Journal tags: min

People talk about the effectiveness (or lack thereof) of large language models as though all tasks are comparable. But it strikes me that there are three broad categories of work that large language models are applied to:

1. Compression.
2. Transformation.
3. Expansion.

Compression is when you feed a large language model something big that you want to make small. Summarise this book. Give me the gist of this meeting. Large language models are generally pretty good at this, which makes sense given that they themselves are kind of like compressed artifacts.

Transformation is when large language models convert from one format into another. Turn this audio into text. Turn this jumble of data into structured JSON. A large language model can handle these tasks pretty well. There’ll probably be a few errors so make sure that’s not a deal-breaker.

Expansion is when you give a large language model a prompt to generate something from scratch. An image. A presentation. An email. A poem. This is where slop lives. The output inevitably betrays its origins, glistening with a sheen of mediocrity.

Laurie spotted this three-way split a while back:

Is what you’re doing taking a large amount of text and asking the LLM to convert it into a smaller amount of text? Then it’s probably going to be great at it. If you’re asking it to convert into a roughly equal amount of text it will be so-so. If you’re asking it to create more text than you gave it, forget about it.

I hope that when the bubble finally bursts, we’ll see the surviving large language models put to work on the first two categories. The boring stuff. The work that’s tedious for humans.

But tedious is as tedious does. Something I consider drudgery might be the very thing that gives you life. Like Giles says:

I have a feeling that everyone likes using AI tools to try doing someone else’s profession. They’re much less keen when someone else uses it for their profession.

The big exception seems to be programming. Apparently there are plenty of coders who never before expressed an interest in being managers who are now happily hanging up their coding spurs in favour being the overseer of non-human workers.

It’s a reasonable outlook. It could even be considered a user-centred approach. Users don’t care about the elegance of your code; they care about accomplishing their tasks.

Programming is something of an exception to the efficacy of large language models in general. Instead of relying on the subjectivity of painting, poetry, or prose, programming can be objectively tested. Throw enough money at the worst people in the world and they’ll give you tokens you can use to get the machines to test their own output. So you can get a large language model to create something reasonably good from scratch as long as that something is code.

If you had asked me about the threat model of large language models two years ago, I probably would’ve been worried for artists, writers, and musicians. I thought that software had enough inherent complexity to be relatively safe.

Now my opinion has completely reversed. Software is almost certainly the killer app for large language models.

I think the artists, writers, and musicians will be okay, or at least as okay as they ever were. It turns out that humans like things made by other humans.

And y’know what? If I had to choose which endeavour I’d rather see automated away—programming or art—it’s no competition.

Don’t get me wrong—it would be nice if everyone got paid for doing what they enjoy. It’s just that I’m okay with software engineers not being at the front of that line.

I remember when I first started getting paid money to make websites. “Really?” I thought, “Someone is willing to pay me to do something I’d do anyway?” I kept waiting for the jig to be up. Instead I saw my profession grow and expand.

Perhaps there’s a long-overdue compression happening.

Or maybe it’s more like a transformation.

If you wanted to make a really crude approximation of project management, you could say there are two main styles: waterfall and agile.

It’s not as simple as that by any means. And the two aren’t really separate things; agile came about as a response to the failures of waterfall. But if we’re going to stick with crude approximations, here we go:

• In a waterfall process, you define everything up front and then execute.
• In an agile process, you start executing and then adjust based on what you learn.

So crude! Much approximation!

It only recently struck me that the agile approach is basically a cybernetic system.

Cybernetics is pretty much anything that involves feedback. If it’s got inputs and outputs that are connected in some way, it’s probably cybernetic. Politics. Finance. Your YouTube recommendations. Every video game you’ve ever played. You. Every living thing on the planet. That’s cybernetics.

Fun fact: early on in the history of cybernetics, a bunch of folks wanted to get together at an event to geek about this stuff. But they knew that if they used the word “cybernetics” to describe the event, Norbert Wiener would show up and completely dominate proceedings. So they invented a new alias for the same thing. They coined the term “artificial intelligence”, or AI for short.

Yes, ironically the term “AI” was invented in order to repel a Reply Guy. Now it’s Reply Guy catnip. In today’s AI world, everyone’s a Norbert Wiener.

The thing that has the Wieners really excited right now in the world of programming is the idea of agentic AI. In this set-up, you don’t do any of the actual coding. Instead you specify everything up front and then have a team of artificial agents execute your plan.

That’s right; it’s a return to waterfall. But that’s not as crazy as it sounds. Waterfall was wasteful because execution was expensive and time-consuming. Now that execution is relatively cheap (you pay a bit of money to line the pockets of the worst people in exchange for literal tokens), you can afford to throw some spaghetti at the wall and see if it sticks.

But you lose the learning. The idea of a cybernetic system like, say, agile development, is that you try something, learn from it, and adjust accordingly. You remember what worked. You remember what didn’t. That’s learning.

Outsourcing execution to machines makes a lot of sense.

I’m not so sure it makes sense to outsource learning.

Not long now ’till Web Day Out — just three weeks!

It’s also not that long until the start of a new financial year so if you’ve got training budget that needs to be used this year, send your team to Web Day Out. Not only is it excellent value for money, it’s also going to have an incredibly high density of knowledge bombs per talk.

CSS! Progressive web apps! Web typography! Browser support! And much more.

If you like the sound of Web Day Out, you’ll also like State Of The Browser, which is just ten days away. In-person tickets for that event are now sold out, but online streaming tickets are still available.

Better yet, if you buy a ticket to Web Day Out, you automatically get a free online streaming ticket for State Of The Browser!

