Tags: form



Marking up help text in forms

Zoe asked a question on Twitter recently:

‘Sfunny—I had been pondering this exact question. In fact, I threw a CodePen together a couple of weeks ago.

Visually, both examples look the same; there’s a label, then a form field, then some extra text (in this case, a validation message).

The first example puts the validation message in an em element inside the label text itself, so I know it won’t be missed by a screen reader—I think I first learned this technique from Derek many years ago.

<div class="first error example">
 <label for="firstemail">Email
<em class="message">must include the @ symbol</em>
 <input type="email" id="firstemail" placeholder="e.g. you@example.com">

The second example puts the validation message after the form field, but uses aria-describedby to explicitly associate that message with the form field—this means the message should be read after the form field.

<div class="second error example">
 <label for="secondemail">Email</label>
 <input type="email" id="secondemail" placeholder="e.g. you@example.com" aria-describedby="seconderror">
 <em class="message" id="seconderror">must include the @ symbol</em>

In both cases, the validation message won’t be missed by screen readers, although there’s a slight difference in the order in which things get read out. In the first example we get:

  1. Label text,
  2. Validation message,
  3. Form field.

And in the second example we get:

  1. Label text,
  2. Form field,
  3. Validation message.

In this particular example, the ordering in the second example more closely matches the visual representation, although I’m not sure how much of a factor that should be in choosing between the options.

Anyway, I was wondering whether one of these two options is “better” or “worse” than the other. I suspect that there isn’t a hard and fast answer.

Unlabelled search fields

Adam Silver is writing a book on forms—you may be familiar with his previous book on maintainable CSS. In a recent article (that for some reason isn’t on his blog), he looks at markup patterns for search forms and advocates that we should always use a label. I agree. But for some reason, we keep getting handed designs that show unlabelled search forms. And no, a placeholder is not a label.

I had a discussion with Mark about this the other day. The form he was marking up didn’t have a label, but it did have a button with some text that would work as a label:

<input type="search" placeholder="…">
<button type="submit">

He was wondering if there was a way of using the button’s text as the label. I think there is. Using aria-labelledby like this, the button’s text should be read out before the input field:

<input aria-labelledby="searchtext" type="search" placeholder="…">
<button type="submit" id="searchtext">

Notice that I say “think” and “should.” It’s one thing to figure out a theoretical solution, but only testing will show whether it actually works.

The W3C’s WAI tutorial on labelling content gives an example that uses aria-label instead:

<input type="text" name="search" aria-label="Search">
<button type="submit">Search</button>

It seems a bit of a shame to me that the label text is duplicated in the button and in the aria-label attribute (and being squirrelled away in an attribute, it runs the risk of metacrap rot). But they know what they’re talking about so there may well be very good reasons to prefer duplicating the value with aria-label rather than pointing to the value with aria-labelledby.

I thought it would be interesting to see how other sites are approaching this pattern—unlabelled search forms are all too common. All the markup examples here have been simplified a bit, removing class attributes and the like…

The BBC’s search form does actually have a label:

<label for="orb-search-q">
Search the BBC
<input id="orb-search-q" placeholder="Search" type="text">
<button>Search the BBC</button>

But that label is then hidden using CSS:

position: absolute;
height: 1px;
width: 1px;
overflow: hidden;
clip: rect(1px, 1px, 1px, 1px);

That CSS—as pioneered by Snook—ensures that the label is visually hidden but remains accessible to assistive technology. Using something like display: none would hide the label for everyone.

Medium wraps the input (and icon) in a label and then gives the label a title attribute. Like aria-label, a title attribute should be read out by screen readers, but it has the added advantage of also being visible as a tooltip on hover:

<label title="Search Medium">
  <span class="svgIcon"><svg></svg></span>
  <input type="search">

This is also what Google does on what must be the most visited search form on the web. But the W3C’s WAI tutorial warns against using the title attribute like this:

This approach is generally less reliable and not recommended because some screen readers and assistive technologies do not interpret the title attribute as a replacement for the label element, possibly because the title attribute is often used to provide non-essential information.

Twitter follows the BBC’s pattern of having a label but visually hiding it. They also have some descriptive text for the icon, and that text gets visually hidden too:

<label class="visuallyhidden" for="search-query">Search query</label>
<input id="search-query" placeholder="Search Twitter" type="text">
<span class="search-icon>
  <button type="submit" class="Icon" tabindex="-1">
    <span class="visuallyhidden">Search Twitter</span>

Here’s their CSS for hiding those bits of text—it’s very similar to the BBC’s:

.visuallyhidden {
  border: 0;
  clip: rect(0 0 0 0);
  height: 1px;
  margin: -1px;
  overflow: hidden;
  padding: 0;
  position: absolute;
  width: 1px;

That’s exactly the CSS recommended in the W3C’s WAI tutorial.

Flickr have gone with the aria-label pattern as recommended in that W3C WAI tutorial:

<input placeholder="Photos, people, or groups" aria-label="Search" type="text">
<input type="submit" value="Search">

Interestingly, neither Twitter or Flickr are using type="search" on the input elements. I’m guessing this is probably because of frustrations with trying to undo the default styles that some browsers apply to input type="search" fields. Seems a shame though.

Instagram also doesn’t use type="search" and makes no attempt to expose any kind of accessible label:

<input type="text" placeholder="Search">
<span class="coreSpriteSearchIcon"></span>

Same with Tumblr:

<input tabindex="1" type="text" name="q" id="search_query" placeholder="Search Tumblr" autocomplete="off" required="required">

…although the search form itself does have role="search" applied to it. Perhaps that helps to mitigate the lack of a clear label?

After that whistle-stop tour of a few of the web’s unlabelled search forms, it looks like the options are:

  • a visually-hidden label element,
  • an aria-label attribute,
  • a title attribute, or
  • associate some text using aria-labelledby.

But that last one needs some testing.

Update: Emil did some testing. Looks like all screen-reader/browser combinations will read the associated text.

A little progress

I’ve got a fairly simple posting interface for my notes. A small textarea, an optional file upload, some checkboxes for syndicating to Twitter and Flickr, and a submit button.

Notes posting interface

It works fine although sometimes the experience of uploading a file isn’t great, especially if I’m on a slow connection out and about. I’ve been meaning to add some kind of Ajax-y progress type thingy for the file upload, but never quite got around to it. To be honest, I thought it would be a pain.

But then, in his excellent State Of The Gap hit parade of web technologies, Remy included a simple file upload demo. Turns out that all the goodies that have been added to XMLHttpRequest have made this kind of thing pretty easy (and I’m guessing it’ll be easier still once we have fetch).

I’ve made a little script that adds a progress bar to any forms that are POSTing data.

Feel free to use it, adapt it, and improve it. It isn’t using any ES6iness so there are some obvious candidates for improvement there.

It’s working a treat on my little posting interface. Now I can stare at a slowly-growing progress bar when I’m out and about on a slow connection.

Enhance! Conf!

Two weeks from now there will be an event in London. You should go to it. It’s called EnhanceConf:

EnhanceConf is a one day, single track conference covering the state of the art in progressive enhancement. We will look at the tools and techniques that allow you to extend the reach of your website/application without incurring additional costs.

As you can probably guess, this is right up my alley. Wild horses wouldn’t keep me away from it. I’ve been asked to be Master of Ceremonies for the day, which is a great honour. Luckily I have some experience in that department from three years of hosting Responsive Day Out. In fact, EnhanceConf is going to run very much in the mold of Responsive Day Out, as organiser Simon explained in an interview with Aaron.

But the reason to attend is of course the content. Check out that line-up! Now that is going to be a knowledge-packed day: design, development, accessibility, performance …these are a few of my favourite things. Nat Buckley, Jen Simmons, Phil Hawksworth, Anna Debenham, Aaron Gustafson …these are a few of my favourite people.

Tickets are still available. Use the discount code JEREMYK to get a whopping 15% off the ticket price.

There’s also a scholarship:

The scholarships are available to anyone not normally able to attend a conference.

I’m really looking forward to EnhanceConf. See you at RSA House on March 4th!

Pseudo and pseudon’t

I like CSS pseudo-classes. They come in handy for adding little enhancements to interfaces based on interaction.

Take the form-related pseudo-classes, for example: :valid, :invalid, :required, :in-range, and many more.

Let’s say I want to adjust the appearance of an element based on whether it has been filled in correctly. I might have an input element like this:

<input type="email" required>

Then I can write some CSS to put green border on it once it meets the minimum requirements for validity:

input:valid {
  border: 1px solid green;

That works, but somewhat annoyingly, the appearance will change while the user is still typing in the field (as soon as the user types an @ symbol, the border goes green). That can be distracting, or downright annoying.

