I woke up today to a very annoying new bug in Firefox. The browser shits the bed in an unpredictable fashion when rounding up single pixel line widths in SVG. That’s quite a problem on The Session where all the sheet music is rendered in SVG. Those thin lines in sheet music are kind of important.
Browser bugs like these are very frustrating. There’s nothing you can do from your side other than filing a bug. The locus of control is very much with the developers of the browser.
That said, I understand the argument for converging on a single rendering engine. I don’t agree with it but I understand it. It’s like this…
Back in the bad old days of the original browser wars, the browser companies just made shit up. That made life a misery for web developers. The Web Standards Project knocked some heads together. Netscape and Microsoft would agree to support standards.
So that’s where the bar was set: browsers agreed to work to the same standards, but competed by having different rendering engines.
There’s an argument to be made for raising that bar: browsers agree to work to the same standards, and have the same shared rendering engine, but compete by innovating in all other areas—the browser chrome, personalisation, privacy, and so on.
Like I said, I understand the argument. I just don’t agree with it.
One reason for zeroing in a single rendering engine is that it’s just too damned hard to create or maintain an entirely different rendering engine now that web standards are incredibly powerful and complex. Only a very large company with very deep pockets can hope to be a rendering engine player. Google. Apple. Heck, even Microsoft threw in the towel and abandoned their rendering engine in favour of Blink and V8.
The ECMAScript, HTML, and CSS specifications today are (for the most part) stellar technical documents whose algorithms can be implemented with considerably less effort and guesswork than in the past.
I’ll be watching that project with interest. Not because I plan to use the brower. I’d just like to see some evidence against the complexity argument.
Meanwhile most other browser projects are building on the raised bar of a shared browser engine. Blisk, Brave, and Arc all use Chromium under the hood.
Arc is the most interesting one. Built by the wonderfully named Browser Company of New York, it’s attempting to inject some fresh thinking into everything outside of the rendering engine.
Experiments like Arc feel like they could have more in common with tools-for-thought software like Obsidian and Roam Research. Those tools build knowledge graphs of connected nodes. A kind of hypertext of ideas. But we’ve already got hypertext tools we use every day: web browsers. It’s just that they don’t do much with the accumulated knowledge of our web browsing. Our browsing history is a boring reverse chronological list instead of a cool-looking knowledge graph or timeline.
For inspiration we can go all the way back to Vannevar Bush’s genuinely seminal 1945 article, As We May Think. Bush imagined device, the Memex, was a direct inspiration on Douglas Engelbart, Ted Nelson, and Tim Berners-Lee.
The article describes a kind of hypertext machine that worked with microfilm. Thanks to Tim Berners-Lee’s World Wide Web, we now have a global digital hypertext system that we access every day through our browsers.
But the article also described the idea of “associative trails”:
Wholly new forms of encyclopedias will appear, ready made with a mesh of associative trails running through them, ready to be dropped into the memex and there amplified.
Our browsing histories are a kind of associative trail. They’re as unique as fingerprints. Even if everyone in the world started on the same URL, our browsing histories would quickly diverge.
Bush imagined that these associative trails could be shared:
The lawyer has at his touch the associated opinions and decisions of his whole experience, and of the experience of friends and authorities.
Heck, making a useful browsing history could be a real skill:
There is a new profession of trail blazers, those who find delight in the task of establishing useful trails through the enormous mass of the common record.
Taking something personal and making it public isn’t a new idea. It was what drove the wave of web 2.0 startups. Before Flickr, your photos were private. Before Delicous, your bookmarks were private. Before Last.fm, what music you were listening to was private.
I’m not saying that we should all make our browsing histories public. That would be a security nightmare. But I am saying there’s a lot of untapped potential in our browsing histories.
Let’s say we keep our browsing histories private, but make better use of them.
From what I’ve seen of large language model tools, the people getting most use of out of them are training them on a specific corpus. Like, “take this book and then answer my questions about the characters and plot” or “take this codebase and then answer my questions about the code.” If you treat these chatbots as calculators for words they can be useful for some tasks.
Large language model tools are getting smaller and more portable. It’s not hard to imagine one getting bundled into a web browser. It feeds on your browsing history. The bigger your browsing history, the more useful it can be.
This is the flyer that Tim Berners-Lee and Robert Cailliau distributed at the Hypertext 91 Conference—the one where their submission was infamously rejected.
The WWW project merges the techniques of information rerieval and hypertext to make an easy but powerful global information system.
The project is based on the philosophy that much academic information should be freely available to anyone. lt aims to allow information sharing within internationally dispersed teams, and the dissemination of information by support groups.
If posts in a social media app do not have URLs that can be linked to and viewed in an unauthenticated browser, or if there is no way to make a new post from a browser, then that program is not a part of the World Wide Web in any meaningful way.
But I must admit that something didn’t sit quite right about the mocking tone he took on the matter of “the fundamentals” (whatever that may mean). Chris shares my misgivings:
Those websites that don’t load on slow connections, or break completely when a JS file fails to load, or don’t work for people with visual or physical impairments?
That’s not an issue of time. It’s an issue of fundamentals.
I think I agree with Laurie that there’s basically no such thing as fundamental technologies (and if there is such a thing, the goalposts are constantly moving). But I agree with Chris with that there is such a thing as fundamental concepts. On the web, for example, accessibility is a core principle of its design that should, in my opinion, be fundamental.
Do I wanna see teenagers building frivolous websites? Absolutely. But when people are getting paid well to build our digital world, they have a responsibility to ensure the right to engage with that world for everyone.
Hello, my name is Jeremy and I am speaking to you today from Brighton on the south coast of England.
I want to tell you about something that happened here in Brighton back in 1985 (pretty sure it took place in one of those buildings along the seafront there). In 1985 Brighton was host to the International Information Theory Symposium. Fascinating.
Something exciting happened there. Word began to go around that there was an unexpected guest attending the event. This unexpected guest was this man, Claude Shannon. The way it was described later was somebody said it was as if Newton had showed up at a physics conference.
He wasn’t even meant to be there, but he was convinced at the dinner after the event to get up and say a few words. And he did, he got up and he started to talk, but he felt like he was losing the audience. So he proceeded to do some juggling.
That’s so Claude Shannon. He was very much into games. He took games very seriously. He was a very playful kind of person.
For example, he invented this machine, which is called the most beautiful machine, also the most useless machine. But I think it’s just wonderful. I mean, it’s like the perfect encapsulation of cybernetics, the ideal feedback loop.