So get your ticket in the next ten days, enjoy State Of The Browser from the comfort of your own home, and then enjoy a trip to Brighton for Web Day Out on Thursday, 12 March. See you there!

When I talk about large language models, I make sure to call them large language models, not “AI”. I know it’s a lost battle, but the terminology matters to me.

The term “AI” can encompass everything from a series of if/else statements right up to Skynet and HAL 9000. I’ve written about this naming collision before.

It’s not just that the term “AI” isn’t useful, it’s so broad as to be actively duplicitous. While talking about one thing—like, say, large language models—you can point to a completely different thing—like, say, machine learning or computer vision—and claim that they’re basically the same because they’re both labelled “AI”.

If a news outlet runs a story about machine learning in the context of disease prevention or archeology, the headline will inevitably contain the phrase “AI”. That story will then gleefully be used by slopagandists looking to inflate the usefulness of large language models.

Conflating these different technologies is the fallacy at the heart of Robin Sloan’s faulty logic:

If these machines churn through all media, and then, in their deployment, discover several superconductors and cure all cancers, I’d say, okay … we’re good.

John Scalzi recently wrote:

“AI” is mostly a marketing phrase for a bunch of different processes and tools which in a different era would have been called “machine learning” or “neural networks” or something else now horribly unsexy.

But I’ve noticed something recently. More than once I’ve seen genuinely-useful services refer to their technology as “traditional machine learning”.

First off, I find that endearing. Like machine learning is akin to organic farming or hand-crafted furniture.

Secondly, perhaps it points to a severing of the ways between machine learning and large language models.

Up until now it may have been mutually benificial for them to share the same marketing term, but with the bubble about to burst, anything to do with large language models might become toxic by association, including the term “AI”. Hence the desire to shake the large-language model grifters from the coattails of machine learning and computer vision.

The Wikimedia Foundation, stewards of the finest projects on the web, have written about the hammering their servers are taking from the scraping bots that feed large language models.

Our infrastructure is built to sustain sudden traffic spikes from humans during high-interest events, but the amount of traffic generated by scraper bots is unprecedented and presents growing risks and costs.

Drew DeVault puts it more bluntly, saying Please stop externalizing your costs directly into my face:

Over the past few months, instead of working on our priorities at SourceHut, I have spent anywhere from 20-100% of my time in any given week mitigating hyper-aggressive LLM crawlers at scale.

And no, a robots.txt file doesn’t help.

If you think these crawlers respect robots.txt then you are several assumptions of good faith removed from reality. These bots crawl everything they can find, robots.txt be damned.

Free and open source projects are particularly vulnerable. FOSS infrastructure is under attack by AI companies:

LLM scrapers are taking down FOSS projects’ infrastructure, and it’s getting worse.

You try to do the right thing by making knowledge and tools freely available. This is how you get repaid. AI bots are destroying Open Access:

There’s a war going on on the Internet. AI companies with billions to burn are hard at work destroying the websites of libraries, archives, non-profit organizations, and scholarly publishers, anyone who is working to make quality information universally available on the internet.

My own experience with The Session bears this out.

Ars Technica has a piece on this: Open source devs say AI crawlers dominate traffic, forcing blocks on entire countries .

So does MIT Technology Review: AI crawler wars threaten to make the web more closed for everyone.

When we talk about the unfair practices and harm done by training large language models, we usually talk about it in the past tense: how they were trained on other people’s creative work without permission. But this is an ongoing problem that’s just getting worse.

The worst of the internet is continuously attacking the best of the internet. This is a distributed denial of service attack on the good parts of the World Wide Web.

If you’re using the products powered by these attacks, you’re part of the problem. Don’t pretend it’s cute to ask ChatGPT for something. Don’t pretend it’s somehow being technologically open-minded to continuously search for nails to hit with the latest “AI” hammers.

If you’re going to use generative tools powered by large language models, don’t pretend you don’t know how your sausage is made.

There are different kinds of buzzwords.

Some buzzwords are useful. They take a concept that would otherwise require a sentence of explanation and package it up into a single word or phrase. Back in the day, “ajax” was a pretty good buzzword.

Some buzzwords are worse than useless. This is when a word or phrase lacks definition. You could say this buzzword in a meeting with five people, and they’d all understand five different meanings. Back in the day, “web 2.0” was a classic example of a bad buzzword—for some people it meant a business model; for others it meant rounded corners and gradients.

The worst kind of buzzwords are the ones that actively set out to obfuscate any actual meaning. “The cloud” is a classic example. It sounds cooler than saying “a server in Virginia”, but it also sounds like the exact opposite of what it actually is. Great for marketing. Terrible for understanding.

“AI” is definitely not a good buzzword. But I can’t quite decide if it’s merely a bad buzzword like “web 2.0” or a truly terrible buzzword like “the cloud”.

The biggest problem with the phrase “AI” is that there’s a name collision.

For years, the term “AI” has been used in science-fiction. HAL 9000. Skynet. Examples of artificial general intelligence.

Now the term “AI” is also used to describe large language models. But there is no connection between this use of the term “AI” and the science fictional usage.

This leads to the ludicrous situation of otherwise-rational people wanted to discuss the dangers of “AI”, but instead of talking about the rampant exploitation and energy usage endemic to current large language models, they want to spend the time talking about the sci-fi scenarios of runaway “AI”.

To understand how ridiculous this is, I’d like you to imagine if we had started using a different buzzword in another setting…

Suppose that when ride-sharing companies like Uber and Lyft were starting out, they had decided to label their services as Time Travel. From a marketing point of view, it even makes sense—they get you from point A to point B lickety-split.

