If I got made king of web browsers, here’s what I’d do

13.41, Friday 7 Aug 2020 Link to this post

It’s hot and it’s lunchtime, so let’s pretend I’m in charge of major global technical infrastructure!

I wrote about how I would improve RSS the other day (because being able to subscribe to text is super neat, but it’s so arcane compared to smartphone apps). And after writing that, it occurred to me that the problem is wider:

The user experience of the web itself sucks.

It is less pleasant to use a web browser than it is to use apps. But that’s because the browser-makers (Google and Apple, primarily) have silently abdicated their responsibility to make browsing good. I get it, they’re conflicted, they’re also running super profitable app stores.

And also, I guess, because browser-makers tend to be engineers, so they do engineering-type things like making the browser an app-delivery platform able to run compiled code. Or fight meaningless user experience battles like hiding the URL, or hiding View Source – both acts that don’t really help early users that much, but definitely impede the user path from being a consumer to being a fully-fledged participant/maker.

You know, and also making humble improvements to the web is unglamorous? It’s hard to measure. It might never be noticed. It’s probably not going to get you your next bonus. Perhaps that’s it.


So what would I do? I would focus on

  • the humble yet meaningful – nothing huge like changing the way the web works. Don’t be revolutionary. But also, no more futzing and tinkering like tweaking the way CSS works.
  • improving what’s already there – it’s fine to add brand new capabilities like being able to talk to external hardware, or peer-to-peer video. But the foundations are crumbling. Look at what’s broken and shore it up. Pave the cowpaths, and all that.
  • the user experience – so much browser effort goes into making things easier for developer, and primarily the developers of crazy high traffic sites. But who cares. Software engineers get paid a fortune, let them sing for their supper. If developer experience matters that much, focus on the long tail of really simply websites instead.
  • browsing not apps – this is a rebalancing, but it feels like recent-ish features like “pinned tabs” etc are about making the web browser a place for web apps. What about reading and writing?
  • making change happen – every improvement carrot needs to be paired with an encouragement stick.

Specifically? Here are three ideas to start, totally off the top of my head.

  • Newsletters. I get bombarded by newsletter signup pop-overs whenever I go to a website. Browsers should block these, they suck. BUT instead there should be a button in the browser toolbar to sign up, and it should glow when available. And the blank “new window” screen should list all the websites I’ve encountered in the last 24 hours with newsletters (with screenshots), and a quick subscribe button for each.
  • Forms. A web browser should never, ever, ever forget something I type into a form. Too often I go to write something long, and the page doesn’t submit properly, or I accidentally hit back, and I lose it. Or I write an email into a contact form, and never get a copy. Browsers need the equivalent of Drafts and Sent folders.
  • Virality. Like it or not, many websites live and die by how much individual pages are shared. The web was successful in the first place because URLs were easy to copy and paste into email… but we can do better now. There’s a Web Share API (technical docs) which pops open the standard share sheet on mobile, but it’s kinda neither here not there. Instead the browser should have social media share icons built in, and APIs to those platforms so the buttons can show a count of how many times the page has been shared. In short, browsers need excellent retweet and fwd buttons.

Or, or, or!

Move the “home” button and the sitemap into the address bar! Let webpages have a standard and exciting way to suggest related articles! Make Bookmarks and History properly smart (highlight my daily visits, for example) and add them as folders on my smartphone home screen! Embrace the trend of personal wikis, and also hypertext and protocols like Quotebacks. More bonkers ideas? I’ve got ‘em: Making Senses, a presentation from 2006.

And then just keep on implementing ideas like that. Find something that sucks. Make the experience better. Repeat 100 times. In two years, look back and see how far we’ve come.

Hey, here’s a bonus idea but it’s a tough one: Google Apps for everything. The experience of writing a Google Doc is awesome. Seeing other people’s cursors, live changes, suggested changes that can be approved/rejected; it’s robust to dropping offline, there are both anonymous and signed-in users, etc. This should be something that any website can do. Sites should be able to identify a user, provide a collection of user handles by whatever method they choose for the collaborating group, and the browser should do the rest.


What I don’t want is for this to lead to a sameness of the web. Websites really are beginning to look the same and that’s a shame. As Benedict Evans put it in his newsletter: maybe this is the same as the way wind tunnel data made all cars look the same.

Rather: provide optional architectures for websites that are good for site users and good for site creators too. Make that space. Make it crazy easy to develop in. Then get out of the way.

There’s probably no immediate commercial reason for this humble kind of work. But the web is the commons of the internet. Looked after, it’s a renewable resource of new ideas and approaches that don’t fit prevailing economic models – and also the one place on the internet that is friendly to history. We mustn’t lose it.

There are people looking at how to completely change web browsing. How to make it social. How to including payments, or publishing, or whatever. Brilliant. Let them. Meanwhile, the web we’ve got is a mess, and we’ll never to these new ideas if the centre doesn’t hold.

So who’s looking at this? Is there a team in Chrome and a team in Safari advocating for the experience of the web, as the web, or does this need to come from somewhere else?