But the reason why people were excited that Claude Shannon was at this event wasn’t because of the most beautiful machine. And it wasn’t because of his juggling. It was because of information theory.
Because Claude Shannon, it’s not like he just revolutionized the field of information theory; Claude Shannon pretty much invented the field of information theory in one fell swoop. In a paper in 1948 called the Mathematical Theory of Communication.
Here’s the TLDR. This is the mathematical part. I won’t go into the details of the mathematical part, but what I recall from Claude Shannon’s work is that he was able to effectively boil information down into fundamental particles. The idea that there’s a single bit of information.
This idea of entropy, the idea that for information to travel between communicator and the receiver, you’ve got the signal that you’re trying to transmit, but there’s also noise. And this noise is unavoidable.
And like I said, this idea of the bit; that any piece of information could be reduced down to a fundamental particle: a one or a zero; on or off, which of course is exactly how computers work. So it’s no exaggeration to say that Claude Shannon is like the father of the digital age.
And one other thing I take from Claude Shannon’s work is that when it comes to communication of information, context matters. In other words, that the expectation between the receiver and the communicator can make a lot of difference.
So to give you an example of shared context being very important in information communication I want to illustrate it with a story from the pre-digital age. This is a story from the age of the electrical telegraph.
Now this story is probably completely apocryphal. In some versions of the story, it involves the novelist Victor Hugo. In other versions, it’s Oscar Wilde. But the point is there’s an author. He’s just published a book and now he’s gone off on holiday after writing the book. But while he’s on holiday, he’s really curious to know, how is the book doing? What are the sales like?
So he sends a telegram to his publisher, but because there’s enough shared context between the publisher and the author, all he sends is a single character. A question mark.
And then, because there’s this shared context between the publisher and the author and the publisher wants to let the author know that actually sales are going really, really well, the publisher also sends back a telegram with a single character. An exclamation mark.
So this is a classic example of the importance of context. I mean, you’re just sending a single character and yet both parties understand the message being conveyed.
Context matters. Shared context matters.
Now I want to try an experiment with you to test how much shared context there is between me (I’m going to try to transmit a message) and you (the receiver of the message).
So we’re starting with a blank slate. And now I’m going to provide one piece of information. Okay. Here’s the piece of information. Probably doesn’t tell you much. A diagonal line. There’s not enough context here.
All right. Back to the blank slate. I’m now going to provide another piece of information.
Okay. Again, in isolation, this probably doesn’t tell you much just another diagonal line. But if I combine it with the first bit of information, then now maybe it starts to become something you can parse. And if I provide just one more bit of information, now maybe it clicks into place that the piece of information I’m trying to convey is ten minutes past ten.
And yet all I’ve done here is I’ve provided you with two diagonal lines in a circle. Yet somehow two diagonal lines in a circle, when we have the shared context of how to read an analogue clock face, is enough to communicate ten minutes past ten.
(The time, by the way, is completely arbitrary. The only reason I chose that time is just that if you ever look at an advertisement for a watch, it’ll usually be ten past ten because the angles of the arms on the watch nicely frame the logo of the watchmaker.)
But anyway, the point is: with enough shared context, two diagonal lines in a circle are enough for me to communicate the piece of information, “ten minutes past ten o’clock.”
I mean, maybe you’re a digital native born in a 21st century, in which case you’re looking at this and thinking, “I just see two diagonal lines in a circle”, but if you can read an analogue clock face, then we have that shared context.
Where did this context come from? Why is it that that clock faces are set up the way they are? Why do clocks go clockwise? It seems like a fairly arbitrary decision.
It is somewhat arbitrary, but one neat solution is that the reason why clocks go in a clockwise direction is that that’s the way that a shadow on a sundial would travel …in the Northern hemisphere.
Now if you look at a sundial in the Southern hemisphere, like this one here—this is in Wellington, in New Zealand—the shadow would actually go in a counterclockwise direction.
So really it’s almost an accident of history that we have clocks that go clockwise. If clocks had been invented in the Southern hemisphere, then they would go in the other direction. It’s pretty arbitrary, but now we’ve decided, we’ve kind of settled on this arbitrary movement of clocks that they go clockwise and we’re stuck with it.
Because inertia is a very powerful force. If you tried to change the way the clocks work you’d have your work cut out for you, even if the reason why clocks work the way they do is arbitrary to begin with.
You know, a very wise person once said the most dangerous phrase in the English language is “We’ve always done it that way.” And that very wise person was the brilliant computer scientist, rear admiral Grace Hopper.
She used to say:
Humans are allergic to change.
I try to fight that. And that’s why I have a clock on my wall that runs counterclockwise.
Right? It kind of drives home this idea that, hey, this is an arbitrary decision.
And it’s kind of weird for us to look at a clock that runs counterclockwise. I actually managed to find a watch a few years ago that worked like this, that ran counterclockwise. And I wore it for a while and I was able to train my brain to read the clock this way. And it worked fine, but it completely broke my brain for reading normal clocks. So I kind of had to just stop doing it.
But I’m fascinated by these examples of fairly arbitrary decisions made sometime in the past that you’re then stuck with, because it’s very hard to change the inertia. But only recently did I find out that there’s a term for this phenomenon. This is called path dependence.
History is full of path dependence. The classic example is if you wanted to make a new train or a new stretch of railway track, you’re gonna have to use the existing gauge of the railway in question. Now it’s not that there’s one gauge of railway that’s better or worse than any other gauge, but if someone’s made that decision in the past, it’s very hard to change. And you really do want to settle on one gauge so that you don’t have to switch trains when you move between different parts of a country (this actually happened down in Australia, where they had different gauges for the railways, it was kind of a mess).
It’s the canonical example of why you need standards. But really the point of standards isn’t necessarily to enshrine the best way of doing something. The point of standards is to enshrine the agreement. “Hey, let’s all agree to do things this.”
Whether the standard is good, bad, better or worse than other ways of doing things is in some ways less important than the agreement. You just need to have everyone agree on something.
Like, there’s a standard for which side of the road you drive on in your country. And it doesn’t really matter whether the standard is for it to be on the left side of the road or the right hand side of the road. But it really matters that you all agree on the same side of the road.
Agreement and standards brings us very nicely to the World Wide web. Because I think the World Wide web is a fantastic example of agreement.