I only want to display the green border when the input is valid and the field is not focused. Luckily for me, those last two words (“not focused”) map nicely to some more pseudo-classes: not and focus:

input:not(:focus):valid {
  border: 1px solid green;

If I want to get really fancy, I could display an icon next to form fields that have been filled in. But to do that, I’d need more than a pseudo-class; I’d need a pseudo-element, like :after

input:not(:focus):valid::after {
  content: '✓';

…except that won’t work. It turns out that you can’t add generated content to replaced elements like form fields. I’d have to add a regular element into my markup, like this:

<input type="email" required>

So I could style it with:

input:not(:focus):valid + span::after {
  content: '✓';

But that feels icky.

Update: See this clever flexbox technique by Hugo Giraudel for a potential solution.

AMPed up

Apple has Apple News. Facebook has Instant Articles. Now Google has AMP: Accelerated Mobile Pages.

The big players sure are going to a lot of effort to reinvent RSS.

That may sound like a flippant remark, but it’s not too far from the truth. In the case of Apple News, its current incarnation appears to be quite literally an RSS reader, at least until the unveiling of the forthcoming Apple News Format.

Google’s AMP project looks a little bit different to the offerings from Facebook and Apple. Rather than creating a proprietary format from scratch, it mandates a subset of HTML …with some proprietary elements thrown in (or, to use the more diplomatic parlance of the extensible web, custom elements).

The idea is that alongside the regular HTML version of your document, you provide a corresponding AMP HTML version. Because the AMP HTML version will be leaner and meaner, user agents can then grab the AMP HTML version and present that to the end user for a faster browsing experience.

So if an RSS feed is an alternate representation of a homepage or a listing of articles, then an AMP document is an alternate representation of a single article.

Now, my own personal take on providing alternate representations of documents is “Sure. Why not?” Here on adactio.com I provide RSS feeds. On The Session I provide RSS, JSON, and XML. And on Huffduffer I provide RSS, Atom, JSON, and XSPF, adding:

If you would like to see another format supported, share your idea.

Also, each individual item on Huffduffer has a corresponding oEmbed version (and, in theory, an RDF version)—an alternate representation of that item …in principle, not that different from AMP. The big difference with AMP is that it’s using HTML (of sorts) for its format.

All of this sounds pretty reasonable: provide an alternate representation of your canonical HTML pages so that user-agents (Twitter, Google, browsers) can render a faster-loading version …much like an RSS reader.

So should you start providing AMP versions of your pages? My initial reaction is “Sure. Why not?”

The AMP Project website comes with a list of frequently asked questions, which of course, nobody has asked. My own list of invented frequently asked questions might look a little different.

Will this kill advertising?

We live in hope.

Alas, AMP pages will still be able to carry advertising, but in a restricted form. No more scripts that track your movement across the web …unless the script is from an authorised provider, like say, Google.

But it looks like the worst performance offenders won’t be able to get their grubby little scripts into AMP pages. This is a good thing.

Won’t this kill journalism?

Of all the horrid myths currently in circulation, the two that piss me off the most are:

  1. Journalism requires advertising to survive.
  2. Advertising requires invasive JavaScript.

Put the two together and you get the gist of most of the chicken-littling articles currently in circulation: “Journalism requires invasive JavaScript to survive.”

I could argue against the first claim, but let’s leave that for another day. Let’s suppose for now that, sure, journalism requires advertising to survive. Fine.

It’s that second point that is fundamentally wrong. The idea that the current state of advertising is the only way of advertising is incredibly short-sighted and misguided. Invasive JavaScript is not a requirement for showing me an ad. Setting a cookie is not a requirement for showing me an ad. Knowing where I live, who my friends are, what my income level is, and where I’ve been on the web …none of these are requirements for showing me an ad.

It is entirely possible to advertise to me and treat me with respect at the same time. The Deck already does this.

And you know what? Ad networks had their chance. They had their chance to treat us with respect with the Do Not Track initiative. We asked them to respect our wishes. They told us get screwed.

Now those same ad providers are crying because we’re installing ad blockers. They can get screwed.


It is entirely possible to advertise within AMP pages …just not using blocking JavaScript.

For a nicely nuanced take on what AMP could mean for journalism, see Joshua Benton’s article on Nieman Lab—Get AMP’d: Here’s what publishers need to know about Google’s new plan to speed up your website.

Why not just make faster web pages?

Excellent question!

For a site like adactio.com, the difference between the regular HTML version of an article and the corresponding AMP version of the same article is pretty small. It’s a shame that I can’t just say “Hey, the current version of the article is the AMP version”, but that would require that I only use a subset of HTML and that I add some required guff to my page (including an unnecessary JavaScript file).

But for most of the news sites out there, the difference between their regular HTML pages and the corresponding AMP versions will be pretty significant. That’s because the regular HTML versions are bloated with third-party scripts, oversized assets, and cruft around the actual content.

Now it is in theory possible for these news sites to get rid of all those things, and I sincerely hope that they will. But that’s a big political struggle. I am rooting for developers—like the good folks at VOX—who have to battle against bosses who honestly think that journalism requires invasive JavaScript. Best of luck.

Along comes Google saying “If you want to play in our sandbox, you’re going to have to abide by our rules.” Those rules include performance best practices (for the most part—I take issue with some of the requirements, and I’ll go into that in more detail in a moment).

Now when the boss says “Slap a three megabyte JavaScript library on it so we can show a carousel”, the developers can only respond with “Google says No.”

When the boss says “Slap a ton of third-party trackers on it so we can monetise those eyeballs”, the developers can only respond with “Google says No.”

Google have used their influence like this before and it has brought them accusations of monopolistic abuse. Some people got very upset when they began labelling (and later ranking) mobile-friendly pages. Personally, I’ve got no issue with that.

In this particular case, Google aren’t mandating what you can and can’t do on your regular HTML pages; only what you can and can’t do on the corresponding AMP page.

Which brings up another question…

Will the AMP web kill the open web?

If we all start creating AMP versions of our pages, and those pages are faster than our regular HTML versions, won’t everyone just see the AMP versions without ever seeing the “full” versions?

Tim articulates a legitimate concern:

This promise of improved distribution for pages using AMP HTML shifts the incentive. AMP isn’t encouraging better performance on the web; AMP is encouraging the use of their specific tool to build a version of a web page. It doesn’t feel like something helping the open web so much as it feels like something bringing a little bit of the walled garden mentality of native development onto the web.

That troubles me. Using a very specific tool to build a tailored version of my page in order to “reach everyone” doesn’t fit any definition of the “open web” that I’ve ever heard.

Fair point. But I also remember that a lot of people were upset by RSS. They didn’t like that users could go for months at a time without visiting the actual website, and yet they were reading every article. They were reading every article in non-browser user agents in a format that wasn’t HTML. On paper that sounds like the antithesis of the open web, but in practice there was always something very webby about RSS, and RSS feed readers—it put the power back in the hands of the end users.

Some people chose not to play ball. They only put snippets in their RSS feeds, not the full articles. Maybe some publishers will do the same with the AMP versions of their articles: “To read more, click here…”

But I remember what generally tended to happen to the publishers who refused to put the full content in their RSS feeds. We unsubscribed.

Still, I share the concern that any one company—whether it’s Facebook, Apple, or Google—should wield so much power over how we publish on the web. I don’t think you have to be a conspiracy theorist to view the AMP project as an attempt to replace the existing web with an alternate web, more tightly controlled by Google (albeit a faster, more performant, tightly-controlled web).

My hope is that the current will flow in both directions. As well as publishers creating AMP versions of their pages in order to appease Google, perhaps they will start to ask “Why can’t our regular pages be this fast?” By showing that there is life beyond big bloated invasive web pages, perhaps the AMP project will work as a demo of what the whole web could be.

I’ve been playing around with the AMP HTML spec. It has some issues. The good news is that it’s open source and the project owners seem receptive to feedback.


No external JavaScript is allowed in an AMP HTML document. This covers third-party libraries, advertising and tracking scripts. This is A-okay with me.

The reasons given for this ban are related to performance and I agree with them completely. Big bloated JavaScript libraries are one of the biggest performance killers on the web. I’m happy to leave them at the door (although weirdly, web fonts—another big performance killer—are allowed in).

But then there’s a bit of an about-face. In order to have a valid AMP HTML page, you must include a piece of third-party JavaScript. In this case, the third party is Google and the JavaScript file is what handles the loading of assets.

