Some people really do want their Instagram back

Taylor Lorenz has written a piece on why you don’t want the old Instagram back and I couldn’t disagree more with it.

No really, a lot of people do want the old Instagram. Insta became a great space for visual artists who adopted the platform as the de facto place where they show their work and build community it.

People think that bringing back the “old” Instagram design, or a chronological feed will somehow recapture the magic of using Instagram in 2014. It won’t. That time is gone and the internet and culture have irrevocably changed. Most importantly, how and what we want to share on the internet has changed.

What Taylor is missing here is that Instagram was always a place for discovery in a way which was controlled by the user. TikTok gives you little control over what you see: the algorithm decides that for you in a way which is much more opaque than even Facebook’s much-maligned news feed. Facebook at least has a notion of what kinds of content is important to you: birthdays, babies, weddings, and all the major events of life. TikTok? Not so much

These days, intimacy is fostered through features like DMs, group chats, or ephemeral posts to Close Friends.

Except that intimacy needs to start somewhere. And that somewhere can’t, by definition, be a DM.

Spaces which are private like the ones Taylor is describing have only two kinds of relationship in them: the already-established relationship between people who already know each other from elsewhere, or the relationship between an advertiser who has paid to be in that space and the people in it.

People think that bringing back the “old” Instagram design, or a chronological feed will somehow recapture the magic of using Instagram in 2014. It won’t. That time is gone and the internet and culture have irrevocably changed.

Culture is not monolithic: platforms have many forms of users, and users find the use for platforms – not the other way round.

This piece is a great example of the tech journalism communities obsession with the newest latest smartest thing, with destroying the old no matter what the cost. And it’s just wrong.

Apple’s “repairwashing”

Cory Doctorow on Apple’s cement overshoes:

“In fact, nearly every part of Apple’s official repair process was worse than the iFixit equivalent. The useless battery-seating press kept knocking the battery out of alignment, and the fancy torx drivers were choresome to use. All of this compounded Apple’s repair-hostile design: swapping the battery requires three different screwdriver bits, removing the speaker, and managing a cluster of hidden fasteners that hold down the fiddly ribbon cables. Apple’s official tools don’t have (industry standard) magnetic tips, so Hollister spent a lot of time chasing minuscule pieces of metal around his workbench.”

It’s tough to see Apple’s self-repair programme as anything but, as Cory puts it, “repairwashing”. What Apple should be doing is building phones and computers which are easy to repair. The company has the best industrial designers in the world. If any business can make things that are upgradeable and serviceable without thousands of pounds of equipment, it’s Apple.

Dipping my toes into Linux (again)

I’m no stranger to Linux. I went through a phase in the noughties of running it, partly inspired by the moves from Mark Pilgrim and Cory Doctorow, and for a while, I used a boxy fat Dell Inspiron running Ubuntu. That wasn’t a great experience: Ubuntu was renowned at the time for its ease of use and friendliness to non-technical users, but all too often, I found myself having to dig into command lines and tweak drivers. Unlike Cory, I didn’t have a hotline into Canonical.

I headed back to using the Mac because there were just too many rough edges around using Linux, and the applications available weren’t as good as those I was used to. For example, I hated OpenOffice, which looked like an old and clunky version of Microsoft Word. In fact, most of the software I was using looked like something from the late 1990s, the time before user interface design was done by designers, rather than coders moonlighting with a copy of Paint.

Times change. An awful lot of what I do on a day to day basis can be done on Linux. The application I do most of my writing in (Typora) has a Linux version. Firefox has recently become my primary browser. Microsoft has a version of Teams for Linux, which means I can use a Linux machine for my day job.

When I recently bought a new Windows machine, I bought a ThinkPad X1 Carbon because I knew I wanted to try Linux out again, and it’s a great laptop to run it on. You can buy the model that I purchased with Ubuntu pre-installed rather than Windows (and I was pretty tempted at the time). It’s massively over-specified for a Linux laptop, with an 11th generation i7, 32Gb of RAM and a 1Tb SSD, but this means it’s also a computer which could last me a long time, especially if I’m not being driven along by the endless hardware upgrade cycle which Microsoft and its partners would love to see you on.