The web is agreement
There’s a friend of mine, Paul Downey, who does these wonderful illustrations. Fantastic. He has this one called “the web is agreement.” Whenever I think about the word agreement, this is what I think of: that the web is agreement. And he does these kind of Hieronymous Bosch and Breughel-like images of all the different formats and standards that we use on the web.
And if you think about what the World Wide Web is, this combination of HTTP and URLs and HTML. This was, you know, when the web was first created. Yeah, these are just agreements.
I mean, HTTP is a protocol and that word protocol literally means agreement. (If you think about diplomatic protocols that are like diplomatic agreements, right?)
URLs, “Hey, let’s all agree to use this addressing scheme.”
And HTML, “If we all agree to use this format, then we get interoperability.”
So these formats, these protocols, this set of standards or agreements came from Sir Tim Berners-Lee. This was back when he was working at CERN at the nuclear physics laboratory in the late 1980s, early ’90s.
I’m somewhat fascinated about the birth of the web, which is why it was a huge honour and pleasure for me to be invited to CERN a few years back. This was in the run up to the 30th anniversary of the original proposal that Tim Berners-Lee submitted for what later became the World Wide web. And we did this project and you can check out the project at worldwideweb30.com.
This wonderful group of people came together for a week to kind of hack on something. And what we were hacking on was this project to recreate the very first web browser …but that you could run it in a modern web browser. This is what it looked like.
The first web browser was also called confusingly WorldWideWeb. It was created by Tim Berners-Lee on his NeXT machine. And this was the first demonstration of those three things working together: HTTP, URLs and HTML.
Now I say we were working on this. I didn’t make this part. This was the really smart bit; much cleverer people were working on the smart bit. What I worked on was the website that accompanied the project.
Remember this is coming up on the 30th anniversary of the web, and I thought it’d be fascinating to not just graph out the 30 years that the web has existed, but also look back at the 30 years before the web existed to see what were the influences that fed into the web.
What I was looking at here really was the path dependencies. What were the path dependencies in computing and networks, hypertext and formats that all fed into that creation of the World Wide Web?
So let’s take formats, for example. Tim Berners-Lee creates HTML along with URLs and HTTP.
And if I show you these elements, they should be quite familiar to you. You recognize what this language is, right? Yeah. Clearly this language is…
SGML. Standard Generalized Markup Language.
Specifically, it’s a flavor of SGML that was being used in CERN at the time. And Tim Berners-Lee thought rather than create his own format, he would use what people were already used to.
He kind of had the same insight that Grace Hopper had, that humans are allergic to change. But instead of trying to change that, he sort of went with the flow.
So he took SGML and basically copied it to make HTML, introducing one new element, which is the, A element.
So, you know, there’s a path dependency, even in the name, right? You think Standard Generalized Markup Language. Oh right. And now we have HyperText Markup Language. So even the name HTML seems to have a path dependency to SGML.
But it goes deeper.
SGML was a specialized version of GML. And GML was supposed to stand for Generalized Markup Language. Except the people who created GML were named Goldfarb, Mosher, and Lorie, which is probably the real reason why it was named GML.
And later we got SGML and then we got HTML. So it turns out there’s a path dependency in the phrase “HTML” that goes back to three dudes wanting to get their names into a format many, many years ago.
What about styling when it came to the early web?
There’s no CSS at this point. But if you look at this first web browser—and this is the very first web page on the web, which is still available at its original URL—you can see that different types of elements are styled differently.
The links are styled differently than the text around them. The heading is styled differently to the body copy. This definition list has formatting going on. You can see the spacing there. So something is doing some styling.
Well, when we were at this hack week at CERN, we had access to the original source code for this project. We found this file in there.
And this is, I guess, the user agent style sheet. This is the bit that tells the first web browser how to style headings, how to style lists, how to style definitions.
It’s not very readable, but you can tell that, you know, there’s a lot of values here, not many property names, but if you squinted it just right, you can imagine, okay, this is some form of a style sheet.
It became clear though that it wasn’t enough to just allow the user agent to do the styling. Authors—that’s developers and designers—authors also needed a way to provide styling information.
Now for a while there, it looked like the way this was gonna happen was going to be in HTML. People started adding these proprietary elements and attributes to HTML that were all about presentation, all about styling. And that’s not what HTML is for. HTML is about structure, about the semantics, the meaning of a document.
It was really important that there’d be some kind of separation of concerns, that you would use one format—HTML—for your structure, and that there should be a separate format, some other kind of language, for styling.
The question is, what should that other format be?
Well, pretty early on, some proposals started coming in early mailing lists to do with the World Wide Web. Pretty much everyone on the web in the first few months was making their own web browser. It was by nerds for nerds.
I think the first proposal for some kind of styling for authors came from Rob Raisch, who was at O’Reilly at the time.
He sent an email to the www-talk mailing list in June of 1993, with this proposal as a way of styling.
Now, again, looking at this, it’s not CSS, but if you squint just right, you can sort of make sense of it. It’s kind of like looking at a clock running counterclockwise. It’s not what we’re used to, but you feel like you could parse it.
Clearly the priority here was to do with brevity. We’ve got these two character things like FO for font, FA for face, HE for Helvetica, SI for size. You put that all together and you can say the font face should be Helvetica and the font size should be 18 of whatever unit we’re talking about here.
So, you know, just about able to parse it, there is the concept here of, you know, some kind of selector, right? The way that we say we’re talking about the BODY, we’re talking about the paragraph or talking about B or I.
The next proposal that came along was by Pei Wei, who was building the Viola web browser. He sent an email to the www-talk mailing list in October of 1993. And he was able to put his entire style sheet in his proposal.
This is what it looks like. Kind of similar to what we saw before. We’ve got the idea of properties and values, but with equal signs rather than colons that we used to now.
But what’s really interesting here is this idea of nesting. We’ve got nesting going on in this proposal which is something that we’re really only just getting in CSS.
Håkon Wium Lie
Now, the next proposal came from Håkon Wium Lie. This was October of 1994 and he called his proposal Cascading HTML Style Sheets. And it looked like this.
h1.font.size = 24pt 100%
And again, you can kind of parse this, right? It’s not what we’re used to, but you squint at it and you can make sense of it. You can see the way that the selector and the properties are kind of scrunched together with this dot syntax. And again, there’s an equal sign rather than a colon, but we get it. It’s like, okay, the font size of an H1 element should be 24 points. Got it.