This seems a bit strange to me; on the one hand claiming that third-party JavaScript is bad for performance and on the other, requiring some third-party JavaScript. As Justin says:

For me this is loading one thing too many… the AMP JS library. Surely the document itself is going to be faster than loading a library to try and make it load faster.

On the plus side, this third-party JavaScript is loaded asynchronously. It seems to mostly be there to handle the rendering of embedded content: images, videos, audio, etc.

Embedded content

If you want audio, video, or images on your page, you must use propriet… custom elements like amp-audio, amp-video, and amp-img. In the case of images, I can see how this is a way of getting around the browser’s lookahead pre-parser (although responsive images also solve this problem). In the case of audio and video, the standard audio and video elements already come with a way of specifying preloading behaviour using the preload attribute. Very odd.

Justin again:

I’m not sure if this is solving anything at the moment that we’re not already fixing with something like responsive images.

To use amp-img for images within the flow of a document, you’ll need to specify the dimensions of the image. This makes sense from a rendering point of view—knowing the width and height ahead of time avoids repaints and reflows. Alas, in many of the cases here on adactio.com, I don’t know the dimensions of the images I’m including. So any of my AMP HTML pages that include images will be invalid.

Overall, the way that AMP HTML handles embedded content looks like a whole lot of wheel reinvention. I like the idea of providing custom elements as an option for authors. I hate the idea of making them a requirement.


If you want to provide metadata about your document, AMP HTML currently requires the use of Google’s Schema.org vocabulary. This has a big whiff of vendor lock-in to it. I’ve flagged this up as an issue and Aaron is pushing a change so hopefully this will be resolved soon.


In its initial release, the AMP HTML spec came with some nasty surprises for accessibility. The biggest is probably the requirement to include this in your viewport meta element:


Yowzers! That’s some slap in the face to decent web developers everywhere. Fortunately this has been flagged up and I’m hoping it will be fixed soon.

If it doesn’t get fixed, it’s quite a non-starter. It beggars belief that Google would mandate to authors that they must make their pages inaccessible to pinch/zoom. I would hope that many developers would rebel against such a draconian injunction. If that happens, it’ll be interesting to see what becomes of those theoretically badly-formed AMP HTML documents. Technically, they will fail validation, but for very good reason. Will those accessible documents be rejected?

Please get involved on this issue if this is important to you (hint: this should be important to you).

There are a few smaller issues. Initially the :focus pseudo-class was disallowed in author CSS, but that’s being fixed.

Currently AMP HTML documents must have this line:

<style>body {opacity: 0}</style><noscript><style>body {opacity: 1}</style></noscript>


That’s a horrible conflation of JavaScript availability and CSS. It’s being fixed though, and soon all the opacity jiggery-pokery will only happen via JavaScript, which will be a big improvement: it should either all happen in CSS or all happen in JavaScript, but not the current mixture of the two.


The AMP HTML version of your page is not the canonical version. You can specify where the real HTML version of your document is by using rel="canonical". Great!

But how do you link from your canonical page out to the AMP HTML version? Currently you’re supposed to use rel="amphtml". No, they haven’t checked the registry. Again. I’ll go in and add it.

In the meantime, I’m also requesting that the amphtml value can be combined with the alternate value, seeing as rel values can be space separated:

rel="alternate amphtml" type="text/html"

See? Not that different to RSS:

rel="alterate" type="application/rss+xml"


When I publish something on adactio.com in HTML, it already gets syndicated to different places. This is the Indie Web idea of POSSE: Publish (on your) Own Site, Syndicate Elsewhere. As well as providing RSS feeds, I’ve also got Twitter bots that syndicate to Twitter. An If This, Then That script pushes posts to Facebook. And if I publish a photo, it goes to Flickr. Now that Medium is finally providing a publishing API, I’ll probably start syndicating articles there as well. The more, the merrier.

From that perspective, providing AMP HTML pages feels like just one more syndication option. If it were the only option, and I felt compelled to provide AMP versions of my content, I’d be very concerned. But for now, I’ll give it a whirl and see how it goes.

Here’s a bit of PHP I’m using to convert a regular piece of HTML into AMP HTML—it’s horrible code; it uses regular expressions on HTML which, as we all know, will summon the Elder Gods.

Building the dConstruct 2015 site

I remember when I first saw Paddy’s illustration for this year’s dConstruct site, I thought “Well, that’s a design direction, but there’s no way that Graham will be able to implement all of it.” There was a tight deadline for getting the site out, and let’s face it, there was so much going on in the design that we’d just have to prioritise.

I underestimated Graham’s sheer bloody-mindedness.

At the next front-end pow-wow at Clearleft, Graham showed the dConstruct site in all its glory …in Lynx.

http://2015.dconstruct.org in Lynx.

I love that. Even with the focus on the gorgeous illustration and futuristic atmosphere of the design, Graham took the time to think about the absolute basics: marking up the content in a logical structured way. Everything after that—the imagery, the fonts, the skewed style—all of it was built on a solid foundation.

One site, two browsers.

It would’ve been easy to go crazy with the fonts and images, but Graham made sure to optimise everything to within an inch of its life. The biggest bottleneck comes from a third party provider—the map tiles and associated JavaScript …so that’s loaded in after the initial content is loaded. It turns out that the site build was a matter of prioritisation after all.


There’s plenty of CSS trickery going on: transforms, transitions, and opacity. But for the icing on the cake, Graham reached for canvas and programmed space elevator traffic with randomly seeded velocity and size.

Oh, and of course it’s all responsive.

So, putting that all together…

The dConstruct 2015 site is gorgeous, semantic, responsive, and performant. Conventional wisdom dictates that you have to choose, but this little site—built on a really tight schedule—shows otherwise.

On The Verge

Quite a few people have been linking to an article on The Verge with the inflammatory title The Mobile web sucks. In it, Nilay Patel heaps blame upon mobile browsers, Safari in particular:

But man, the web browsers on phones are terrible. They are an abomination of bad user experience, poor performance, and overall disdain for the open web that kicked off the modern tech revolution.

Les Orchard says what we’re all thinking in his detailed response The Verge’s web sucks:

Calling out browser makers for the performance of sites like his? That’s a bit much.

Nilay does acknowledge that the Verge could do better:

Now, I happen to work at a media company, and I happen to run a website that can be bloated and slow. Some of this is our fault: The Verge is ultra-complicated, we have huge images, and we serve ads from our own direct sales and a variety of programmatic networks.

But still, it sounds like the buck is being passed along. The performance issues are being treated as Somebody Else’s Problem …ad networks, trackers, etc.

The developers at Vox Media take a different, and in my opinion, more correct view. They’re declaring performance bankruptcy:

I mean, let’s cut to the chase here… our sites are friggin’ slow, okay!

But I worry about how they can possibly reconcile their desire for a faster website with a culture that accepts enormously bloated ads and trackers as the inevitable price of doing business on the web:

I’m hearing an awful lot of false dichotomies here: either you can have a performant website or you have a business model based on advertising. Here’s another false dichotomy:

If the message coming down from above is that performance concerns and business concerns are fundamentally at odds, then I just don’t know how the developers are ever going to create a culture of performance (which is a real shame, because they sound like a great bunch). It’s a particularly bizarre false dichotomy to be foisting when you consider that all the evidence points to performance as being a key differentiator when it comes to making moolah.

It’s funny, but I take almost the opposite view that Nilay puts forth in his original article. Instead of thinking “Oh, why won’t these awful browsers improve to be better at delivering our websites?”, I tend to think “Oh, why won’t these awful websites improve to be better at taking advantage of our browsers?” After all, it doesn’t seem like that long ago that web browsers on mobile really were awful; incapable of rendering the “real” web, instead only able to deal with WAP.

As Maciej says in his magnificent presentation Web Design: The First 100 Years:

As soon as a system shows signs of performance, developers will add enough abstraction to make it borderline unusable. Software forever remains at the limits of what people will put up with. Developers and designers together create overweight systems in hopes that the hardware will catch up in time and cover their mistakes.

We complained for years that browsers couldn’t do layout and javascript consistently. As soon as that got fixed, we got busy writing libraries that reimplemented the browser within itself, only slower.

I fear that if Nilay got his wish and mobile browsers made a quantum leap in performance tomorrow, the result would be even more bloated JavaScript for even more ads and trackers on websites like The Verge.

If anything, browser makers might have to take more drastic steps to route around the damage of bloated websites with invasive tracking.

We’ve been here before. When JavaScript first landed in web browsers, it was quickly adopted for three primary use cases:

  1. swapping out images when the user moused over a link,
  2. doing really bad client-side form validation, and
  3. spawning pop-up windows.