With all that in mind, I decided that the time was right to dip my toes back into the world of Linux. So, after watching a few videos about the various distributions available, I downloaded and installed Zorin OS, a pretty Mac-like experience. It’s also very beginner-friendly: it even includes an ingenious way of handling Windows applications by installing Wine and any required supporting files when you first click on a Windows installer.

The installation experience was as smooth as you would expect: Linux distributions are now mature and stable enough that installing on anything but peculiar hardware will be easy. It lets me quickly partition my drive, splitting it between Windows and Linux. It was then a quick task to customise the interface to look more Mac-like, install the applications I wanted, and away I went.

Then two days later, I changed my mind and decided I wanted to use the latest Long Term release of Ubuntu instead. Canonical issues a new long-term support version of Ubuntu every year, which guarantees five years of updates and support. LTS releases tend to be stable, which was also a selling point.

Ubuntu 22.04 LTS uses the latest version of Gnome, which has an updated display engine which is smoother than previous versions. I also wanted to work on something as mainstream as possible. I was sold on using just Linux on my ThinkPad and getting rid of Windows altogether. The overall experience had been more than good enough, and I’ve fallen out of love with Windows 11. So I installed Ubuntu, taking the whole 1Tb SSD for it. Goodbye, Windows 11.

As I had expected, I’ve had no problems at all with hardware. In fact, some of the hardware that I had expected to have to do some fiddling and driver installation with worked immediately. For example, the fingerprint reader, which I used to log in to the machine, works just as well with Ubuntu as it did with Windows. My CalDigit Thunderbolt dock, which I use to connect to the webcam, display and a few other peripherals, also works perfectly.

The only place where I tripped up a little was adjusting the display to get the UI sized correctly. The display on my ThinkPad is a 16:10 panel which runs at 1920×1200. Running at the native resolution makes text and UI elements a little small; I enabled fractional sizing and bumped it up to 125%, the same as Windows runs. That works, but it leaves the text blurry – not enough to make it unusable, but annoying. The answer was to leave it running at 100% but turn on Large Text in the Accessibility options. The text was now crisp and large enough for me.

And that is all the interface tweaking I’ve done, other than moving the dock to the bottom and making it look a little more Mac-like, all using the built-in options with no additional software installed.

Operating systems should get out of the way and let you get on with the actual work you do. If you’re worried about and twiddling with the OS, you’re playing rather than producing. There’s nothing wrong with that if you enjoy it. But for people who want to do some real work, the OS shouldn’t be something you think about much.

So far, for me, Ubuntu is doing precisely that. I’m writing, not computing. I’m creating, not being bamboozled by a thousand different options. I’m delighted with it, so I really can’t write much more about the experience. It’s just been simple.

Paul Thurrott is very unhappy

Paul Thurrott is really unhappy with the current direction of Windows (subscriber only link, and I think he has a lot of good points:

“Naturally, this made me think of Windows, and of Microsoft’s incessant, slow boil moves to forever ruin its user experience with crapware bundling, forced telemetry tracking, and, yes, advertising. These are the times that try one’s soul, as Thomas Paine once opined of an admitted more serious historical crisis. But I feel the pain all the same. And as time goes on, and Windows 8 becomes Windows 10 becomes Windows 11, it just gets worse.”

I think there’s a lot to like about Windows 11, but after using it extensively for a while I tend to agree with Paul. I like Windows 11’s simplicity and the way that’s stripped away a lot of what I see as legacy cruft in favour of something that looks clean and modern. But Microsoft being Microsoft, you can already see the temptation to put in ads, prod you towards using Bing, make you love Microsoft News (no really that’s not going to happen guys).

You can see how this happens: every team in Microsoft sees how it can “add value” to Windows and without really strong leadership this turns into a mess. It seems like the company has learned nothing since the days which Steven Sinofsky is describing in his excellent “Learning by shipping” series of emails.

Some thoughts on the Surface Duo

Despite Microsoft pricing the device to fail in the UK I’ve somehow ended up buying a Surface Duo. Yes, the cost here is ridiculous – £1300 at a time when the price has been reduced to $999 in the US – but I’ve always loved Microsoft’s Surface line and was curious about it. And I’m in the incredibly fortunate position of being able to support curiosity purchases. Not spending £700 on commuting tends to do that.