But wait, what’s this percentage after it, like this 100% or this 40%?
Well, this is a really interesting part of this proposal. This was this idea of influence. The idea that an author should be able to effectively say how much they care about a particular style being applied.
So if you really want that heading to be that size, you say 100%, I care about this. But if you only half care, you could say 50%.
And the idea was that users would also be providing styles and users would also specify how much they cared, how much influence they wanted to exert exert on the styles.
And then there’s kind of a bit of hand-wavy logic where it’s like, “And then the user agent figures out what the final style should be.”
And that last part it turned out was really hard to do. So this idea of influence somewhat fell by the wayside. But I think it’s very powerful and it definitely matched the ideas of Tim Berners-Lee with his first web browser, this idea that the web should be a read and write medium.
Because that first web browser—WorldWideWeb—wasn’t just a browser. It was also an editor.
The idea was you would open a document from the web, you’re looking at it and you think “I wanna make changes to this document. I’m gonna create my own copy, put it on my server and make the changes.”
Now it turned out that was really hard. And so that was one of the first things that got dropped from the World Wide Web, which is a bit of a shame because I think it is a very, very powerful idea, a very empowering idea.
We somewhat got back this idea of a read/write web with things like wikis and blogs and even social media to a certain extent.
And the idea that users should have influence over the styles of a website? Well, that survived in web browsers for quite a while, with this concept of user style sheets.
This is different to user agent style sheets. This was literally that in your browser, you could specify styles to override what an author has specified.
This got dropped from browsers over time because it turned out to be a real power-user feature. Most people weren’t using this. These days, if you want to apply styles as a user, you have to install a browser extension, some kind of plugin. Or your operating system has some kind of translation of like, “these are my preferences at the operating system level” and those get translated to the browser.
I think it’s a real shame that we lost user style sheets. I thought it was a very empowering feature. I get it. It was, you know, somewhat of an edge case. It was power-user feature, but I think it’s a shame we lost it.
I do see, however, a bit of a resurgence in the idea of giving users control over styling with some of the things we’re seeing in CSS, particularly in the media queries level five, this idea of what are being called preference queries. You can say, you know, prefers a color scheme, like dark mode prefers reduced motion, prefers reduced data.
Now it’s a bit different because it’s still up to the author of the style sheet on that website to honour these preferences, right? You still have to write the styles to do the right thing to respect the colour scheme or reduced motion or reduced data.
Though, you know, some browsers are looking into automatically applying some of this stuff automatically: inverting colors and reducing motion.
And on the whole, I welcome the idea that users should have more of a say in how websites are styled. I think it’s a good thing.
So we’re seeing a bit of a resurgence of this idea of influence in modern CSS.
And speaking of modern CSS being somewhat foreshadowed in Håkon’s original proposal, here’s something else that was in that original proposal…
If you look at this, you can kind of figure out what’s going on. That there’s kind of a declaration at the top to say there’s a variable, if you like, that’s 12 points. And then that variable is used throughout the style sheet. It’s multiplied by different numbers.
And really, we’ve got this now in CSS, thanks to custom properties and calc(), right? The ability to set variables and do calculations on those variables. But it took a long time between this original proposal and this very modern CSS that we have today.
I think this is the first time we started to see colons rather than equal signs for properties and values. But again, you can see the way that the selectors and the properties sort of munged together with this dot syntax.
It’s parsable, right? Again, it’s like looking at a clock running counterclockwise, but you can understand what’s going on here.
In other words, the thinking behind the proposal. Because like I was kind of saying, you know, the standard itself, in some ways, isn’t the important thing. The important thing is the agreement. So let’s all try and agree on what we’re trying to accomplish with some kind of style sheet language. And I will also freely admit I’m just a sucker for design principles.
I’m fascinated by design principles. I even collect them. This is like my equivalent of my interesting rock collection. A collection of design principles at principles.adactio.com.
And if you go there, I’ve collected design principles from individuals, from organizations, and I have Bert’s principles there.
And they’re worth reading through, but one of the issues with Bert’s principles is there’s a lot of them. These are all the different things that feed into the design of a style sheet language. And these are all good things, but I think what’s missing here is some kind of prioritization.
Because the hard part about design principles, isn’t saying what you value. The hard part about design principles is saying we value one thing over another.
So let’s take two of these. We see simplicity and longevity. Well, do we value simplicity more than longevity? Do we value longevity more than simplicity? That’s actually the hard part, to specify the priorities.
So I think it’s a bit of a shame that there isn’t prioritization here, but I think it’s still fascinating that we can look at all of the things that Bert was imagining we have to balance in some kind of style sheet language.
Well, it became pretty clear that Bert and Håkon were working on the same sort of thing. And so they pooled their resources together and kids, that’s where CSS comes from. Jointly from Bert and Håkon.
And what they settled on—with all of those different design principles and all of the ideas from the different proposals that came before—this is what we got:
This one pattern. You’ve got a selector, a property and a value. Then we’ve got these special characters for syntax, right? Curly braces, colons, semicolons, but really it’s somewhat arbitrary. The point is that all of CSS pretty much can be boiled down to this one pattern: selector, property, value.
It’s a very simple pattern. And yet it allows for endless complexity. I mean, this is our shared context on the web for styling. If you think about the number of websites out there, right? Billions. And every one of them has a different style sheet and every one of them is different. And yet all of them use the same pattern at its root.
It’s the classic example of how a simple rule can create a complex system. And I think this might also be at the heart of why CSS can be misunderstood. Because this pattern is very simple and because it’s very simple, people might think well, CSS is therefore easy.
But there’s a difference between simplicity and easiness.
Like, you can learn the idea of CSS in an hour, right? Because effectively this pattern is it. You need to get your head around selector, property, value.
But you can then spend a lifetime trying to master CSS because of all the possible combinations of selectors and properties and values, right? It’s a lifetime of learning.
So this is where I think some of the disconnect comes with people thinking, “Oh yeah, I’ll pick up CSS. No problem. It’s easy.” And actually, no. It’s simple, but it’s not easy.
And CSS has grown over time, right? We keep getting more selectors, we keep getting more properties and we keep getting more values. It grows and grows while still maintaining this fundamental pattern.
And if we look at where the growth of CSS has come from, you know, a lot of the time it came from hacks. And I don’t mean literal CSS hacks, like the box model hack or tan hack for any anybody old enough to remember that.