The first use case was so popular, it was moved from a procedural language (JavaScript) to a declarative language (CSS). The second use case is still with us today. The third use case was solved by browsers. They added a preference to block unwanted pop-ups.

Tracking and advertising scripts are today’s equivalent of pop-up windows. There are already plenty of tools out there to route around their damage: Ghostery, Adblock Plus, etc., along with tools like Instapaper, Readability, and Pocket.

I’m sure that business owners felt the same way about pop-up ads back in the late ’90s. Just the price of doing business. Shrug shoulders. Just the way things are. Nothing we can do to change that.

For such a young, supposedly-innovative industry, I’m often amazed at what people choose to treat as immovable, unchangeable, carved-in-stone issues. Bloated, invasive ad tracking isn’t a law of nature. It’s a choice. We can choose to change.

Every bloated advertising and tracking script on a website was added by a person. What if that person refused? I guess that person would be fired and another person would be told to add the script. What if that person refused? What if we had a web developer picket line that we collectively refused to cross?

That’s an unrealistic, drastic suggestion. But the way that the web is being destroyed by our collective culpability calls for drastic measures.

By the way, the pop-up ad was first created by Ethan Zuckerman. He has since apologised. What will you be apologising for in decades to come?


When I give talks or workshops, I sometimes get a bit ranty. One of the richest seams of rantiness comes from me complaining about how we web designers and developers are responsible for making the web a hostile place. “Stop getting the web wrong!” I might shout, like an old man yelling at a cloud. I point to services like Instapaper and Readability and describe their existence as a damning indictment of our work.

Don’t get me wrong—I really like Instapaper, Readability, RSS readers, or any other tools that allow people to read what they want when they want it. But think about their fundamental selling point: get to the content you want without having to wade through the cruft. That cruft was put there by us.

So-called modern web design and development is damage that people have to route around.

(Ooh, I can feel myself coming over all ranty and angry again! Calm down, Jeremy, calm down!)

And. Breathe.

Now there’s a new tool to the add to the list: Facebook Instant. Again, I think it’s actually pretty great that this service exists. But once again, it should make us ashamed of the work we’re collectively producing.

In this case, the service is—somewhat ironically—explicitly touting the performance benefits of not going to a website to read an article. Quite right.

PPK points to tools as the source of the problem and Marco Arment agrees:

The entire culture dominant among web developers today is bizarrely framework-heavy, with seemingly no thought given to minimizing dependencies and page weight.

But I think it’s a bit more subtle than that. As John Gruber says:

Business development deals have created problems that no web developer can solve. There’s no way to make a web page with a full-screen content-obscuring ad anything other than a shitty experience.

Now you might be saying to yourself “Well, I’ve never made a bloated web page!” or “I’ve never slapped loads of intrusive crap over the content!” I’d certainly like to think that I can look at my track record and hold my head up reasonably high. But that doesn’t matter. If the overall perception is that going to a URL to read an article is a pain in the ass, it hurts all of us.

Take this article from M.G. Siegler:

Not only is the web not fast enough for apps, it’s not fast enough for text either. …on mobile, the web browser just isn’t cutting it. … Native apps provide a better user experience on mobile than a web browser.

On the face of it, this is kind of a bizarre claim. After all, there’s nothing inherent in web browsers that makes them slow at rendering text—quite the opposite! And native apps still use HTTP (and often HTML) to fetch content; the network doesn’t suddenly get magically faster just because the piece of software requesting a resource doesn’t happen to be a web browser.

But this conflation of slow websites and slow web browsers is perfectly understandable. If it looks like a slow duck, and it quacks like a slow duck, then why not conclude that ducks are slow? Even if we know that there’s nothing inherently slow about making web pages:

My hope is that Facebook Instant will shake things up a bit. M.G. Siegler again:

At the very least, Facebook has put everyone else on notice. Your content better load fast or you’re screwed. Publication websites have become an absolutely bloated mess. They range from beautiful (The Verge) to atrocious (Bloomberg) to unusable (Forbes). The common denominator: they’re all way too slow.

There needs to be a cultural change in how we approach building for the web. Yes, some of the tools we choose are part of the problem, but the bigger problem is that performance still isn’t being recognised as the most important factor in how people feel about websites (and by extension, the web). This isn’t just a developer issue. It’s a design issue. It’s a UX issue. It’s a business issue. Performance is everybody’s collective responsibility.

I’d better stop now before I start getting all ranty again.

I’ll leave you with some other writings on this topic…

Tim Kadlec talks about choosing performance:

It’s not because of any sort of technical limitations. No, if a website is slow it’s because performance was not prioritized. It’s because when push came to shove, time and resources were spent on other features of a site and not on making sure that site loads quickly.

Jim Ray points out that “we learned the wrong lesson from the rise of mobile and the app ecosystem”:

We’ve spent far too long trying to compete with native experiences by making our websites look and behave like apps. This includes not just thousands of lines of JavaScript to mimic native app swipes and scrolling but even the lower overhead aesthetics of fixed position headers and persistent navigation.


Finally, Baldur Bjarnason has written a terrific piece:

The web doesn’t suck. Your websites suck.

All of your websites suck.

You destroy basic usability by hijacking the scrollbar. You take native functionality (scrolling, selection, links, loading) that is fast and efficient and you rewrite it with ‘cutting edge’ javascript toolkits and frameworks so that it is slow and buggy and broken. You balloon your websites with megabytes of cruft. You ignore best practices. You take something that works and is complementary to your business and turn it into a liability.

The lousy performance of your websites becomes a defensive moat around Facebook.

Go read the whole thing—it’s terrific:

This is a long-standing debate. Except it’s only long-standing among web developers. Columnists, managers, pundits, and journalists seem to have no interest in understanding the technical foundation of their livelihoods. Instead they are content with assuming that Facebook can somehow magically render HTML over HTTP faster than anybody else and there is nothing anybody can do to make their crap scroll-jacking websites faster. They buy into the myth that the web is incapable of delivering on its core capabilities: delivering hypertext and images quickly to a diverse and connected readership.

100 words 058

PPK writes of modern web development:

Tools don’t solve problems any more, they have become the problem.

I think he’s mostly correct, but I think there is some clarification required.

Web development tools fall into two broad categories:

  1. Local tools like preprocessors, task managers, and version control systems that help the developer output their own HTML, CSS, and JavaScript.
  2. Tools written in HTML, CSS, and JavaScript that the end user has to download for the developer to gain benefit.

It’s that second category that contain a tax on the end user. Stop solving problems you don’t yet have.

Inlining critical CSS for first-time visits

After listening to Scott rave on about how much of a perceived-performance benefit he got from inlining critical CSS on first load, I thought I’d give it a shot over at The Session. On the chance that this might be useful for others, I figured I’d document what I did.

The idea here is that you can give a massive boost to the perceived performance of the first page load on a site by putting the most important CSS in the head of the page. Then you cache the full stylesheet. For subsequent visits you only ever use the external stylesheet. So if you’re squeamish at the thought of munging your CSS into your HTML (and that’s a perfectly reasonable reaction), don’t worry—this is a temporary workaround just for initial visits.

My particular technology stack here is using Grunt, Apache, and PHP with Twig templates. But I’m sure you can adapt this for other technology stacks: what’s important here isn’t the technology, it’s the thinking behind it. And anyway, the end user never sees any of those technologies: the end user gets HTML, CSS, and JavaScript. As long as that’s what you’re outputting, the specifics of the technology stack really don’t matter.

Generating the critical CSS

Okay. First question: how do you figure out which CSS is critical and which CSS can be deferred?

To help answer that, and automate the task of generating the critical CSS, Filament Group have made a Grunt task called grunt-criticalcss. I added that to my project and updated my Gruntfile accordingly:

    // All my existing Grunt configuration goes here.
    criticalcss: {
        dist: {
            options: {
                url: 'http://thesession.dev',
                width: 1024,
                height: 800,
                filename: '/path/to/main.css',
                outputfile: '/path/to/critical.css'

I’m giving it the name of my locally-hosted version of the site and some parameters to judge which CSS to prioritise. Those parameters are viewport width and height. Now, that’s not a perfect way of judging which CSS matters most, but it’ll do.

Then I add it to the list of Grunt tasks:

// All my existing Grunt tasks go here.

grunt.registerTask('default', ['sass', etc., 'criticalcss']);

The end result is that I’ve got two CSS files: the full stylesheet (called something like main.css) and a stylesheet that only contains the critical styles (called critical.css).