Now imagine if otherwise-sensible people began to sound the alarm about the potential harms of Time Travel. Given the explosive growth we’ve seen in this sector, sooner or later they’ll be able to get you to point B before you’ve even left point A. There could be terrible consequences from that—we’ve all seen the sci-fi scenarios where this happens.

Meanwhile the actual present-day harms of ride-sharing services around worker exploitation would be relegated to the sidelines. Clearly that isn’t as important as the existential threat posed by Time Travel.

It sounds ludicrous, right? It defies common sense. Just because a vehicle can get you somewhere fast today doesn’t mean it’s inevitably going to be able to break the laws of physics any day now, simply because it’s called Time Travel.

And yet that is exactly the nonsense we’re being fed about large language models. We call them “AI”, we look at how much they can do today, and we draw a straight line to what we know of “AI” in our science fiction.

This ridiculous situation could’ve been avoided if we had settled on a more accurate buzzword like “applied statistics” instead of “AI”.

It’s almost as if the labelling of the current technologies was more about marketing than accuracy.

I’ve been deep-diving into HTML web components over the past few weeks. I decided to refactor the JavaScript on The Session to use custom elements wherever it made sense.

I really enjoyed doing this, even though the end result for users is exactly the same as before. This was one of those refactors that was for me, and also for future me. The front-end codebase looks a lot more understandable and therefore maintainable.

Most of the JavaScript on The Session is good ol’ DOM scripting. Listen for events; when an event happens, make some update to some element. It’s the kind of stuff we might have used jQuery for in the past.

Chris invoked Betteridge’s law of headlines recently by asking Will Web Components replace React and Vue? I agree with his assessment. The reactivity you get with full-on frameworks isn’t something that web components offer. But I do think web components can replace jQuery and other approaches to scripting the DOM.

I’ve written about my preferred way to do DOM scripting: element.target.closest. One of the advantages to that approach is that even if the DOM gets updated—perhaps via Ajax—the event listening will still work.

Well, this is exactly the kind of thing that custom elements take care of for you. The connectedCallback method gets fired whenever an instance of the custom element is added to the document, regardless of whether that’s in the initial page load or later in an Ajax update.

So my client-side scripting style has updated over time:

1. Adding event handlers directly to elements.
2. Adding event handlers to the document and using event.target.closest.
3. Wrapping elements in a web component that handles the event listening.

None of these progressions were particularly ground-breaking or allowed me to do anything I couldn’t do previously. But each progression improved the resilience and maintainability of my code.

Like Chris, I’m using web components to progressively enhance what’s already in the markup. In fact, looking at the code that Chris is sharing, I think we may be writing some very similar web components!

A few patterns have emerged for me…

Naming custom elements

Naming things is famously hard. Every time you make a new custom element you have to give it a name that includes a hyphen. I settled on the convention of using the first part of the name to echo the element being enhanced.

If I’m adding an enhancement to a button element, I’ll wrap it in a custom element that starts with button-. I’ve now got custom elements like button-geolocate, button-confirm, button-clipboard and so on.

Likewise if the custom element is enhancing a link, it will begin with a-. If it’s enhancing a form, it will begin with form-.

The name of the custom element tells me how it’s expected to be used. If I find myself wrapping a div with button-geolocate I shouldn’t be surprised when it doesn’t work.

Naming attributes

You can use any attributes you want on a web component. You made up the name of the custom element and you can make up the names of the attributes too.

I’m a little nervous about this. What if HTML ends up with a new global attribute in the future that clashes with something I’ve invented? It’s unlikely but it still makes me wary.

So I use data- attributes. I’ve already got a hyphen in the name of my custom element, so it makes sense to have hyphens in my attributes too. And by using data- attributes, the browser gives me automatic reflection of the value in the dataset property.

Instead of getting a value with this.getAttribute('maximum') I get to use this.dataset.maximum. Nice and neat.

The single responsibility principle

My favourite web components aren’t all-singing, all-dancing powerhouses. Rather they do one thing, often a very simple thing.

Here are some examples:

• Jason’s aria-collapsable for toggling the display of one element when you click on another.
• David’s play-button for adding a play button to an audio or video element.
• Chris’s ajax-form for sending a form via Ajax instead of a full page refresh.
• Jim’s user-avatar for adding a tooltip to an image.
• Zach’s table-saw for making tables responsive.

All of those are HTML web components in that they extend your existing markup rather than JavaScript web components that are used to replace HTML. All of those are also unambitious by design. They each do one thing and one thing only.

But what if my web component needs to do two things?

I make two web components.

The beauty of custom elements is that they can be used just like regular HTML elements. And the beauty of HTML is that it’s composable.

What if you’ve got some text that you want to be a level-three heading and also a link? You don’t bemoan the lack of an element that does both things. You wrap an a element in an h3 element.

The same goes for custom elements. If I find myself adding multiple behaviours to a single custom element, I stop and ask myself if this should be multiple custom elements instead.

Take some of those button- elements I mentioned earlier. One of them copies text to the clipboard, button-clipboard. Another throws up a confirmation dialog to complete an action, button-confirm. Suppose I want users to confirm when they’re copying something to their clipboard (not a realistic example, I admit). I don’t have to create a new hybrid web component. Instead I wrap the button in the two existing custom elements.

Rather than having a few powerful web components, I like having lots of simple web components. The power comes with how they’re combined. Like Unix pipes. And it has the added benefit of stopping my code getting too complex and hard to understand.

Communicating across components

Okay, so I’ve broken all of my behavioural enhancements down into single-responsibility web components. But what if one web component needs to have awareness of something that happens in another web component?