I mean hacks in a sense of its original use of a clever solution to a problem, but probably not a great long term solution.
So the classic example of hacks on the web would’ve been layout. You know, in the early days we were using tables for layout. We had transparent gifs, one pixel by one pixel gifs that we would give width and height to allow us to make all the layouts we wanted. And it worked, but it was a hack.
So then we got CSS and we switched to using floats for layout, which was better. But, you know, it was still a hack because floats weren’t intended for layout. They were intended for, you know, having text flow around images.
And it’s only relatively recently in the history of the web that we finally are able to throw away our hacks and use proper layout tools.
Because now we’ve got flexbox and we’ve got grid and these are made for layout. It took quite a while for us to get there.
But you know, in the early days of the web, it wasn’t even clear if CSS should attempt to do layout or whether there should be a third format specifically for layout. Because maybe there needed to be that separation of concerns between structure (you’ve got HTML), styling (you’ve got CSS), and some third technology for layout which could be considered like its own its own thing.
I mean, if you think about it today, we kind of sequester layout into media queries. So you could imagine that being a separate technology.
But anyway, it became clear over time that CSS should be the home for layout as well as other kinds of styling.
And that’s what we’ve got now. We’ve got flexbox. We’ve got grid. We got proper layout on the web so we were able to stop using our hacks and use the real native tools.
It’s a similar story with typography.
If you wanted to use a font that wasn’t one of the system fonts that most people would have installed, well, you went into Photoshop and you made an image of text using the font you wanted, and now the user would have to download that image file and the text wasn’t selectable, it was a fixed width, it came with all sorts of problems. And we came up with very clever solutions to do what’s called image replacement in CSS, but they were all hacks.
And now we don’t need the hacks because we’ve got the @font-face rule. So we can literally specify the typeface we want to use. We can stop using the hacks.
We used a lot of hacks for graphic design as well. Things that we weren’t quite able to do natively in CSS.
This is the very 2005-feeling design. You see the way we’ve got that element there with rounded corners? And you see the way that there’s a gradient in the background of the page? Well, back in 2005, we didn’t have rounded corners in CSS and we didn’t have gradients.
So those rounded corners? Those are images that have been absolutely positioned into that element.
And that gradient is actually an image. It’s a one pixel wide, but very, very tall image that is tiled across the entire background.
So these were hacks and they worked, but obviously I wouldn’t need to do that today. Today, I’ve got border-radius to do my rounded corners and I got linear-gradient to do my gradients.
Ironically, right as we got the power to do rounded corners and gradients in CSS natively, we all collectively decided that, nah, actually what we want is flat design …which we could have been doing all along!
Anyway, let me show you another example. This is a website that dates back even further. This is my own personal website, adactio.com.
This design hasn’t really changed in over 20 years, but I have updated the technology.
Let’s the take image at the top. You can see that’s been given some treatment as a corner has been sliced out of one edge and a gradient has been applied so it fades out.
Now it used to be that I would have to do that in Photoshop. I’d take the original image, I’d slice off the corner, I’d add a gradient layer on top of it.
Well now I don’t need Photoshop because I can use clip-path to take off the corner. I can use a linear gradient using generated content—using the ::after pseudo element—to get the exact same result.
So now I’ve got something that looks pretty much the same, except where I had to do it in Photoshop before, now I’m able to do it natively in CSS.
And that might not sound like much of a win because the end result looks the same, right?
Except now when I’m applying a prefers-color-scheme style sheet and I give a dark mode to my site, I don’t have to make a separate image. Because this is being done natively in CSS. The gradient is in CSS. The clip path is in CSS. It’s all native to CSS.
This comes down to this idea of material honesty. About using the right material for the job, rather than using a material that’s pretending to be something else, whether that’s, you know, an image of text, pretending to be a font or an image of a gradient, pretending to be a real gradient or, you know, any of those graphic design tricks.
We can now be materially honest in CSS because we’ve got grid and flexbox and border-radius and @font-face. It’s more honest.
And also it’s easier, right? It’s less work to avoid the hacks.
Like, one of my favorite examples of something we got recently in CSS to avoid the hacks. It’s a small thing, but it makes a big difference in my opinion, is just styling things like check boxes and radio buttons.
We’ve always been able to do it, but it involved a hack where you’d hide the real checkbox off screen. And then you’d use a background image to show the different states of the checkbox. And it was fine and it worked and we could make it accessible.
But now we can just use accent-color and it’s easier.
So there’s been this movement from hacks to native CSS. And in a way, the hacks show the direction of travel. The hacks show us what the future could be.
The other place CSS has borrowed from or learned from, has been in our tools. Like the tools we use to generate our CSS.
Sass is the classic example, right? Sass is this enormously popular pre-processor for CSS and people were using Sass to do things you couldn’t do natively in CSS.
And I feel like one of the genius bits of design in Sass was that it worked a lot like the way HTML worked in the early days compared to SGML.
Remember, I was talking about how Tim Berners-Lee took SGML and literally turned it into HTML? Like, you were able to take an SGML document and change the file extension, and it would be a valid HTML document. And so that really helped with the adoption of HTML.
Well, that’s the same with Sass. You could take your existing style sheet document and just changed the file extension to .scss and it was already valid Sass, right? You didn’t have to learn a new syntax.
This isn’t supposition on my part—that this was a reason for the huge success of Sass—because actually Sass had two options, two different syntaxes, and you could choose.
There was this .sass syntax and the .scss syntax.
And with the .sass syntax, it used significant white space. It was more condensed, right? It used indentation.
But with .scss it used the syntax you were already familiar with from CSS.
So you could say that the .sass syntax was more efficient. You could say it’s a better format, but humans are allergic to change. And the .scss syntax was familiar enough that people go, “oh yeah, I get this.”
You could, you could take your existing CSS and just start using the features of Sass you wanted to.
And people overwhelmingly chose the .scss syntax over the .sass syntax. I got to meet Hampton Catlin who invented Sass and he confirmed the numbers for me. He said, yeah, it was a no brainer. Basically the .sass syntax lacks familiarity. It’s like looking at a clock running counter-clockwise.
But anyway, people started using Sass to do things like nesting, calculations (these mix-ins), variables; all these things that we now do in CSS.
Of course, the reason we can do these things in CSS is because Sass proved that there was a desire for these things.
So now you really don’t need Sass for a lot of stuff, but the reason you don’t need Sass is because of Sass. Sass paved the way. Sass showed that there was a demand for this stuff. And now it’s native in the CSS. We don’t need that tool anymore.