Cache-busting CSS

Okay, this is a bit of a tangent but trust me, it’s going to be relevant…

Most of the time it’s a very good thing that browsers cache external CSS files. But if you’ve made a change to that CSS file, then that feature becomes a bug: you need some way of telling the browser that the CSS file has been updated. The simplest way to do this is to change the name of the file so that the browser sees it as a whole new asset to be cached.

You could use query strings to do this cache-busting but that has some issues. I use a little bit of Apache rewriting to get a similar effect. I point browsers to CSS files like this:

<link rel="stylesheet" href="/css/main.20150310.css">

Now, there isn’t actually a file named main.20150310.css, it’s just called main.css. To tell the server where the actual file is, I use this rewrite rule:

RewriteCond %{REQUEST_FILENAME} !-f
RewriteRule ^(.+).(d+).(js|css)$ $1.$3 [L]

That tells the server to ignore those numbers in JavaScript and CSS file names, but the browser will still interpret it as a new file whenever I update that number. You can do that in a .htaccess file or directly in the Apache configuration.

Right. With that little detour out of the way, let’s get back to the issue of inlining critical CSS.

Differentiating repeat visits

That number that I’m putting into the filenames of my CSS is something I update in my Twig template, like this (although this is really something that a Grunt task could do, I guess):

{% set cssupdate = '20150310' %}

Then I can use it like this:

<link rel="stylesheet" href="/css/main.{{ cssupdate }}.css">

I can also use JavaScript to store that number in a cookie called csscached so I’ll know if the user has a cached version of this revision of the stylesheet:

document.cookie = 'csscached={{ cssupdate }};expires="Tue, 19 Jan 2038 03:14:07 GMT";path=/';

The absence or presence of that cookie is going to be what determines whether the user gets inlined critical CSS (a first-time visitor, or a visitor with an out-of-date cached stylesheet) or whether the user gets a good ol’ fashioned external stylesheet (a repeat visitor with an up-to-date version of the stylesheet in their cache).

Here are the steps I’m going through:

First of all, set the Twig cssupdate variable to the last revision of the CSS:

{% set cssupdate = '20150310' %}

Next, check to see if there’s a cookie called csscached that matches the value of the latest revision. If there is, great! This is a repeat visitor with an up-to-date cache. Give ‘em the external stylesheet:

{% if _cookie.csscached == cssupdate %}
<link rel="stylesheet" href="/css/main.{{ cssupdate }}.css">

If not, then dump the critical CSS straight into the head of the document:

{% else %}
{% include '/css/critical.css' %}

Now I still want to load the full stylesheet but I don’t want it to be a blocking request. I can do this using JavaScript. Once again it’s Filament Group to the rescue with their loadCSS script:

    // include loadCSS here...
    loadCSS('/css/main.{{ cssupdate }}.css');

While I’m at it, I store the value of cssupdate in the csscached cookie:

    document.cookie = 'csscached={{ cssupdate }};expires="Tue, 19 Jan 2038 03:14:07 GMT";path=/';

Finally, consider the possibility that JavaScript isn’t available and link to the full CSS file inside a noscript element:

<link rel="stylesheet" href="/css/main.{{ cssupdate }}.css">
{% endif %}

And we’re done. Phew!

Here’s how it looks all together in my Twig template:

{% set cssupdate = '20150310' %}
{% if _cookie.csscached == cssupdate %}
<link rel="stylesheet" href="/css/main.{{ cssupdate }}.css">
{% else %}
{% include '/css/critical.css' %}
// include loadCSS here...
loadCSS('/css/main.{{ cssupdate }}.css');
document.cookie = 'csscached={{ cssupdate }};expires="Tue, 19 Jan 2038 03:14:07 GMT";path=/';
<link rel="stylesheet" href="/css/main.{{ cssupdate }}.css">
{% endif %}

You can see the production code from The Session in this gist. I’ve tweaked the loadCSS script slightly to match my preferred JavaScript style but otherwise, it’s doing exactly what I’ve outlined here.

The result

According to Google’s PageSpeed Insights, I done good.

Optimising https://thesession.org/

Cerf rocks

After I wrote about digital preservation and the need to save everything, not just the so-called “important” stuff, Jason wrote a lovely piece with his own thoughts on the matter:

In order to write a history, you need evidence of what happened. When we talk about preserving the stuff we make on the web, it isn’t because we think a Facebook status update, or those GeoCities sites have such significance now. It’s because we can’t know.

In a timely coincidence, Vint Cerf also spoke about the importance of digital preservation:

When you think about the quantity of documentation from our daily lives that is captured in digital form, like our interactions by email, people’s tweets, and all of the world wide web, it’s clear that we stand to lose an awful lot of our history.

He warns of the dangers of rapidly-obsoleting file formats:

We are nonchalantly throwing all of our data into what could become an information black hole without realising it. We digitise things because we think we will preserve them, but what we don’t understand is that unless we take other steps, those digital versions may not be any better, and may even be worse, than the artefacts that we digitised.

It was a little weird that the Guardian headline refers to Vint Cerf as “Google boss”. On the BBC he’s labelled as “Google’s Vint Cerf”. Considering he’s one of the creators of the internet itself, it’s a bit like referring to Neil Armstrong as a NASA employee.

I have to say, I just love listening to him talk. He’s so smooth. I’m sure that the character of The Architect from The Matrix Reloaded is modelled on him.

Vint Cerf knows a thing or two about long-term thinking when it comes to data formats. He has written many RFCs for the IETF (my favourite being RFC 2468). Back in 1969, he wrote RFC 20, proposing the ASCII format for network interchange. If you’ve ever used the keypress event in JavaScript and wondered why, for example, the number 13 corresponds to a carriage return, this is where all those numbers come from.

Last month, over 45 years after the RFC’s original publication, it became an official standard.

So when Vint Cerf warns about the dangers of digitising into file formats that could become unreadable, I think we should pay attention to him.

A question of timing

I’ve been updating my collection of design principles lately, adding in some more examples from Android and Windows. Coincidentally, Vasilis unveiled a neat little page that grabs one list of principles at random —just keep refreshing to see more.

I also added this list of seven principles of rich web applications to the collection, although they feel a bit more like engineering principles than design principles per se. That said, they’re really, really good. Every single one is rooted in performance and the user’s experience, not developer convenience.

Don’t get me wrong: developer convenience is very, very important. Nobody wants to feel like they’re doing unnecessary work. But I feel very strongly that the needs of the end user should trump the needs of the developer in almost all instances (you may feel differently and that’s absolutely fine; we’ll agree to differ).

That push and pull between developer convenience and user experience is, I think, most evident in the first principle: server-rendered pages are not optional. Now before you jump to conclusions, the author is not saying that you should never do client-side rendering, but instead points out the very important performance benefits of having the server render the initial page. After that—if the user’s browser cuts the mustard—you can use client-side rendering exclusively.

The issue with that hybrid approach—as I’ve discussed before—is that it’s hard. Isomorphic JavaScript (terrible name) can theoretically help here, but I haven’t seen too many examples of it in action. I suspect that’s because this approach doesn’t yet offer enough developer convenience.

Anyway, I found myself nodding along enthusiastically with that first of seven design principles. Then I got to the second one: act immediately on user input. That sounds eminently sensible, and it’s backed up with sound reasoning. But it finishes with:

Techniques like PJAX or TurboLinks unfortunately largely miss out on the opportunities described in this section.

Ah. See, I’m a big fan of PJAX. It’s essentially the same thing as the Hijax technique I talked about many years ago in Bulletproof Ajax, but with the new addition of HTML5’s History API. It’s a quick’n’dirty way of giving the illusion of a fat client: all the work is actually being done in the server, which sends back chunks of HTML that update the interface. But it’s true that, because of that round-trip to the server, there’s a bit of a delay and so you often end up briefly displaying a loading indicator.

I contend that spinners or “loading indicators” should become a rarity

I agree …but I also like using PJAX/Hijax. Now how do I reconcile what’s best for the user experience with what’s best for my own developer convenience?

I’ve come up with a compromise, and you can see it in action on The Session. There are multiple examples of PJAX in action on that site, like pretty much any page that returns paginated results: new tune settings, the latest events, and so on. The steps for initiating an Ajax request used to be:

  1. Listen for any clicks on the page,
  2. If a “previous” or “next” button is clicked, then:
  3. Display a loading indicator,
  4. Request the new data from the server, and
  5. Update the page with the new data.

In one sense, I am acting immediately to user input, because I always display the loading indicator straight away. But because the loading indicator always appears, no matter how fast or slow the server responds, it sometimes only appears very briefly—just for a flash. In that situation, I wonder if it’s serving any purpose. It might even be doing the opposite to its intended purpose—it draws attention to the fact that there’s a round-trip to the server.