Here’s an example from The Session: the results page when you search for sessions in London.

There’s a map. That’s one web component. There’s a list of locations. That’s another web component. There are links for traversing backwards and forwards through the locations via Ajax. Those links are in web components too.

I want the map to update when the list of locations changes. Where should that logic live? How do I get the list of locations to communicate with the map?

Events!

When a list of locations is added to the document, it emits a custom event that bubbles all the way up. In fact, that’s all this component does.

You can call the event anything you want. It could be a newLocations event. That event is dispatched in the connectedCallback of the component.

Meanwhile in the map component, an event listener listens for any newLocations events on the document. When that event handler is triggered, the map updates.

The web component that lists locations has no idea that there’s a map on the same page. It doesn’t need to. It just needs to dispatch its event, no questions asked.

There’s nothing specific to web components here. Event-driven programming is a tried and tested approach. It’s just a little easier to do thanks to the connectedCallback method.

I’m documenting all this here as a snapshot of my current thinking on HTML web components when it comes to:

• naming custom elements,
• naming attributes,
• the single responsibility principle, and
• communicating across components.

I may well end up changing my approach again in the future. For now though, these ideas are serving me well.

This is a little follow-up to my post about web components for date inputs.

If you try the demo on iOS it doesn’t work. There’s nothing stopping you selecting any date.

That’s nothing to do with the web components. It turns out that Safari on iOS doesn’t support min and max on date inputs. This is also true of any other browser on iOS because they’re all just Safari in a trenchcoat …for now.

I was surprised — input type="date" has been around for a long time now. I mean, it’s not the end of the world. You’d have to do validation on inputted dates on the server anyway, but it sure would be nice for the user experience of filling in forms.

Alas, it doesn’t look like this is something on the interop radar.

What really surprised me was looking at Can I Use. That shows Safari on iOS as fully supporting date inputs.

Maybe it’s just semantic nitpickery on my part but I would consider that the lack of support for the min and max attributes means that date inputs are partially supported.

Can I Use gets its data from here. I guess I need to study the governance rules and try to figure out how to submit a pull request to update the currently incorrect information.

I just described prototype code as code to be thrown away. On that topic…

I’ve been observing how people are programming with large language models and I’ve seen a few trends.

The first thing that just about everyone agrees on is that the code produced by a generative tool is not fit for public consumption. At least not straight away. It definitely needs to be checked and tested. If you enjoy debugging and doing code reviews, this might be right up your street.

The other option is to not use these tools for production code at all. Instead use them for throwaway code. That could be prototyping. But it could also be the code for those annoying admin tasks that you don’t do very often.

Take content migration. Say you need to grab a data dump, do some operations on the data to transform it in some way, and then pipe the results into a new content management system.

That’s almost certainly something you’d want to automate with bespoke code. Once the content migration is done, the code can be thrown away.

Read Matt’s account of coding up his Braggoscope. The code needed to spider a thousand web pages, extract data from those pages, find similarities, and output the newly-structured data in a different format.

I’ve noticed that these are just the kind of tasks that large language models are pretty good at. In effect you’re training the tool on your own very specific data and getting it to do your drudge work for you.

To me, it feels right that the usefulness happens on your own machine. You don’t put the machine-generated code in front of other humans.

I’ve already written about how much I enjoyed hosting Leading Design San Francisco last week.

All the speakers were terrific. Lola’s talk was particularly …um, interesting:

In this talk, Lola will share her adventures in the world of blockchain, the hostility she experienced in her first go-round in 2018, and why she’s chosen to head back to a technology that is going through its largest reputational and social crisis to date.

Wait …I was supposed to stand on stage and introduce a talk that was (at least partly) about blockchain? I have opinions.

As it turned out, Lola warned me that I’d be making an appearance in her talk. She was going to quote that blog post. Before the talk, I asked her how obnoxious I could be about blockchain in her intro. She told me to bring it.

So in the introduction, I deployed all the sarcasm I had in me and said:

Listen, we designers have a tendency to be over-critical of things sometimes. There are all these ideas that we dismiss: phrenology, homeopathy, flat-earthism …blockchain. Haters gonna hate.

I remember somebody asking online a while back, “Why the hate for web3?” And someone I know responded by saying “We hate it because we understand it.” I think there’s a lot of truth to that.

But look, just because blockchains are powering crypto ponzi schemes and N F fucking Ts, it’s worth remembering that it’s also simply a technology. It’s a technological solution in search of a problem.

To be fair, it’s still early days. After all, it’s only been over a decade now.

It’s like the law of instrument says; when all you have is a hammer, everything looks like a nail. Blockchain is like that. Except the hammer is also made of glass.

Anyway, Lola is going to defend the indefensible and talk about blockchain. One thing to keep in mind is this: remember when everyone was talking about “The Cloud”? And then it turned out that you could substitute the phrase “someone else’s server” for “The Cloud?” Well, every time you hear Lola say the word “blockchain”, I’d like you to mentally substitute the phrase “multiple copies of a spreadsheet.”

Please give an open mind and a warm welcome to Lola Oyelayo Pearson!

I got some laughs. I also got lots of gasps and pearl-clutching, as though I were saying something taboo. Welcome to San Francisco.

Lola gave as good as she got. I got a roasting in her talk.

And just to clarify, Lola and I are friends—this was a consensual smackdown.

There was a very serious point to Lola’s talk. Cryptobollocks and other blockchain-powered schemes have historically been very bro-y, and exploitative of non-bro communities. Lola wants to fight that trend.

I get it. But it reminds me a bit of the justifications you hear from people who go to work at Facebook claiming that they can do more good from the inside. Whatever helps you sleep at night.

