origin of this term is slightly messy to attribute, but I think it might be Justin McCarthy in February 2026’s Software Factories And The Agentic Moment.) The software factory pattern is looping on a broad goal, and then relying on the harness to drive progress towards that goal. Our first pass at implementation is fairly basic: An agent skill /linear-project-loop which reads in a Linear project and starts by auditing that project’s goal definition on these dimensions: An RFC in Notion that describes the project’s goals, how those goals are measured, and the general approach A Datadog dashboard or Snowflake queries that measure progress against those goals If those are missing, or the Linear project is missing in its entirety, it iterates with you on creating those missing tools. Then it reviews the state of the metrics and issues for the project. If new work is identified, it adds those issues to the project. It updates the state of issues that have moved. It works on the non-blocked tasks based on the project’s current state. This is often writing a pull request, updating a pull request, pinging for review, asking a clarifying question, etc. When a task completes, if the project description is fresh, it takes on the next task. If the description hasn’t been updated in a while, it reruns the loop starting with the first step. Right now I am running this locally in a local harness, but it’s working well enough that I anticipate moving the behavior to be driven by the same orchestrated harness that we assign one-off tasks to. What I particularly like about the factory pattern is that it parallels very closely how I’ve been working locally, while forcing me to recognize the places where I was accidentally hording parts of the state for myself regarding the goals of the project. I was already asking agents to iterate on specific Linear projects, but they didn’t have the ability to evaluate if they were going in the right direction, or if it was missing necessary tasks. Now it does. The other place this has been extremely helpful for me is checking in on projects post release. For example, I shipped our passkeys implementation earlier this year, but some months go by without my checking in on how it’s going. If we saw adoption spike, or error rates start to turn, I might miss it, but running the factory in a less frequent post-release mode would catch it immediately. The final thought that’s been interesting to me is how much all of the pieces here compound only to the extent that you have the other pieces. For example, this factory pattern depends on having Datadog MCP and Snowflake access available to manage goal-tracking, but it also depends on Linear being the single source of state for the company’s work, and an orchestrated harness that can perform work independently from your laptop. Keeping up with this many migrations is a fascinating industry moment.

yesterday 2 votes
CSS-Tricks could be a co-op

I owe a lot of my professional identity and success to CSS-Tricks. CSS-Tricks repeatedly gave me the opportunity to write for them. In doing so, they helped to both socialize and normalize accessibility as a mainstream frontend concern. I’m deeply thankful to them for this. The team was also a joy to work with, notably Geoff Graham. He’s a mensch, and one of the nicest people you can interact with in the frontend web space. If you have not been following the news about the site, Kevin Powell has a good video about the whole situation: Content skipped. I’m not speaking on behalf of Geoff, Chris, or others involved with running the current version of CSS-Tricks. I’ve got skin in the game as an author. This is my personal opinion, born of my feelings and beliefs. I think a lot of the web’s infrastructure should be co-ops, and CSS-Tricks is knowledge infrastructure. To that point, I should also point out that the website covers far more than just CSS. The corporate model of ownership can be a risk. If infrastructure is not part of a corporation’s core strategy, it is not a priority. As Kevin’s video touched on, it seems like promotion via owning the frontend content space isn’t part of Digital Ocean’s strategy anymore. It is not that CSS-Tricks does not have value. It is that Digital Ocean cannot see it. It is deeply, tragically ironic to me that Digital Ocean allowed this to transpire. This is because I know for a fact that the techniques and philosophies shared by CSS-Trick authors helped to shape iterations of their product’s UI. Some may be quick to point out that this knowledge now—illegally—exists inside of LLM training data, so the risk of the website going away is mitigated. To this, know that we should be striving to keep resources like CSS-Tricks going. Human creativity is the force that creates new techniques, strategies, and technologies. The web will calcify without voices sharing what they know, forever locking us into endless permutations of a fixed point in time. Unlike corporations, co-ops don’t have to be motivated by profit. By not needing to prioritize growth at all costs it means co-ops can instead prioritize and incentivise things like preservation and cultivation. It is also a successful model of operation, one that even already exists, and flourishes in the tech space. Collective ownership can also serve as checks and balances for, and protection against hierarchical decision-making. I only need to point to the chaotic and aberrant decisions many CEOs in the technology space have been making as of late to demonstrate the value of this approach. Paddy Srinivasan, if you somehow wind up reading this: Save some face and take a big swing. Give CSS-Tricks back to the people who love it.

3 days ago
fibre broadband anticlimax

How can something that “just works” be so annoying? situation We live in Cambridge off a little road down a drive in shared ownership between us and our neighbouring houses. All the utilities are buried under this drive, including the phone line. anticipation Over the last few years we have been canvassed repeatedly by CityFibre saying that they can deliver fibre all way to our house. I saw them digging trenches and leaving tails of purple fibre cladding along nearby roads, ready to hook up all the houses. I thought they would need to do something similar to deliver fibre to us. So when they turned up and knocked on our door, I talked to their salesbods and walked them up and down the drive and pointed out where the existing BT line goes. Then they gave up trying to sell to us. This happened about three times. disaffection We were not eager enough for an upgrade to deal with these impediments. notification A few months ago we were told that CityFibre would soon come and do the upgrade, since there’s a nationwide deadline for turning off the copper phone network at the end of the year. We expected that this would force them to actually plan some digging works, so we talked to our neighbours about it. We were all ready for some huge faff to follow the next visit by the CityFibre bods. installation CityFibre turned up on the promised morning bright and early. To our enormous surprise, a brown fibre housing was already poking out of the ground next to our copper phone line. It had been fed through 50 metres of 5cm duct without us being aware they were even working on the street. Within a couple of hours, the technicians had drilled through our wall, installed the ONT, blown fibre through the unexpected pipe, plugged in the CPE (superficially identical to the old one), and left telling us to anticipate that it might not work properly until tomorrow. activation Around lunch time, the copper phone line stopped working completely. Some faff ensued, switching all our devices over to the new WiFi network. For a while we thought this was the death of our land line, but in the course of debugging other issues, I realised that the router has a built-in VoIP adapter (I don’t think we were told it has a built-in VoIP adapter) so I plugged the phone in and it Just Worked: they had ported our phone number across and everything. Flawless. I was seriously impressed. rumination It has been a few weeks since the switchover, and apart from a couple of horrible Clown-afflicted IoT devices, it has been fairly smooth. What prompted me to write this up was realising that we delayed this upgrade for years because the sales people were not given enough technical information about how the installation process works: the fact that houses typically have a 5cm duct containing the copper lines (probably standard for the last 40 years) and the fact that fibre can be shoved through a few tens of metres without difficulty. And worse, the sales people didn’t have an esclation path for difficult cases: they just gave up instead. From a technical point of view, the installation was impeccable. (I guess the loose 24 hour window for the cutover time was because OpenReach and CityFibre don’t have tight requirements on ISP reconfiguration schedules.) From the sales point of view, it was crap. Maybe it would have gone faster if we offered to switch early without asking if the drive would be a problem? But I guess the difference between “yes!” and “yes, but will this be a problem?” is too much to expect from a minimum-wage door-to-door salesbod whose employer didn’t give them enough information or any escalation path.

4 days ago
A Simple Guide for Calm UI

Read the post here.

4 days ago
📚 BoredReading

You seem to be enjoying this.

Join free to unlock everything.

Create free account

Already have an account? Sign in