“What if”, I asked myself, “I only showed the loading indicator if the server is taking too long to send a response back?”

The updated flow now looks like this:

  1. Listen for any clicks on the page,
  2. If a “previous” or “next” button is clicked, then:
  3. Start a timer, and
  4. Request the new data from the server.
  5. If the timer reaches an upper limit, show a loading indicator.
  6. When the server sends a response, cancel the timer and
  7. Update the page with the new data.

Even though there are more steps, there’s actually less happening from the user’s perspective. Where previously you would experience this:

  1. I click on a button,
  2. I briefly see a loading indicator,
  3. I see the new data.

Now your experience is:

  1. I click on a button,
  2. I see the new data.

…unless the server or the network is taking too long, in which case the loading indicator appears as an interim step.

The question is: how long is too long? How long do I wait before showing the loading indicator?

The Nielsen Norman group offers this bit of research:

0.1 second is about the limit for having the user feel that the system is reacting instantaneously, meaning that no special feedback is necessary except to display the result.

So I should set my timer to 100 milliseconds. In practice, I found that I can set it to as high as 200 to 250 milliseconds and keep it feeling very close to instantaneous. Anything over that, though, and it’s probably best to display a loading indicator: otherwise the interface starts to feel a little sluggish, and slightly uncanny. (“Did that click do any—? Oh, it did.”)

You can test the response time by looking at some of the simpler pagination examples on The Session: new recordings or new discussions, for example. To see examples of when the server takes a bit longer to send a response, you can try paginating through search results. These take longer because, frankly, I’m not very good at optimising some of those search queries.

There you have it: an interface that—under optimal conditions—reacts to user input instantaneously, but falls back to displaying a loading indicator when conditions are less than ideal. The result is something that feels like a client-side web thang, even though the actual complexity is on the server.

Now to see what else I can learn from the rest of those design principles.

Celebrating CSS

Cascading Style Sheets turned 20 years old this week. Happy birthtime, CeeSusS!

Bruce interviewed Håkon about the creation of CSS, and it makes for fascinating reading. If you want to dig even deeper, here’s Håkon’s 1994 thesis comparing competing approaches to style sheets.

CSS gets a tough rap. I remember talking to Douglas Crockford about CSS. I’ll paraphrase his stance as “Kill it with fire!” To be fair, he was mostly talking about the lack of a decent layout system in CSS—something that’s only really getting remedied now.

Most of the flak directed at CSS comes from smart programmers, decrying its lack of power. As a declarative language, it lacks even the most basic features of even the simplest procedural language. How are serious programmers supposed to write their serious programmes with such a primitive feature set?

But I think this mindset misses out a crucial facet of understanding CSS: it’s not about us. By us, I mean professional web developers. And when I say it’s not about us, I mean it’s not only about us.

The web is for everyone. That doesn’t just mean that it’s for everyone to use—the web is for everyone to create. That means that the core building blocks of the web need to be learnable by everyone, not just programmers.

I get nervous when I see web browsers gaining powerful features that can only be accessed via a JavaScript API. Geolocation is one example: it doesn’t have any declarative equivalent to its JavaScript implementation. Counter-examples would be video and audio: you can use the JavaScript API to get exactly the behaviour you want, if you’ve got that level of knowledge …or you can use the video and audio elements if you’re okay with letting web browsers handle the complexity of display and playback.

I think that CSS hits a nice sweet spot, balancing learnability and power. I love the fact that every bit of CSS ever written comes down to the same basic pattern:

selector {
    property: value;

That’s it!

How amazing is it that one simple pattern can scale to encompass a whole wide world of visual design variety?

Think about the revolution that CSS has gone through in recent years: OOCSS, SMACSS, BEM …these are fundamentally new ways of approaching front-end development, and yet none of these approaches required any changes to be made to the CSS specification. The power and flexibility was already available within its simple selector-property-value pattern.

Mind you, that modularity was compromised when we got things like named animations; a pattern that breaks out of the encapsulation model of CSS. Variables in CSS also break out of the modularity pattern.

Personally, I don’t think there’s any reason to have variables in the CSS language; it’s enough to have them in pre-processing tools. Variables add enormous value for developers, and no value at all for end users. As long as developers can use variables—and they can, with Sass and LESS—I don’t think we need to further complicate CSS.

Bert Bos wrote an exhaustive list of design principles for web standards. There’s some crossover with Tim Berners-Lee’s principles of design, with ideas such as modularity and robustness. Personally, I think that Bert and Håkon did a pretty damn good job of balancing principles like learnability, extensibility, longevity, interoperability and a host of other factors while still producing something powerful enough to scale for the whole web.

There’s one important phrase I want to highlight in the abstract of the 20 year old CSS proposal:

The proposed scheme provides a simple mapping between HTML elements and presentation hints.


Every line of CSS you write is a suggestion. You are not dictating how the HTML should be rendered; you are suggesting how the HTML should be rendered. I find that to be a very liberating and empowering idea.

My only regret is that—twenty years on from the birth of CSS—web browsers are killing the very idea of user stylesheets. Along with “view source”, this feature really drove home the idea that professional web developers are not the only ones who have a say in what gets rendered in web browsers …and that the web truly is for everyone.


Cennydd wrote a really good post recently called Why don’t designers take Android seriously?

I completely agree with his assessment that far too many developers are ignoring or dismissing Android for two distasteful reasons:

  1. Android is difficult
  2. User behaviours are different:

Put uncharitably, the root issue is “Android users are poor”.

But before that, Cennydd compares the future trajectories of other platforms and finds them wanting in comparison to Android: Windows, iOS, …the web.

On that last comparison, I (unsurprisingly) disagree. But it’s not because I think the web is a superior platform; it’s because I don’t think the web is a platform at all.

I wrote about this last month:

The web is not a platform. It’s a continuum.

I think it’s a category error to compare the web to Android or Windows or iOS. It’s like comparing Coca-Cola, Pepsi, and liquid. The web is something that permeates the platforms. From one point of view, this appears to make the web less than the operating system that someone happens to be using to access it. But in the same way that a chicken is an egg’s way of reproducing and a scientist is the universe’s way of observing itself, an operating system is the web’s way of providing access to itself.

Wait a minute, though …Cennydd didn’t actually compare Android to the web. He compared Android to the web browser. Like I’ve said before:

We talk about “the browser” when we should be talking about the browsers. I’m guilty of this. I’ll use phrases like “designing in the browser” or talk about “what we can do in the browser”, when really I should be talking about designing in the browsers and what we can do in the browsers.

But Cennydd’s comparison does raise an interesting question: what is a web browser exactly? Answering that question probably requires an answer to the question: what is the web?

(At this point you might be thinking, “Ah, this is just semantics!” and you’d be right. Abandon ship here if you feel that way. But to describe something as “just semantics” is like pointing at all the written works in every library and saying “but they’re just words”, or taking in the entire trajectory of human civilisation and saying “but those are just ideas”. So yeah, this is “just” semantics.)

So what is the web? Well the unsexy definition I’ve used in the past is that the web consists of files (e.g. HTML, CSS, JavaScript), accessible at URLs, delivered over HTTP. So FTP is not the web. Email is not the web. Gopher is not the web.

But to be honest, I don’t think that the Hypertext Transfer Protocol is the important part of the web; it’s the URLs that really matter. It’s the addressability of the files that’s the killer app of the web in my opinion.

I also don’t think that it’s the file formats themselves that define the web. Don’t get me wrong: I love HTML …and I have nothing against CSS or JavaScript. But if HTML were to disappear, the tears I would weep would not be so much for the format itself, but for the two decades of culture that have been stored with it.

I was re-reading Weaving The Web and in that book, Tim Berners-Lee describes his surprise when people started using HTML to mark up their content. He expected HTML to be used for indices that would point to the URLs of the actual content, which could be in any file format (PDF, word processessing documents, or whatever). It turned out that HTML had just enough expressiveness and grokability to be used instead of those other formats.

So I certainly don’t consider anything that happens to be written using HTML, CSS, and JavaScript to automatically be a part of the web. I can open up a text editor and make an HTML document but as long as it sits on my computer instead of being addressable by a URL, it’s not part of the web. Likewise, a native app might be powered by CSS and JavaScript under the hood, but without a URL, it’s not part of the web.

Perhaps then, a web browser is something that can access URLs. Certainly in pretty much every example of a web browser throughout the web’s history, the URL has been front and centre: if the web were a platform, the URL bar would be its command line.

But, like the rise of HTML, the visibility of the URL in a web browser is an accident of history. It was added almost as an afterthought as a power-user feature: why would most people care what the URL of the content happens to be? It’s the content itself that matters, and you’d get to that content not by typing URLs, but by following hyperlinks.