The crux of Lola’s belief is this: blockchain technology is inevitable, therefore it is uncumbent on us as ethical designers to ensure that the technology is deployed in a way that empowers people instead of exploiting them.

But I take issue with the premise. Blockchain technology is not inevitable. That’s the worst kind of technological determinism. It’s defeatist. It’s a depressing view of “progress” driven not by people, but by technological forces beyond our control.

I refuse to accept that anti-humanist deterministic view.

In any case, for technological determinism to have any validity, there needs to be something to it. At least virtual reality and machine learning are based on some actual technologies. In the case of cryptobollocks, there is no there there. There is nothing except the hype, which is why you’ll see blockchain enthusiasts trying to ride the coattails of trending technologies in a logical fallacy that goes something like this:

1. There are technologies that will be really big in the future,
2. blockchain is a technology, therefore
3. blockchain will be really big in the future.

Blockchain is bullshit. It isn’t even very clever bullshit. And it certainly isn’t inevitable.

I’ve mentioned before that I’m not a fan of initialisms and acronyms. They can be exclusionary.

It bothers me doubly when everyone is talking about AI.

First of all, the term is so vague as to be meaningless. Sometimes—though rarely—AI refers to general artificial intelligence. Sometimes AI refers to machine learning. Sometimes AI refers to large language models. Sometimes AI refers to a series of if/else statements. That’s quite a spectrum of meaning.

Secondly, there’s the assumption that everyone understands the abbreviation. I guess that’s generally a safe assumption, but sometimes AI could refer to something other than artificial intelligence.

In countries with plenty of pastoral agriculture, if someone works in AI, it usually means they’re going from farm to farm either extracting or injecting animal semen. AI stands for artificial insemination.

I think that abbreviation might work better for the kind of things currently described as using AI.

We were discussing this hot topic at work recently. Is AI coming for our jobs? The consensus was maybe, but only the parts of our jobs that we’re more than happy to have automated. Like summarising some some findings. Or perhaps as a kind of lorem ipsum generator. Or for just getting the ball rolling with a design direction. As Terence puts it:

Midjourney is great for a first draft. If, like me, you struggle to give shape to your ideas then it is nothing short of magic. It gets you through the first 90% of the hard work. It’s then up to you to refine things.

That’s pretty much the conclusion we came to in our discussion at Clearleft. There’s no way that we’d use this technology to generate outputs for clients, but we certainly might use it to generate inputs. It’s like how we’d do a quick round of sketching to get a bunch of different ideas out into the open. Terence is spot on when he says:

Midjourney lets me quickly be wrong in an interesting direction.

To put it another way, using a large language model could be a way of artificially injecting some seeds of ideas. Artificial insemination.

So now when I hear people talk about using AI to create images or articles, I don’t get frustrated. Instead I think, “Using artificial insemination to create images or articles? Yes, that sounds about right.”

I’ve always liked the idea that your website can be your API. Like, you’ve already got URLs to identify resources, so why not make that URL structure predictable and those resources parsable?

That’s why the (read-only) API for The Session doesn’t live at a separate subdomain. It uses the same URL structure as the regular site, but you can request the resources in an alternative format: JSON, XML, RSS.

This works out pretty well, mostly because I put a lot of thought into the URL structure of the site. I’m something of a URL fetishist, but I think that taking a URL-first approach to information architecture can be a good exercise.

Most of the resources on The Session involve nouns like tunes, events, discussions, and so on. There’s a consistent and predictable structure to the URLs for those sections:

• /things
• /things/new
• /things/search

And then an idividual item can be found at:

• things/ID

That’s all nice and predictable and the naming of the URLs matches what you’d expect to find:

• tunes
• tunes/new
• tunes/search?q=Ireland
• tunes/320

Tunes, events, discussions, sessions. Those are all fine. But there’s one section of the site that has this root URL:

/recordings

When I was coming up with the URL structure twenty years ago, it was clear what you’d find there: track listings for albums of music. No one would’ve expected to find actual recordings of music available to listen to on-demand. The bandwidth constraints and technical limitations of the time made that clear.

Two decades on, the situation has changed. Now someone new to the site might well expect to hit a link called “recordings” and expect to hear actual recordings of music.

So I should probably change the label on the link. I don’t think “albums” is quite right—what even is an album any more? The word “discography” is probably the most appropriate label.

Here’s my dilemma: if I update the label, should I also update the URL structure?

Right now, the section of the site with /tunes URLs is labelled “tunes”. The section of the site with /events URLs is labelled “events”. Currently the section of the site with /recordings URLs is labelled “recordings”, but may soon be labelled “discography”.

If you click on “tunes”, you end up at /tunes. But if you click on “discography”, you end up at /recordings.

Is that okay? Am I the only one that would be bothered by that?

I could update the URLs to match the labelling (with redirects for the old URLs, of course), but I’m not so keen on this URL structure:

• /discography
• /discography/new
• /discography/search
• /discography/ID

It doesn’t seem as tidy as:

• /recordings
• /recordings/new
• /recordings/search
• /recordings/ID

But if I don’t update the URLs to match the label, then I’m just going to have to live with the mismatch.

I’m just thinking out loud here. I think I should definitely update the label. I just won’t make any decision on changing URLs for a while yet.

A recurring theme in my writing and talks is “lay off the JavaScript, people!” But I have to make a conscious effort to specify that I mean client-side JavaScript.

I thought it would be obvious from the context that I was talking about the copious amounts of JavaScript being shipped to end users to download, parse, and execute. But nothing’s ever really obvious. If I don’t explicitly say JavaScript in the browser, then someone inevitably thinks I’m having a go at JavaScript, the language.