Some quick thoughts:

  • Microsoft is on to something when it talks about the productivity of dual screens, and I think it’s correct that dual screens are better than a single large foldable screen for this purpose. I’d really like to see this in a larger format as a tablet, which was obviously the intention behind the now-canned Surface Neo. Even though it’s a lot smaller, I prefer having two screens versus a single one which splits virtually. I can’t quite explain why, but it feels much more natural.
  • Apps which span are few and far between — basically Microsoft’s own and a hand full of others. However, I’ve actually found myself almost never using it in this mode, even with Microsoft apps. It works nicely, and the apps are — in my experience — now pretty rock solid but I just don’t use it. Much more often I’m using a pair of apps to do something, such as Outlook on one side and To Do on the other, dragging and dropping neatly between the two.
  • The hardware really is beautifully designed, and you can see that the team who created it have poured their heart and soul into it. The hinge on its own is a thing of joy: incredibly smooth, just the right resistance, and firm enough to hold in place without accidentally getting knocked into another position. Well done, Microsoft.
  • When I say it is beautifully designed, I mean that it pases my “pick up” test: this is a device that you constantly want to pick up and use.
  • Lack of 5G, NFC and so on do not feel like limitations at this point. Likewise, the camera: it’s good enough for the things it is designed to so (video calling, quick captures of documents and whiteboards). No one is going to use this as their main camera.
  • Software performance is fine. Microsoft seems to have got the bugs out, and it never feels slow, which probably goes to prove that if you take out the need for lots of performance for image processing and AI, you really don’t need to have the latest generation of chip.
  • The one area that Microsoft needs to work on more is typing, as the experience of the on-screen keyboard is hit and miss. It’s perfectly fine when you are thumb typing with it folded over, and when you have a split screen and are holding it like a little laptop, thumb typing away. However, when it tries to shift the keyboard over to the side a little so you can type one-handed, it doesn’t quite shift over enough if you have small hands. Most of the time I find myself holding it with one hand and swiping with the other instead – and that is fine.

Overall I think Microsoft is on to something with this form factor, but I really wish it was larger and a tablet rather than smaller and a sort-of phone. Microsoft is absolutely correct not to market this as a smartphone, because it never really feels like one — but it does feel like a tiny, interesting and highly usable mini-tablet.

John Gruber on Jason Snell on iOS Markdown Editors


“I have no idea why there are now apps that use Markdown as their back end storage format but only show styled text without the Markdown source code visible… If you want Markdown, show the Markdown. Trust me, it’s meant to be shown.”

Simple answer to the “why”: for compatibility. I use Ulysses for writing most of my content. But it’s not always the place where words get originated. Quite often, anything longer than a quick post (like this one) will start its life on my Freewrite rather than my Mac, particularly if it’s written somewhere other than my home office1.

Because Ulysses uses Markdown as its underlying format, if I want to switch the app I use to write that’s not a problem. Likewise, it makes getting words into Ulysses incredibly easy.

Not everyone wants to see every piece of source code behind an interface, but everyone should be able to write without worrying that the underlying format will mean no ability to read those words in the future. As a kind of lingua franca for structured words, Markdown is great, but not everyone wants to see [links looking like this]( when they are writing. For some people – like me – it’s distracting and a little irritating to see everything. If anything, I would like an option in Ulysses to just hide all the formatting stuff.

For me, Ulysses offers just the right balance: a really good editor, using Markdown so I know I can always read those words no matter what platform I am using in the future, and an excellent client on both Mac and iPad. Markdown might have “meant to be shown” but it doesn’t have to be shown.

  1. After using the Freewrite’s keyboard I bought a DasKeyboard Pro with the same switches, and since then the keyboard on the 16in MacBook Pro feels like mushy crap. Sorry Apple.

Writing with the Freewrite

For anyone that has spent the last 20 years or so typing on ever less satisfying keyboards, writing with the Astrohaus Freewrite is a strange experience. In fact, in some ways it is profoundly disconcerting. Going back to a normal computer keyboard and regular large screen is like emerging from a wilderness retreat.

Like a wilderness retreat, the Freewrite is an attempt to regain some simplicity and with it that often talked of but little understood condition of flow. Flow has attained almost mystical status in the productivity and artistic communities as a kind of meditative state in which, thanks to an extreme of focus, the words just come. In an age when your phone goes ping and there’s an incoming message request every few seconds from one or more demanding social networks, flow is power.