There’s an argument to be made that, with the rise of search engines, the visibility of URLs has become less important. See, for example, the way that every advertisement for a website on the Tokyo subway doesn’t show a URL; it shows what to type into a search engine instead (and I’ve started seeing this in some TV adverts here in the UK too).

So a web browser that doesn’t expose the URLs of what it’s rendering is still a web browser.

Now imagine a browser that you install on your device that doesn’t expose URLs, but under the hood it is navigating between URLs using HTTP, and rendering the content (images, JavaScript, CSS, HTML, JSON, whatever). That’s a pretty good description of many native apps. There’s a whole category of native apps that could just as easily be described as “artisanal web browsers” (and if someone wants to write a browser extension that replaces every mention of “native app” with “artisanal web browser” that would be just peachy).

Instagram’s native app is a web browser.

Facebook’s native app is a web browser.

Twitter’s native app is a web browser.

Like Paul said:

Monolithic browsers are not the only User Agent.

I was initially confused when Anna tweeted:

Reading the responses to @Cennydd’s tweet about designers needing to pay attention to Android. The web is fragmented. That’s our job.

I understood Cennydd’s point to be about native apps, not the web. But if, as I’ve just said, many native apps are in fact web browsers, does that mean that making native apps is a form of web development?

I don’t think so. I think making a native app has much more in common with making a web browser than it does with making a web site/app/thang. Certainly the work that Clearleft has done in this area felt that way: the Channel 4 News app is a browser for Channel 4 News; the Evo iPad app is a browser for Evo.

So if your job involves making browsers like those, then yes, you absolutely should be paying more attention to Android, for all the reasons that Cennydd suggests.

But if, like me, you have zero interest in making browsers—whether it’s a browser for Android, iOS, OS X, Windows, Blackberry, Linux, or NeXT—you should still be paying attention to Android because it’s just one of the many ways that people will be accessing the web.

It’s all too easy for us to fall into the trap of thinking that people will only be using traditional monolithic web browsers to access what we build. The truth is that our work will be accessed on the desktop, on mobile, and on tablets, but also on watches, on televisions, and sure, even fridges, but also on platforms that may not even have screens.

It’s certainly worth remembering that what you make will be viewed in the context of an artisanal browser. Like Jen says:

The “native apps are better” argument ignores the fact one of the most popular things to do in apps is read the web.

But just because we know that our work will be accessed on a whole range of devices and platforms doesn’t mean that we should optimise for those specific devices and platforms. That just won’t scale. The only sane future-friendly approach is to take a device-agnostic, platform-agnostic approach and deliver something that’s robust enough to work in this stunningly-wide range of browsers and user-agents (hint: progressive enhancement is your friend).

I completely agree with Cennydd: I think that ignoring Android is narrow-minded, blinkered and foolish …but I feel the same way about ignoring Windows, Blackberry, Nokia, or the Playstation. I also think it would be foolish to focus on any one of those platforms at the expense of others.

I love the fact that the web can be accessed on so many platforms and devices by so many different kinds of browsers. I only wish there more: more operating systems, more kinds of devices, more browsers. Any platform that allows more people to access the web is good with me. That’s why I, like Cennydd, welcome the rise of Android.

Stop seeing fragmentation. Start seeing diversity.

Making progress

When I was talking about Async, Ajax, and animation, I mentioned the little trick I’ve used of generating a progress element to indicate to the user that an Ajax request is underway.

I sometimes use the same technique even if Ajax isn’t involved. When a form is being submitted, I find it’s often good to provide explicit, immediate feedback that the submission is underway. Sure, the browser will do its own thing but a browser doesn’t differentiate between showing that a regular link has been clicked, and showing that all those important details you just entered into a form are on their way.

Here’s the JavaScript I use. It’s fairly simplistic, and I’m limiting it to POST requests only. At the moment that a form begins to submit, a progress element is inserted at the end of the form …which is usually right by the submit button that the user will have just pressed.

While I’m at it, I also set a variable to indicate that a POST submission is underway. So even if the user clicks on that submit button multiple times, only one request is set.

You’ll notice that I’m attaching an event to each form element, rather than using event delegation to listen for a click event on the parent document and then figuring out whether that click event was triggered by a submit button. Usually I’m a big fan of event delegation but in this case, it’s important that the event I’m listening to is the submit event. A form won’t fire that event unless the data is truly winging its way to the server. That means you can do all the client-side validation you want—making good use of the required attribute where appropriate—safe in the knowledge that the progess element won’t be generated until the form has passed its validation checks.

If you like this particular pattern, feel free to use the code. Better yet, improve upon it.


Stop me if you’ve heard this one before. You’re reading an article on Smashing Magazine or A List Apart or some other publication. The article is about a specific feature of CSS, or maybe JavaScript, or perhaps it’s exploring some of the newer additions to HTML. The article is good. It explains how to use this particular feature in your work. Then you read the comments. The first comment is inevitably from someone bemoaning the fact that they can’t use this feature because it isn’t supported in every browser. Specifically, it isn’t supported in some older version of Internet Explorer that they have to support. Therefore the entire article is rendered null and void.

That attitude infuriates and depresses me. It seems to me that it demonstrates a fundamental mismatch between how that person views the job of web development and the way the web actually works.

It is entirely possible—nay, desirable—to use features long before they are supported in every browser. That’s how we move the web forward. If we waited until there was universal support for a feature before we used it, we’d still be using CSS 1.0 and HTML 2.0.

If you use a CSS feature that isn’t supported in a particular browser—like, say, an older version of Internet Explorer—that browser will simply ignore that CSS rule. So you don’t get that rounded corner, or text shadow, or whatever it was. Browsers have the same error-handling mechanism for HTML: if they see something they don’t understand, they just ignore it. The browser will not throw an error. The browser will not stop rendering the page. Browsers are very liberal in what they accept.

It’s a bit trickier with JavaScript: browsers will throw an error; browsers will stop processing the script. That’s why it’s important to use feature detection. That’s also why you definitely don’t want to rely on JavaScript for rendering your content—it’s the most fragile layer of the front-end stack. Note, I’m not saying don’t use JavaScript; I’m saying don’t rely on JavaScript. Otherwise you’ve got yourself a SPOF.

Anyway, my point is—and I can’t believe I still have to repeat this after all these years—websites do not need to look exactly the same in every browser.

“But my client!”, cries the Smashing Magazine commenter, “But my boss!”

If your client or boss expects that a website will look and behave the same in every browser on every device, then where did they get those expectations from? And rather than spending your time trying to meet those impossible expectations, I think your time would be better spent explaining why those expectations don’t match the reality of the web.

It’s like Mike Monteiro says about clients: if they just don’t get something about your design, that’s not their fault; it’s yours. Explaining your design work is part of your design work. It’s the same with web development. It’s our job to explain how the web works …and how the unevenly-distributed nature of browser capabilities is not a bug, it’s a feature.

That was true fourteen years ago when John Allsopp wrote A Dao Of Web Design, and it’s still true today. Back then, designers and developers were comparing the web to print and finding it wanting. These days, designers and developers are comparing the web to native and finding it wanting. In both cases, I feel like they’re missing the fundamental point of the web: you can provide universal access to content and tasks without providing exactly the same experience for every single browser or device. That’s not a failing of the web—that’s its killer app.

Paul Kinlan published a post called This Is the Web Platform where he tabulates the current state of browser support for various features. “Pretty damning” he says:

the feature support that is ubiquitous across the web is actually pretty small especially if you are supporting IE8.

That’s true …from a certain point of view. But it depends on your definition of “support”. If your definition of “support” is “must look and work identically to the latest version of Chrome”, then yes, you’re going to have a smaller set of features you can use (you’re also going to live a miserable existence). But if your definition of “support” is “must be able to access the content and accomplish the task”, then as long as you’re using progressive enhancement, you can use all the features you want and support Internet Explorer 8, 7, 6, 5 …you can support every browser capable of connecting to the internet.

Like Brad said:

There is a difference between support and optimization.

I think part of the problem may be with the language we use. We talk about “the browser” when we should be talking about the browsers. I’m guilty of this. I’ll use phrases like “designing in the browser” or talk about “what we can do in the browser”, when really I should be talking about designing in the browsers and what we can do in the browsers.

It’s a subtle Lakoffian thing, but when we talk about “the browser” as if it were a single entity we might be unconsiously reinforcing the expectation that there is one Platonic ideal of browser rendering and that’s what we’re designing for.