I have absolutely nothing against JavaScript the language. Just like I have nothing against Python or Ruby or any other language that you might write with on your machine or your web server. But as soon as you deliver bytes over the wire, I start having opinions. It just so happens that JavaScript is the universal language for client-side coding so that’s why I call for restraint with JavaScript specifically.

There was a time when JavaScript only existed in web browsers. That changed with Node. Now it’s possible to write code for your web server and code for web browsers using the same language. Very handy!

But just because it’s the same language doesn’t mean you should treat it the same in both circumstance. As Remy puts it:

There are two JavaScripts.

One for the server - where you can go wild.

One for the client - that should be thoughtful and careful.

I was reading something recently that referred to Eleventy as a JavaScript library. It really brought me up short. I mean, on the one hand, yes, it’s a library of code and it’s written in JavaScript. It is absolutely technically correct to call it a JavaScript library.

But in my mind, a JavaScript library is something you ship to web browsers—jQuery, React, Vue, and so on. Whereas Eleventy executes its code in order to generate HTML and that’s what gets sent to end users. Conceptually, it’s like the opposite of a JavaScript library. Eleventy does its work before any user requests a URL—JavaScript libraries do their work after a user requests a URL.

To me it seems obvious that there should an entirely different mindset for writing code intended for a web browser. But nothing’s ever really obvious.

I remember when Node was getting really popular and npm came along as a way to manage all the bundles of code that people were assembling in their Node programmes. Makes total sense. But then I thought I heard about people using npm to do the same thing for client-side code. “That can’t be right!” I thought. I must’ve misunderstood. So I talked to someone from npm and explained how I must be misunderstanding something.

But it turned out that people really were treating client-side JavaScript no different than server-side JavaScript. People really were pulling in megabytes of other people’s code to ship to end users so that they could, I dunno, left pad numbers or something.

Listen, I don’t care what you get up to in the privacy of your own codebase. But don’t poison the well of the web with profligate client-side JavaScript.

If you’re travelling around Ireland, you may come across some odd pieces of 19th century architecture—walls, bridges, buildings and roads that serve no purpose. They date back to The Great Hunger of the 1840s. These “famine follies” were the result of a public works scheme.

The thinking went something like this: people are starving so we should feed them but we can’t just give people food for nothing so let’s make people do pointless work in exchange for feeding them (kind of like an early iteration of proof of work for cryptobollocks on blockchains …except with a blockchain, you don’t even get a wall or a road, just ridiculous amounts of wasted energy).

This kind of thinking seems reprehensible from today’s perspective. But I still see its echo in the work ethic espoused by otherwise smart people.

Here’s the thing: there’s good work and there’s working hard. What matters is doing good work. Often, to do good work you need to work hard. And so people naturally conflate the two, thinking that what matters is working hard. But whether you work hard or not isn’t actually what’s important. What’s important is that you do good work.

If you can do good work without working hard, that’s not a bad thing. In fact, it’s great—you’ve managed to do good work and do it efficiently! But often this very efficiency is treated as laziness.

Sensible managers are rightly appalled by so-called productivity tracking because it measures exactly the wrong thing. Those instruments of workplace surveillance measure inputs, not outputs (and even measuring outputs is misguided when what really matters are outcomes).

They can attempt to measure how hard someone is working, but they don’t even attempt to measure whether someone is producing good work. If anything, they actively discourage good work; there’s plenty of evidence to show that more hours equates to less quality.

I used to think that must be some validity to the belief that hard work has intrinsic value. It was a position that was espoused so often by those around me that it seemed a truism.

But after a few decades of experience, I see no evidence for hard work as an intrinsically valuable activity, much less a useful measurement. If anything, I’ve seen the real harm that can be caused by tying your self-worth to how much you’re working. That way lies burnout.

We no longer make people build famine walls or famine roads. But I wonder how many of us are constructing little monuments in our inboxes and calendars, filling those spaces with work to be done in an attempt to chase the rewards we’ve been told will result from hard graft.

I’d rather spend my time pursuing the opposite: the least work for the most people.

In two of my recent talks—In And Out Of Style and Design Principles For The Web—I finish by looking at three different components:

1. a button,
2. a dropdown, and
3. a datepicker.

In each case you could use native HTML elements:

1. button,
2. select, and
3. input type="date".

Or you could use divs with a whole bunch of JavaScript and ARIA.

In the case of a datepicker, I totally understand why you’d go for writing your own JavaScript and ARIA. The native HTML element is quite restricted, especially when it comes to styling.

In the case of a dropdown, it’s less clear-cut. Personally, I’d use a select element. While it’s currently impossible to style the open state of a select element, you can style the closed state with relative ease. That’s good enough for me.

Still, I can understand why that wouldn’t be good enough for some cases. If pixel-perfect consistency across platforms is a priority, then you’re going to have to break out the JavaScript and ARIA.

Personally, I think chasing pixel-perfect consistency across platforms isn’t even desirable, but I get it. I too would like to have more control over styling select elements. That’s one of the reasons why the work being done by the Open UI group is so important.

But there’s one more component: a button.

Again, you could use the native button element, or you could use a div or a span and add your own JavaScript and ARIA.

Now, in this case, I must admit that I just don’t get it. Why wouldn’t you just use the native button element? It has no styling issues and the browser gives you all the interactivity and accessibility out of the box.

I’ve been trying to understand the mindset of a developer who wouldn’t use a native button element. The easy answer would be that they’re just bad people, and dismiss them. But that would probably be lazy and inaccurate. Nobody sets out to make a website with poor performance or poor accessibility. And yet, by choosing not to use the native HTML element, that’s what’s likely to happen.