So, how does this relate to the chunky little box on which I’m typing this draft? It helps at this point to describe what the Freewrite is and what it’s intended to do. Launched in 2017 by a small company called Astrohaus, Freewrite is a combination of excellent mechanical keyboard and e-ink screen, all in a box which looks like a steampunk typewriter. You type, words appear on screen. That, in a nutshell is what Freewrite is and does.


Compared to any computer you’re likely to use daily the Freewrite is an absolute chonk. It weighs around four pounds and the case is made from some kind of metal — probably aluminium but whatever it is it’s substantial. The base is plastic with a slightly grippy feel, which obviously ensures it firmly stays even on your expensive executive glass meeting room table, Mr Hipster Executive.

Everything about this machine is designed to look and feel analogue. The switches which allow you to change which folder you’re writing in and turn Wi-Fi on and off are big, mechanical beasts rather than wimpy little buttons. The power switch is a red button with a satisfying push to it.

Then, of course, there is the keyboard. It’s mechanical and uses Cherry MX Brown switches. For those relatively new to technical switches, the Brown switch is often favoured as it has a relatively short depth of travel before the key is activated, but then an extended depression which gives it a satisfying thunk if you bottom it out. In other words, you can go relatively gently on it and be about as quiet as a mechanical keyboard can get, or you can hammer it a little harder and thunk to your heart’s content. Because the activation depth is relatively short, if you are coming from a normal computer keyboard it will not feel like you have to do insane finger workouts before you can actually type anything at speed.

Just as importantly, like all quality keyboards (and like no laptop keyboards) this is raked at an angle, which gives your hands a more comfortable position and lets them travel across the board a little faster. Once you get used to it, don’t be surprised if your typing speed goes up, but do be prepared to make many errors in the meantime. If, like me, you have typed millions of words on laptop keyboards and just got used to them, it will take a while to adjust. But it’s worth it in the end. Your hands and your typing speed will thank you.

The screen

The screen on the Freewrite is a pair of e-ink panels. The top one, which is about the size of a small smartphone, is where your words appear as you type. As you would expect from an e-ink screen it is a little laggy, but not dramatically so. There’s a backlight for typing in dimmer conditions, or you can turn that light off and it’s perfectly readable in almost every light.

The second screen just underneath serves as an information panel, giving you useful information like a word count, a timer, or a couple of different kinds of clock. I suspect that for me this will stay on the word count permanently. I don’t often need a timer, and annoyingly the timer isn’t persistent. If the company upgrades the software, I would love to see an information panel which combines word count and timer, which I think would be the most useful option of all.

A digression

While I was finishing off that sentence I heard the “bong” of Outlook demanding my attention on my laptop and quickly had to break to answer an email. And I instantly realised that this keyboard is going to spoil my experience of my lovely new MacBook Pro unless I get an external Cherry MX Brown keyboard to go with it. The feeling is that different, and even now I can tell it’s that much better. Which leaves me wondering why I resisted getting into mechanical keyboards for this long?


In line with the rest of the device the software on the Freewrite is minimal. In fact, there really isn’t anything which anyone raised typing on computers would recognise as software. There are no apps. You just type, and words appear on the screen. That’s it.

You have three folders to store files in and you access them not through a fancy touch screen but by moving a mechanical switch. You turn it off and get a screensaver. Turn it on and you are back where you left off in whatever document you’re working on. There’s no spell check, no grammar checker, nothing which could potentially get between you and writing.

The philosophical typewriter

This is because, unlike most devices, the Freewrite is opinionated about the way you should work. Computers, particularly modern ones, are built on the principle that the user is always right, at least about the way they want to work. Don’t like this word processor? Fine! Use this other one instead! Like to write and edit as you go along? Word will let you do that all day long, with its squiggly red lines and attention-grabbing autocorrects.

The Freewrite does not care about your approach. It is built for one approach and one only: you draft, by writing in an uninterrupted way as possible. Then you use another tool to edit and turn your steam of consciousness and raw words into something polished.

Freewrite has absolutely no aspirations to be a device that you edit on. It lacks every editing capability except for a backspace key to get rid of that typo in the previous word. There are no arrow keys. Although you can page up and page down if you want to look back into your text, there’s no cursor that you can more anywhere backwards.