There’s another phrase that bothers me, and it’s the phrase that Paul used in the title of his article: “the web platform”. This is something I talked about back in November in my presentation The Power of Simplicity:

But this idea of the web as a platform, I get why from a marketing perspective, we’d want to use that phrase, because it puts the web on equal footing with genuine platforms.

I would say Flash is a platform, and native: iOS and Android and these things. They are platforms, in that it’s all one bundle. And the web isn’t like that.

What I mean is, if you use the Flash platform, then anyone with the Flash plug-in can get your content. It’s on or off. It’s one or zero; it’s binary. Either they have the platform or they don’t. Either they get all your content, or none of your content.

And it’s similar with native apps. If you’ve got the right phone, you can get my app. All of my app. You don’t get bits of my app, you get all my app. Or you get none of it because you don’t have that particular phone that I’m supporting.

And the web is not like that. The web is not binary, one or zero, on or off. It’s not a platform where you get one hundred per cent or zero per cent. It’s this continuum.

The web is not a platform. It’s a continuum.


Aral and his trusty sidekick Victor have taken up residency for a while at the Clearleft office in Middle Street while they work on their very exciting project. It’s nice having them around.

I got chatting to Aral about a markup pattern that’s become fairly prevalent since the rise of Github: linking to the source code for a website or project. You know, like when you see “fork me on Github” links.

We were talking about how it would be nice to have some machine-readable way of explicitly marking up those kind of links, whether they’re in the head of the document, or visible in the body. Sounds like a job for the rel attribute, I thought.

The rel attribute describes the relationship of the current document to the linked document. You can use it on the link element (in the head of your document) and the a element (in the body). The example that everyone is familiar with is rel=”stylesheet” when linking off to a CSS file—the linked document has the relationship of being a stylesheet for the current document.

The rel attribute could theoretically take a space-separated list of any values, just like the class attribute. In practice, there’s much more value in having everyone agree on which rel values should be used.

There used to be a page on the WHATWG site for listing rel values, but it tended to stagnate. So now the official registry for rel values is on the microformats wiki. That’s where you can see which values are recommended for use today and you can also brainstorm new ideas.

The benefit of having one centralised for this is that you can see if someone else has had the same idea as you. Then you can come to agreement on which value to use, so that everyone’s using the same vocabulary instead of just making stuff up.

It doesn’t look like there’s an existing value for the use case of linking to a document’s (or a project’s) source code so I’ve proposed rel=”source”.

Now I should document some use cases of people linking their site to its source code. It might be that wikis qualify as another use case: every “edit” button points to the source of the document in wiki markup.

If you have any thoughts on this pattern, or examples to add, please feel free to add them.

Defining the damn thang

Chris recently documented the results from his survey which asked:

Is it useful to distinguish between “web apps” and “web sites”?

His conclusion:

There is just nothing but questions, exemptions, and gray area.

This is something I wrote about a while back:

Like obscenity and brunch, web apps can be described but not defined.

The results of Chris’s poll are telling. The majority of people believe there is a difference between sites and apps …but nobody can agree on what it is. The comments make for interesting reading too. The more people chime in an attempt to define exactly what a “web app” is, the more it proves the point that the the term “web app” isn’t a useful word (in the sense that useful words should have an agreed-upon meaning).

Tyler Sticka makes a good point:

By this definition, web apps are just a subset of websites.

I like that. It avoids the false dichotomy that a product is either a site or an app.

But although it seems that the term “web app” can’t be defined, there are a lot of really smart people who still think it has some value.

I think Cennydd is right. I think the differences exist …but I also think we’re looking for those differences at the wrong scale. Rather than describing an entire product as either a website or an web app, I think it makes much more sense to distinguish between patterns.

Let’s take those two modifiers—behavioural and informational. But let’s apply them at the pattern level.

The “get stuff” sites that Jake describes will have a lot of informational patterns: how best to present a flow of text for reading, for example. Typography, contrast, whitespace; all of those attributes are important for an informational pattern.

The “do stuff” sites will probably have a lot of behavioural patterns: entering information or performing an action. Feedback, animation, speed; these are some of the possible attributes of a behavioural pattern.

But just about every product out there on the web contains a combination of both types of pattern. Like I said:

Is Wikipedia a website up until the point that I start editing an article? Are Twitter and Pinterest websites while I’m browsing through them but then flip into being web apps the moment that I post something?

Now you could make an arbitrary decision that any product with more than 50% informational patterns is a website, and any product with more than 50% behavioural patterns is a web app, but I don’t think that’s very useful.

Take a look at Brad’s collection of responsive patterns. Some of them are clearly informational (tables, images, etc.), while some of them are much more behavioural (carousels, notifications, etc.). But Brad doesn’t divide his collection into two, saying “Here are the patterns for websites” and “Here are the patterns for web apps.” That would be a dumb way to divide up his patterns, and I think it’s an equally dumb way to divide up the whole web.

What I’m getting at here is that, rather than trying to answer the question “what is a web app, anyway?”, I think it’s far more important to answer the other question I posed:


Why do you want to make that distinction? What benefit do you gain by arbitrarily dividing the entire web into two classes?

I think by making the distinction at the pattern level, that question starts to become a bit easier to answer. One possible answer is to do with the different skills involved.

For example, I know plenty of designers who are really, really good at informational patterns—they can lay out content in a beautiful, clear way. But they are less skilled when it comes to thinking through all the permutations involved in behavioural patterns—the “arrow of time” that’s part of so much interaction design. And vice-versa: a skilled interaction designer isn’t necessarily the best at old-skill knowledge of type, margins, and hierarchy. But both skillsets will be required on an almost every project on the web.

So I do believe there is value in distinguishing between behaviour and information …but I don’t believe there is value in trying to shoehorn entire products into just one of those categories. Making the distinction at the pattern level, though? That I can get behind.


Incidentally, some of the respondents to Chris’s poll shared my feeling that the term “web app” was often used from a marketing perspective to make something sound more important and superior:

Perhaps it’s simply fashion. Perhaps “website” just sounds old-fashioned, and “web app” lends your product a more up-to-date, zingy feeling on par with the native apps available from the carefully-curated walled gardens of app stores.

Approaching things from the patterns perspective, I wonder if those same feelings of inferiority and superiority are driving the recent crop of behavioural patterns for informational content: parallaxy, snowfally, animation patterns are being applied on top of traditional informational patterns like hierarchy, measure, and art direction. I’m not sure that the juxtaposition is working that well. Taking the single interaction involved in long-form informational patterns (that interaction would be scrolling) and then using it as a trigger for all kinds of behavioural patterns feels …uncanny.

Parsing webmentions

Thanks to everyone who helped me test webmentions that I hacked together at Indie Web Camp last weekend.

Let me explain what web mentions are all about…

Basically, it’s an equivalent to pingback. Let’s say I write something here on adactio.com. Suppose that prompts you to write something in response on your own site. A web mention is a way for you to let me know that your response exists.

If you look in the head of any of my journal posts, you’ll see this link element:

<link rel="webmention" href="http://adactio.com/webmention.php" />

That’s my web mention endpoint: http://adactio.com/webmention.php …it’s kind of like a webhook: a URL that’s intended to be hit by machines rather than people. So when you publish your response to my post, you ping that URL with a POST request that sends two parameters:

  1. target: the URL of my post and
  2. source: the URL of your response.

Ideally your own CMS or blogging system would take care of doing the pinging, but until that’s more widely implemented, I’m providing this form at the end of each of my posts:

Either way, once you ping my web mention endpoint—discoverable through that link rel="webmention"—with those two parameters, I just need to confirm that your post does indeed contain a link to my post—by making a cURL request and parsing your source—and then I return a server response of 202 (Accepted).

Here’s the code for a minimum viable web mention in PHP.

That’s as far as I got at Indie Web Camp but it was enough for me to start collecting responses to posts.

Webmentions as links

The next step is to do something with the responses. After all, I’ve already got the source of each response from those cURL requests.

Barnaby has a written a nice straightforward microformats parser in PHP. I’m using that to check the cURLed source for any responses that have been marked up using h-entry. That’s one of the microformats 2 vocabularies—a much simpler way of writing structured content with microformats.

Aaron, Amber, and Barnaby all sent responses that were marked up with h-entry so now their responses appear in full.

Webmentions as comments

So there you have it. Comments are now open on every journal post on adactio.com …the only catch is that you have to write the comment on your own site. And if you want the content of your post to appear here (instead of just a link) then update your blog post template to include a handful of h-entry classes.

Feel free to use this post as a test. Mark up your blog with h-entry, write a post that links to this URL, and enter the URL of your post in the form below.