And I feel like that’s a test of a really good tool. A really successful tool is when it becomes redundant.
With jQuery you were able to do things using a CSS syntax, whereas otherwise you had to use this long winded DOM syntax of document.getElementsByTagName or getElementById.
Whereas in jQuerythe idea was, “Hey, if you already know how to select something in CSS, just use that syntax again!”
It’s using what people are familiar with. Humans are allergic to change.
But these days we don’t need jQuery because in the DOM, we’ve got querySelectorAll, we’ve got querySelector. We can use CSS selectors to do our DOM scripting.
Why don’t we need jQuery anymore? Because of jQuery.
jQuery showed that this was a really clever idea. It was something people wanted. And so now it’s been standardized. We don’t need jQuery.
So I really feel like the goal of any good library should be to make itself obsolete. It’s so successful, it’s no longer needed.
And you could kind of see this in the history of the web with Sass, with jQuery, even with something like Flash.
You know, Flash showed the way. It showed that, “Hey, we want a way to do animation. We need some way to do video on the web.” And people were using Flash because there was no other way to do those things.
Now we don’t need Flash or jQuery or Sass because we get them natively.
So all of these are almost like research and development for the web.
They’re kind of like hacks, but I think a better way of thinking about them is they’re more like polyfills—these are things we can use until we get a standardized way of doing it.
I think it’s fascinating to look at our tools and see what can they tell us about what’s coming into standards.
A whole set of tools are these methodologies that people came up with, like OOCSS from Nicole Sullivan. And there’s BEM. And SMACSS was another one. There’s a whole bunch.
But I’m fascinated by these because these aren’t an example of some technology we needed to lobby browser makers to implement. Because these are really agreements.
These are agreements. These are saying, “let’s all agree to structure our CSS in a certain way.” Nothing needed to change in browsers, right?
And all of these are testament to the power of the cascade. Because what they do is they almost deliberately limit the cascade, which is seen as almost being too powerful.
So they’re not really tools, they’re methodologies. Or another way of putting it is they are agreements.
Again, the power of saying “let’s all agree to do something.”
And the problem that most of them are trying to solve is trying to do CSS at scale, trying to do CSS when you’ve got a large team. Which is interesting to think about like, why wasn’t CSS designed to scale well like this?
Large companies find HTML and CSS frustrating “at scale” because the web is fundamentally an anticapitalist mashup art experiment designed to give consumers all the power.
Okay, it’s funny, but it’s kind of funny ’cause it’s true. If you look at those design principles that Bert Bos came up with, it was very much about empowering the end user, that CSS needed to be accessible. It needed to be something you could learn quickly.
So, you know, thinking about CSS as something that needs to be able to scale to multiple teams of people? That wasn’t really on the cards for CSS back then. It wasn’t a priority.
And maybe that’s why we came up with these methodologies like BEM and OOCSS and SMACSS to try and manage this stuff.
But even these methodologies, now the ideas behind them are finding their way into the standards. Now we’re getting cascade layers and scope in CSS.
To me, this feels like a return of this idea of influence that Håkon Wium Lie was talking about all those years ago.
Now it’s not so much about the influence between an author and a user; it’s about the influence between multiple authors working on, on a giant code base.
So it’s a very exciting time for CSS to see these new tools arrive that can solve these scale problems.
What’s still missing?
And you can answer the question by looking at what are we still using tools for? What are we still having to polyfil because we don’t yet have it natively in the browser?
And to answer this question, I’m gonna just quickly finish with three components that kind of demonstrate where CSS is still missing some features.
Let’s start with a button component.
Really there’s absolutely no reason to do that. In this case, it’s a no brainer. You use a button element and you style it with CSS. I cannot think of any reason why you would not do that. There used to be reasons like ages ago in Internet Explorer it used to be hard to style buttons. Those days are long gone.
So in this case, really simple answer to the question. Material honesty. Use a button element. Use CSS to style it.
All right. What about a dropdown component? There’s a number of options, you click in and you get a select dropdown.
I mean, I personally wouldn’t do it. But I guess I can see it, because CSS isn’t there yet. We don’t yet have the power to style a dropdown.
What about a date picker component? You click into it, the user chooses a date.
Again, there appears to be an obvious solution here, which is use input type="date". Boom. You’re done style it how you want, right?
Yeah, it’s a real shame.
So I think these three components kind of show the battle ground, if you will, where CSS is still falling short a bit, where we have to still use hacks. It’s kind of this battle between the under-engineered solution—just use the native HTML element—and the over-engineered solution, right? We’re gonna have to create all the functionality and all the accessibility by ourselves.
And I used to get mad at people choosing the hacks, choosing the over-engineered solution. But I realized that it’s kind of like, you know, trying to reduce teen pregnancy by telling people to just stop having sex. Abstinence isn’t realistic. People are going to do it anyway. And the question is, well, how do we make it better and safer in the long run?
So I think that’s the real battleground, is how do we style elements like select and date pickers natively in CSS? And that’s why I think the work being done by the open UI group is really, really important.
The purpose of open UI to the web platform is to allow web developers to style and extend built in web controls, such as select dropdowns, check boxes, radio buttons, and date and color pickers.
I think it’s really important work. And I think that’s where we should be putting our effort.
Web browsers are the shared context between users and authors.
Whereas, if you want to use a framework or a library, you have to ship that context to the end user. And that puts a burden on them. It’s not good for performance. It’s not good for the user experience.
So web browsers are where past agreements live on today and they live on into the future.
When something lands in a browser, it stays in a browser. So by using what’s in web browsers, you are benefiting from decades of work by multitudes of people. And it’s better for users.
They can point the way to a shared context in the future, but they themselves are not the future. So don’t get too attached. Treat them as cattle, not pets.
Use frameworks and libraries as scaffolding to help you build. But they are not a foundation.
Web standards in the browser are your foundation to build upon.
You know, having an awareness of the history of technologies from sun dials to web browsers, it can help you understand the way things are today. And in some ways the lessons of path dependence and inertia are sort of grim, right? Because of some arbitrary decision in the past, we are now stuck with the consequences in our clockfaces, in CSS. And it’s very, very hard to change that.
But there’s another way to look at this.
Nothing was inevitable. Which means nothing is inevitable (you know, except for entropy and the heat death of the universe).