I think I might have finally figured out what might be going on in the mind of such a developer. I think the issue is one of control.

When I hear that there’s a native HTML element—like button or select—that comes with built-in behaviours around interaction and accessibility, I think “Great! That’s less work for me. I can just let the browser deal with it.” In other words, I relinquish control to the browser (though not entirely—I still want the styling to be under my control as much as possible).

But I now understand that someone else might hear that there’s a native HTML element—like button or select—that comes with built-in behaviours around interaction and accessibility, and think “Uh-oh! What if there unexpected side-effects of these built-in behaviours that might bite me on the ass?” In other words, they don’t trust the browsers enough to relinquish control.

I get it. I don’t agree. But I get it.

If your background is in computer science, then the ability to precisely predict how a programme will behave is a virtue. Any potential side-effects that aren’t within your control are undesirable. The only way to ensure that an interface will behave exactly as you want is to write it entirely from scratch, even if that means using more JavaScript and ARIA than is necessary.

But I don’t think it’s a great mindset for the web. The web is filled with uncertainties—browsers, devices, networks. You can’t possibly account for all of the possible variations. On the web, you have to relinquish some control.

Still, I’m glad that I now have a bit more insight into why someone would choose to attempt to retain control by using div, JavaScript and ARIA. It’s not what I would do, but I think I understand the motivation a bit better now.

Last month I wrote about writing on web.dev. At that time, the first five parts of a fourteen-part course on responsive design had been published. I’m pleased to say that the next five parts are now available. They are:

1. Typography
2. Responsive images
3. The picture element
4. Icons
5. Theming

It wasn’t planned, but these five modules feel like they belong together. The first five modules were concerned with layout tools—media queries, flexbox, grid, and even container queries. The latest five modules are about the individual elements of design—type, colour, and images. But those elements are examined through the lens of responsiveness; responsive typography with clamp, responsive colour with prefers-color-scheme, and responsive images with picture and srcset.

The final five modules should be available later this month. In the mean time, I hope you like the first ten modules.

If you download Safari Technology Preview you can test drive features that are on their way in Safari 15. One of those features, announced at Apple’s World Wide Developer Conference, is coloured browser chrome via support for the meta value of “theme-color.” Chrome on Android has supported this for a while but I believe Safari is the first desktop browser to add support. They’ve also added support for the media attribute on that meta element to handle “prefers-color-scheme.”

This is all very welcome, although it does remind me a bit of when Internet Explorer came out with the ability to make coloured scrollbars. I mean, they’re nice features’n’all, but maybe not the most pressing? Safari is still refusing to acknowledge progressive web apps.

That’s not quite true. In her WWDC video Jen demonstrates how you can add a progressive web app like Resilient Web Design to your home screen. I’m chuffed that my little web book made an appearance, but when you see how you add a site to your home screen in iOS, it’s somewhat depressing.

The steps to add a website to your home screen are:

1. Tap the “share” icon. It’s not labelled “share.” It’s a square with an arrow coming out of the top of it.
2. A drawer pops up. The option to “add to home screen” is nowhere to be seen. You have to pull the drawer up further to see the hidden options.
3. Now you must find “add to home screen” in the list

• Copy
• Add to Reading List
• Add Bookmark
• Add to Favourites
• Find on Page
• Add to Home Screen
• Markup
• Print

It reminds of this exchange in The Hitchhiker’s Guide To The Galaxy:

“You hadn’t exactly gone out of your way to call attention to them had you? I mean like actually telling anyone or anything.”

“But the plans were on display…”

“On display? I eventually had to go down to the cellar to find them.”

“That’s the display department.”

“With a torch.”

“Ah, well the lights had probably gone.”

“So had the stairs.”

“But look you found the notice didn’t you?”

“Yes,” said Arthur, “yes I did. It was on display in the bottom of a locked filing cabinet stuck in a disused lavatory with a sign on the door saying ‘Beware of The Leopard.’”

Safari’s current “support” for adding progressive web apps to the home screen feels like the minimum possible …just enough to use it as a legal argument if you happen to be litigated against for having a monopoly on app distribution. “Hey, you can always make a web app!” It’s true in theory. In practice it’s …suboptimal, to put it mildly.

Still, those coloured tab bars are very nice.

It’s a little bit weird that this stylistic information is handled by HTML rather than CSS. It’s similar to the meta viewport value in that sense. I always that the plan was to migrate that to CSS at some point, but here we are a decade later and it’s still very much part of our boilerplate markup.

Some people have remarked that the coloured browser chrome can make the URL bar look like part of the site so people might expect it to operate like a site-specific search.

I also wonder if it might blur “the line of death”; that point in the UI where the browser chrome ends and the website begins. Does the unified colour make it easier to spoof browser UI?

Probably not. You can already kind of spoof browser UI by using the right shade of grey. Although the removal any kind of actual line in Safari does give me pause for thought.

I tend not to think of security implications like this by default. My first thought tends to be more about how I can use the feature. It’s only after a while that I think about how bad actors might abuse the same feature. I should probably try to narrow the gap between those thoughts.

On the Coil developers site there’s a page proudly answering the question who is web monetized?

You’ll some familiar sites in there: CSS Tricks, A List Apart, and even this humble website, adactio.com.

But lest you think that this social proof is in any way an endorsement, I should probably clarify what my experience with Coil has been like.

Coil itself is grand. You get an identifier and you add it to your website in a meta element, much like you would do with indie web endpoints for webmentions or micropub.