Like I said, the Freewrite has an opinion about the way you should write, and if you don’t like that opinion you should go somewhere else.

However, the Freewrite does have another trick up its sleeve: it does everything possible to make sure that you can get your words into another tool for editing which suits you.

Postbox and sync

On the back of the Freewrite is a USB Type-C port which you can use for charging the device, or for connecting it to a computer and pulling files off it. However, remember that Wi-Fi switch? It’s there to connect you with Postbox, which is Astrohaus’ cloud sync service. When connected to Wi-Fi every thirty seconds or so Freewrite connects to Postbox and syncs the latest version of your document to the cloud. From there, you can sync it to Google Drive or Dropbox, and these synced versions are in Word format. This means you can seamlessly move from drafting with the Freewrite to editing in Word, with all its excellent editing features.

There’s just one caveat: it’s not really sync as you normally understand it. Although Postbox includes a capable little distraction-free writer called Sprinter, changes you make in Sprinter are not synced back to the Freewrite. In fact, if you open a Freewrite-originated document in Sprinter it moves it to a different folder which the Freewrite can’t even access, and removes it from your device.

This is a bit of a shame. It’s not that I want to get into a cycle of drafting then editing on the Mac then back to the Freewrite, but there are times when an idea appears while I’m in front of the Mac and I would like to jot it down quickly there and then and expand it later on the Freewrite. But I can’t — yet. There have been some murmurings that it might be possible in the future to edit and roundtrip documents to and from the Freewrite through Sprinter, but it’s not there yet, and if that feature never comes along at all, I won’t be crying over it.


When my friend, colleague and now published author Thomas McMullan reviewed the Freewrite a few years ago, he concluded with this:

It is easy to write the Freewrite off as an expensive oddity, angled to nostalgic retirees and well-heeled posers, but it shows that the progress of writing technology doesn’t need to travel in a straight line. With its USB Type-C port and automatic cloud syncing, the Freewrite doesn’t ignore internet connectivity, but instead keeps it under tight control. It shows an alternate path, perhaps into a cul-de-sac, where typing doesn’t happen across 20 tabs.

Other critics gave the device a lot of stick, arguing it was a bit of a hipster toy, but I think Tom was on to something. There is a lot more to the Freewrite. Or rather: there is a lot less to it. It is devoted to one task and one task alone: hammering out a first rough draft of whatever you’re writing. I wrote the first draft of this review on the Freewrite and the experience of it was excellent. Not the experience of the device: the device just got out of the way. What was delightful was the experience of the writing itself. I don’t think I have actually enjoyed the physicality of writing quite so much for years.

In and hour and a half, I’ve written as little over 1800 words on the Freewrite and I have experienced a sense of flow while I’ve been doing it. That is well worth the money this machine cost. Now you’ll have to excuse me — I’m off to order a decent keyboard for the Mac too.


I was an early user of ChromeOS. Not as early as David Ruddock, who has written a post on how Google’s flagship desktop operating system has stalled, but as soon as the first commercially available Chromebook was out. I was in. Since then I’ve always had a Chromebook in my life, and usually it’s been a pretty high-end one. Sometimes, as was the case with the Pixelbook, they have spent quite a chunk of time as my main or even only laptop.

I’m pretty sure that I made a similar argument to David’s a few years ago. When rumours that ChromeOS and Android were going to merge, or that Android apps would come to ChromeOS, I was not only sceptical, but actually antagonistic towards the idea. Android apps would mean less focus on web apps, and that means less focus on what ChromeOS is really, really good at: the web.

I think, though, that David’s piece doesn’t really focus on what Google was trying to achieve with ChromeOS and so he misses the mark. To understand that, you need to look at what people actually use computers for now.

First of all, it’s important to split how people use laptops between work and home. David says:

“I say this even as one of the few people who can do 95% of my job on a Chromebook: that 5%, when you really, really need it, is more than enough reason to avoid a platform entirely. And for many others, it’s much more than 5%: it’s their entire workflow.”