So if someone tells you, “Hey, that’s just the way things are; accept it”, don’t believe it.
Understand your position in the timeline.
Yes, the present moment is the result of decisions made in the past, many of them arbitrary, but that also means the future will be highly influenced by the decisions you make today even if those decisions seem small and inconsequential.
The choices you make now could turn out to have long-lasting repercussions into the future.
So make your decisions wisely. You are literally creating the future.
There’s a lot of emphasis put on decision-making: making sure you’re making the right decision; evaluating all the right factors before making a decision. But we rarely talk about revisiting decisions.
I think perhaps there’s a human tendency to treat past decisions as fixed. That’s certainly true when it comes to evaluating technology.
I’ve been guilty of this. I remember once chatting with Mark about something written in PHP—probably something I had written—and I made some remark to the effect of “I know PHP isn’t a great language…” Mark rightly called me on that. The language wasn’t great in the past but it has come on in leaps and bounds. My perception of the language, however, had not updated accordingly.
I try to keep that lesson in mind whenever I’m thinking about languages, tools and frameworks that I’ve investigated in the past but haven’t revisited in a while.
The carousel is like one of those on a game show that shows the prizes that can be won. The tool will sit on there until I think it’s gone through enough maturing to actually be a viable tool for me, the team I’m working with and the clients I’m working for.
Crucially a carousel is circular: tools and technologies come back around for re-evaluation. It’s all too easy to treat technologies as being on a one-way conveyer belt—once they’ve past in front of your eyes and you’ve weighed them up, that’s it; you never return to re-evaluate your decision.
This doesn’t need to be a never-ending process. At some point it becomes clear that some technologies really aren’t worth returning to:
It’s a really useful strategy because some tools stay on the carousel and then I take them off because they did in fact, turn out to be useless after all.
See, for example, anything related to cryptobollocks. It’s been well over a decade and blockchains remain a solution in search of problems. As Molly White put it, it’s not still the early days:
How long can it possibly be “early days”? How long do we need to wait before someone comes up with an actual application of blockchain technologies that isn’t a transparent attempt to retroactively justify a technology that is inefficient in every sense of the word? How much pollution must we justify pumping into our atmosphere while we wait to get out of the “early days” of proof-of-work blockchains?
Back to the web (the actual un-numbered World Wide Web)…
Nolan Lawson wrote an insightful article recently about how he senses that the balance has shifted away from single page apps. I’ve been sensing the same shift in the zeitgeist. That said, both Nolan and I keep an eye on how browsers are evolving and getting better all the time. If you weren’t aware of changes over the past few years, it would be easy to still think that single page apps offer some unique advantages that in fact no longer hold true. As Nolan wrote in a follow-up post:
My main point was: if the only reason you’re using an SPA is because “it makes navigations faster,” then maybe it’s time to re-evaluate that.
Perhaps the best example of a technology that warrants regular re-evaluation is the World Wide Web itself. Over the course of its existence it has been seemingly bettered by other more proprietary technologies.
Flash was better than the web. It had vector graphics, smooth animations, and streaming video when the web had nothing like it. But over time, the web caught up. Flash was the hare. The World Wide Web was the tortoise.
In more recent memory, the role of the hare has been played by native apps.
I remember talking to someone on the Twitter design team who was designing and building for multiple platforms. They were frustrated by the web. It just didn’t feel as fully-featured as iOS or Android. Their frustration was entirely justified …at the time. I wonder if they’ve revisited their judgement since then though.
Just the other day I was chatting with one of my colleagues about an online service that’s available on the web and also as a native app. He was showing me the native app on his phone and said it’s not a great app.
“Why don’t you add the website to your phone?” I asked.
“You know,” he said. “The website’s going to be slow.”
He hadn’t tested this. But years of dealing with crappy websites on his phone in the past had trained him to think of the web as being inherently worse than native apps (even though there was nothing this particular service was doing that required any native functionality).
It has become a truism now. Native apps are better than the web.
And you know what? Once upon a time, that would’ve been true. But it hasn’t been true for quite some time …at least from a technical perspective.
But even if the technologies in browsers have reached parity with native apps, that won’t matter unless we can convince people to revisit their previously-formed beliefs.
The technologies are the easy bit. Getting people to re-evaluate their opinions about technologies? That’s the hard part.
It’s a bit of a cliché to talk about living in the future. It’s also a bit pointless. After all, any moment after the big bang is a future when viewed from any point in time before it.
Still, it’s kind of fun when a sci-fi date rolls around. Like in 2015 when we reached the time depicted in Back To The Future 2, or in 2019 when we reached the time of Blade Runner.
In 2022 we are living in the future of web standards. Again, technically, we’re always living in the future of any past discussion of web standards, but this year is significant …in a very insignificant way.
It all goes back to 2008 and an interview with Hixie, editor of the HTML5 spec at the WHATWG at the time. In it, he mentioned the date 2022 as the milestone for having two completely interoperable implementations.
The far more important—and ambitious—date was 2012, when HTML5 was supposed to become a Candidate Recommendation, which is standards-speak for done’n’dusted.
But the mere mention of the year 2022 back in the year 2008 was too much for some people. Jeff Croft, for example, completely lost his shit (Jeff had a habit of posting angry rants and then denying that he was angry or ranty, but merely having a bit of fun).
God knows where I’ll be in 13 years. Quite frankly, I’ll be pretty fucking disappointed in myself (and our entire industry) if I’m writing HTML in 13 years.
That always struck me as odd. If I thought like that, I’d wonder what the point would be in making anything on the web to begin with (bear in mind that both my own personal website and The Session are now entering their third decade of life).
If you’re thinking that planning how the web will look and work 13 years from now is a little bit ridiculous, you’re not alone.
Even if your 2022 ronc-o-matic web-enabled toaster (It slices! It dices! It browses! It arouses!) does ship with Firefox v22.3, will HTML still be the dominant language of web? Given that no one can really answer that question, does it make sense to propose a standard so far in the future?
(I’m re-reading that article in the current version of Firefox: 95.0.2.)
Two-thousand-twenty-two. That’s 14 years from now. Can any of us think that far? Wouldn’t our robot overlords, whether you welcome them or not, have taken over by then? Will the internet even matter then?
2022: God knows what the Internet will look like at that point. Will we even have websites?
Dan Rubin, who has indeed successfully moved from web work to photography, wrote:
I certainly don’t intend to be doing “web work” by that time. I’m very curious to see where the web actually is in 14 years, though I can’t imagine that HTML5 will even get that far; it’ll all be obsolete before 2022.