The problem is with how you then actually get hold of any money that is owed to you from micropayments. Coil doesn’t handle this directly. You have to set up a “wallet” with a third-party service and therein lies the problem.

They are all terrible.

I’m not talking about the hoops you have to jump through to set up an account. I get it. This is scary financial stuff so of course I’ll need to scan my passport and hand over loads of information (more than is needed to open an actual bank account with, say, Monzo).

No, the problem is the stench of crypto.

I tried Stronghold for a while. They really, really don’t want you to use boring old-fashioned currencies like the euro or the pound. There’s also Gatehub. Same. And there’s Uphold. Also a shell game.

I’ve been using Coil and Uphold for a while now, and I’ve amassed a grand total of £6.06 — woo-hoo! So I log into my account and attempt to transfer that sweet, sweet monetisation and …I can’t.

The amount needs to be greater than or equal to £11.53 GBP

But I can still exchange that £6.06 for magic beans like Bitcoin, XRP, and Ether.

The whole thing smells of grift and it feels icky to be in any way associated with it. I understand why Coil needs to partner with existing payment providers, but it would be nice if just one of them weren’t propping up ponzi schemes. If anyone has found a way to get web monetisation to work without needing like you need to take a shower afterwards, I’d love to hear about it.

When I was braindumping my thoughts prompted by last week’s UX Fest conference, I wrote about dark patterns.

Well, actually I wrote about deceptive dark patterns. That was a deliberate choice.

The phrase “dark pattern” is …problematic. We really don’t need to be associating darkness with negativity any more than we already do in our language and culture.

This is something I discussed with Melissa Smith after her talk on this topic. The consensus in general seems to be that the terminology is far from ideal, but it’s a bit late to change it now (I’m sure if Harry were coining the term today, he would choose a different phrase).

The defining characteristic of a “dark” pattern is that intentionally deceptive. How about we shift the terminology to talk about deceptive patterns?

Now, I get that inertia is a powerful force and it would be confusing to try do to a find-and-replace on all the resources that already exist on documenting “dark” patterns. So here’s a compromise:

From here on out, let’s start using the adjective “deceptive” in addition to the existing adjective “dark.” That’s what I did in my blog post. I only used the phrase “deceptive dark patterns.”

If we do that consistently, then after a while we’ll be able to drop one of those adjectives—“dark”—and refer to “deceptive patterns.”

Personally I’d love it if we could change the terminology overnight—and I’m quite heartened by the speed at which we changed our Github branches from “master” to “main”—but being pragmatic, I think this approach stands a greater chance of success.

Who’s with me?

I’m subscribed to a lot of blogs in my RSS reader. I follow some people because what they write about is very different to what I know about. But I also follow lots of people who have similar interests and ideas to me. So I’m not exactly in an echo chamber, but I do have the reverb turned up pretty high.

Sometimes these people post thoughts that are eerily similar to what I’ve been thinking about. Ethan has been known to do this. Get out of my head, Marcotte!

But even if Ethan wasn’t some sort of telepath, he’d still be in my RSS reader. We’re friends. Lots of the people in my RSS reader are my friends. When I read their words, I can hear their voices.

Then there are the people I’ve never met. Like Desirée García, Piper Haywood, or Jim Nielsen. Never met them, don’t know them, but damn, do I enjoy reading their blogs. Last year alone, I ended up linking to Jim’s posts ten different times.

Or Baldur Bjarnason. I can’t remember when I first came across his writing, but it really, really resonates with me. I probably owe him royalties for the amount of times I’ve cited his post Over-engineering is under-engineering.

His latest post is postively Marcottian in how it exposes what’s been fermenting in my own mind. But because he writes clearly, it really helps clarify my own thinking. It’s often been said that you should write to figure out what you think, and I can absolutely relate to that. But here’s a case where somebody else’s writing really helps to solidify my own thoughts.

Which type of novelty-seeking web developer are you?

It starts with some existentialist stock-taking. I can relate, what with the whole five decades thing. But then it turns the existential questioning to the World Wide Web itself, or rather, the people building the web.

In a way, it’s like taking the question of the great divide (front of the front end and back of the front end), and then turning it 45 degrees to reveal an entirely hidden dimension.

In examining the nature of the web, he hits on the litmus of how you view encapsulation:

I mention this first as it’s the aspect of the web that modern web developers hate the most without even giving it a label. Single-Page-Apps and GraphQL are both efforts to eradicate the encapsulation that’s baked into the foundation of every layer of the web.

Most modern devs are trying to get rid of it but it’s one of the web’s most strategic advantages.

I hadn’t thought of this before.

By default, if you don’t go against the grain of the web, each HTTP endpoint is encapsulated from each other.

Moreover, all of this can happen really fast if you aren’t going overboard with your CSS and JS.

He finishes with a look at another of the web’s most powerful features: distribution. In between are the things that make the web webby: hypertext and flexibility (The Dao of the Web).

It’s the idea that the web isn’t a single fixed thing but a fluid multitude whose shape is dictated by its surroundings.

This resonates with me because it highlights two different ways of viewing the web.

On the one hand, you can see the web purely as a distribution channel. In the past you might have been distributing a Flash movie. These days you might be distributing a single page app. Either way, the web is there as a low-friction way of getting your creation in front of other people.

The other way of building for the web is to go with the web’s grain, embracing flexibility and playing to the strengths of the medium through progressive enhancement. This is the distinction I was getting at when I talked about something being not just on the web, but of the web.

With that mindset, Baldur then takes us through some of the technologies that he’s excited about, like SvelteKit and Hotwire. I think it’s the same mindset that got me excited about service workers. As Baldur says:

They are helping the web become better at being its own thing.

That’s my tagline right there.