Actually, probably 90% of workers who use a laptop can do their jobs on a Chromebook. If your life revolves around office applications – and that’s most people who use a computer for work – then web apps are not only fine, in many cases they are the only option on a laptop. The Microsoft Office suite is a first-class citizen on the web, and some of Microsoft’s PWA’s are excellent. I’ve known Office 365 deployments where people don’t even bother to have the desktop apps installed. Salesforce… does it even have a Windows or Mac desktop version? It’s all about the web.

In that environment, ChromeOS is fine – and it has been for years.

Of course, there are plenty of people, typically in the creative industries, who require applications that either don’t exist on the web or where the web apps just aren’t good enough. Because journalists work in this area they often think everyone does.

But as Benedict Evans has repeatedly pointed out, there are seven million Adobe Creative Cloud subscribers out of the 1.5bn PCs in the world. That’s less than 0.5% of all computer users. Adobe isn’t the be-all and end-all of creative software or of software you can’t do with web apps, but even being generous it’s hard to make the case that the total number of PC users who can’t live in a world of web apps gets to more than between 2-5% of users.

Home users are a little different. Yes, there are home users who fall into that nebulous category of “prosumer” – the kind of people who do “proper” photography or video editing, and want/need Photoshop or Premiere/Final Cut to do that. But again, they’re rare. The majority of the world’s photographs and videos are created, edited, published and viewed on smartphones, never touching a PC at all.

The one thing that lots of home users do with laptops which Chromebooks struggle with is games, and even here the majority of casual gaming is phone based. This is where having Android app support for ChromeOS makes sense: being able to play the huge range of casual and less casual mobile games on a ChromeOS laptop would be awesome. Sadly though – and this is where David’s piece is correct – to do it properly you need the developers to have optimised for larger screens, and to put it bluntly most simply haven’t bothered.

Remember I said that for web apps, “ChromeOS is fine – and has been for years”. That’s actually Google’s biggest challenge. In the space of nine years, ChromeOS has run slap into the same challenge that it took Windows and MacOS three decades to get to: there really aren’t a lot of improvements to make. Of course, ChromeOS needs to keep pace with web APIs and – in true Google style – push them forward. And there are definitely areas for improvement, like biometric support (as David highlights). You can improve the interface, as all OS’ should.

But what else do you want ChromeOS to do, other than be the best platform for running the web apps which 90% of users care about?

This post was written and posted on an iPad Pro, another device where you can’t do any creative work and that isn’t really suitable for professionals.

You’ve got to know when to fold ’em

Lots to agree with in Owen Williams piece on “Foldable phones: right tech, wrong place?”, starting with this:

If I consider it for a while, the largest advantage would be to merge the tablet category with the phone, but it’s a confusing proposal: do consumers desire a phone and tablet that will require charging even more often? Perhaps foldables are like adding a second monitor to your desktop: useful for power users to get a bit more space, but most people don’t care.

There’s two other problems with foldable phones of the kind we’re seeing at the moment. First, when unfolded they are too wide to thumb type on, but too narrow to have a meaningful keyboard. Second, the most popular size of tablet is around 10-11in: halve that and you have a very wide or very narrow phone, depending on which axis you put the fold.

Foldables actually make more sense applied to larger devices, such as the iPad Pro or something more complex along the lines of the Surface Book. While I never really wished my phone was a tablet, I’ve certainly mulled the idea of a tablet and laptop combined, adaptable to the situation I’m in, without the awkward detached keyboard kicking around somewhere.

Surface Book, for example, could simply be a single ultra-thin device that folds into itself to become a slate, rather than detaching. The lower half, in ‘laptop’ mode could transform into a simple display-based keyboard, and adapt dynamically to the form-factor it finds itself in at any moment.

I’m not sure about this. Part of the advantage of tablet form devices is that you don’t have to have the additional weight and bulk of the keyboard. Surface Book works because if you take off the base, it weighs less than half as much, making it easy to use when held. Surface Pro and iPad Pro work because you can remove the keyboard altogether and still have a usable slate.

In theory I would love an iPad Pro that folded in half to reduce its size. In practice, that would more than double the thickness once you account for the inevitable hinge mechanism. Folding devices are just bulky devices, and who wants that?

Most of the tech news sites are going absolutely crazy over folding devices. I just don’t think they have any kind of use case yet. Perhaps they’ll find one, but I can think of a million technologies in the past which never made it past this stage.