Joshua Works made a prediction that’s worryingly close to reality:
I’ll be surprised if website-as-HTML is still the preferred method for moving around the tons of data we create, especially in the manner that could have been predicted in 2003 or even today. Hell, iPods will be over 20 years old by then and if everything’s not run as an iPhone App, then something went wrong.
The world in 2022 will be pretty much like the world in 2009.
The world in 2009 is pretty much like 1996 which was pretty much like the world in 1983 which was pretty much like the world in 1970. Some changes are fairly sudden, others are slow, some are dramatic, others subtle, but as a whole “pretty much the same” covers it.
The Web in 2022 will not be dramatically different from the Web in 2009. It will be less hot and it will be less cool. The Web is a project, and as it succeeds it will fade out of our attention and into the background. We don’t care about things when they work.
Now that’s a sensible perspective!
So who else is looking forward to seeing what the World Wide Web is like in 2036?
I must remember to write a blog post then and link back to this one. I have no intention of trying to predict the future, but I’m willing to bet that hyperlinks will still be around in 14 years.
I’m glad I did. While it’s probably of little interest to anyone else, I enjoy scrolling back to see how the same date unfolded over the years.
’Sfunny, when I look back at older journal entries they’re often written out of frustration, usually when something in the dev world is bugging me. But when I look back at all the links I’ve bookmarked the vibe is much more enthusiastic, like I’m excitedly pointing at something and saying “Check this out!” I feel like sentiment analyses of those two sections of my site would yield two different results.
But when I scroll down through my “on this day” page, it also feels like descending deeper into the dark waters of linkrot. For each year back in time, the probability of a link still working decreases until there’s nothing but decay.
Too much has been lost already. The glue that holds humanity’s knowledge together is coming undone.
In one sense, linkrot is the price we pay for the web’s particular system of hypertext. We don’t have two-way linking, which means there’s no centralised repository of links which would be prohibitively complex to maintain. So when you want to link to something on the web, you just do it. An a element with an href attribute. That’s it. You don’t need to check with the owner of the resource you’re linking to. You don’t need to check with anyone. You have complete freedom to link to any URL you want to.
But it’s that same simple system that makes the act of linking a gamble. If the URL you’ve linked to goes away, you’ll have no way of knowing.
As I scroll down my “on this day” page, I come across more and more dead links that have been snapped off from the fabric of the web.
If I stop and think about it, it can get quite dispiriting. Why bother making hyperlinks at all? It’s only a matter of time until those links break.
And yet I still keep linking. I still keep pointing to things and saying “Check this out!” even though I know that over a long enough timescale, there’s little chance that the link will hold.
In a sense, every hyperlink on the World Wide Web is little act of hope. Even though I know that when I link to something, it probably won’t last, I still harbour that hope.
If hyperlinks are built on hope, and the web is made of hyperlinks, then in a way, the World Wide Web is quite literally made out of hope.
Thirty years later, it is easy to overlook the web’s origins as a tool for sharing knowledge. Key to Tim Berners-Lee’s vision were open standards that reflected his belief in the Rule of Least Power, a principle that choosing the simplest and least powerful language for a given purpose allows you to do more with the data stored in that language (thus, HTML is easier for humans or machines to interpret and analyze than PostScript). Along with open standards and the Rule of Least Power, Tim Berners-Lee wanted to make it easy for anyone to publish information in the form of web pages. His first web browser, named Nexus, was both a browser and editor.
It was very hot here in England last week. By late afternoon, the stuffiness indoors was too much to take.
If you can’t stand the heat, get out of the kitchen. That’s exactly what Jessica and I did. The time had come for us to avail of someone else’s kitchen. For the first time in many months, we ventured out for an evening meal. We could take advantage of the government discount scheme with the very unfortunate slogan, “eat out to help out.” (I can’t believe that no one in that meeting said something.)
Just to be clear, we wanted to dine outdoors. The numbers are looking good in Brighton right now, but we’re both still very cautious about venturing into indoor spaces, given everything we know now about COVID-19 transmission.
Fortunately for us, there’s a new spot on the seafront called Shelter Hall Raw. It’s a collective of multiple local food outlets and it has ample outdoor seating.
We found a nice table for two outside. Then we didn’t flag down a waiter.
Instead, we followed the instructions on the table. I say instructions, but it was a bit simpler than that. It was a URL: shelterhall.co.uk (there was also a QR code next to the URL that I could’ve just pointed my camera at, but I’ve developed such a case of QR code blindness that I blanked that out initially).
Just to be clear, under the current circumstances, this is the only way to place an order at this establishment. The only (brief) interaction you’ll have with another persn is when someone brings your order.
It worked a treat.
We had frosty beverages chosen from the excellent selection of local beers. We also had fried chicken sandwiches from Lost Boys chicken, purveyors of the best wings in town.
The whole experience was a testament to what the web can do. You browse the website. You make your choice on the website. You pay on the website (you can create an account but you don’t have to).
Thinking about it, I can see why they chose the web over a native app. Online ordering is the only way to place your order at this place. Telling people “You have to go to this website” …that seems reasonable. But telling people “You have to download this app” …that’s too much friction.
It hasn’t been a great week for the web. Layoffs at Mozilla. Google taking aim at URLs. It felt good to see experience an instance of the web really shining.
The timeline of Gopher and HTTP can be evidenced by their default port numbers. Gopher is 70, HTTP 80. The HTTP port was assigned (likely by Jon Postel at the IANA) at the request of Tim Berners-Lee sometime between 1990 and 1992.
Kimberly was spelunking down the original source code, when she came across this line in the HTUtils.h file:
#define TCP_PORT 80 /* Allocated to http by Jon Postel/ISI 24-Jan-92 */
We showed this to Jean-François Groff, who worked on the original web technologies like libwww, the forerunner to libcurl. He remembers that day. It felt like they had “made it”, receiving the official blessing of Jon Postel (in the same RFC, incidentally, that gave port 70 to Gopher).
Then he told us something interesting about the next line of code:
#define OLD_TCP_PORT 2784 /* Try the old one if no answer on 80 */
Port 2784? That seems like an odd choice. Most of us would choose something easy to remember.
Well, it turns out that 2784 is easy to remember if you’re Tim Berners-Lee.
Those were the last four digits of his parents’ phone number.