The other day I was trying to use my iPad to do some work and realized that WWDC had come and gone without any mention of a single feature that would actually make it easier for me to do what most people would consider real work on it, and started typing a few notes.
Somehow, that balooned into a full-blown rant…
TL;DR: I Miss Netbooks
I miss netbooks–or rather, the entire category of small devices with physical keyboards, sub-10” screens, and a “normal” operating system. Tablets and e-readers are starting to blend, but productive computing still requires functionality that just isn’t available as in a mainstream ultraportable form factor.
The Status Quo of Small Computing
Even though I actually started drafting this on an iPad Mini’s onscreen keyboard and have been using an iPad as my main (personal) machine for over a decade, there are a lot of things I still miss.
Some of which I can hack around, but most of them are a sad reflection of the state of computing these days:
- I miss having a proper file system with seamless syncing across my devices. The iPad has forced me to fragment my documents, notes, and code across at least three different syncing mechanisms (iA Writer + iCloud,
git
, and SyncThing) and still doesn’t work reliably with third-party storage providers (Dropbox, OneDrive, you name it – they all mysteriously fail after a while). - Largely because of the above, document-oriented work is still a challenge. Office has to take great pains to use direct connections to cloud storage (which introduces its own quirks), and I can’t really use Apple’s iWork suite for anything “professional”, although I must grant the experience is slightly better.
- Developing for non-Apple systems on an iPad is still almost impossible unless, like me, you’re willing (or insane enough) to use sandboxes like a-Shell or iSH (which are brilliant in their own right, but completely hamstrung by Apple’s runtime restrictions).
- Developing or working via an iPad and using it as a terminal or thin client is actually completely feasible (and a large part of what I end up doing), but you have to have a Linux or Windows machine to remote into.
The experience can be amazingly good (webcam and audio redirection work great, for instance, so I can travel and attend meetings via a corporate virtual desktop in a pinch), but 99.9% of ordinary people will never even attempt any of it.
The situation on Android is… technically better (you can do a lot more locally on an Android tablet if you can use Termux, including building perfectly ordinary ARM binaries you can just upload to a server and run), but the overall quality of applications is just plain worse–as an example, the Android Remote Desktop clients I’ve used are just horrible at handling input, let alone audio or video redirection.
The hardware options are not that great either, since Google has stopped being an effective steward of their platform a decade ago. So there is a privacy and proprietary UX trade-off involved in using any high-end Android device.
But hey, at least I can have a (mostly) normal filesystem and build binaries locally. But the sandboxing paradigm is becoming closer to Apple’s, and I expect that to get worse over time.
I miss the 12” MacBook Air (the form factor, not the performance) and would love to have a slightly smaller laptop with the footprint of an iPad Pro but able to run some form of Linux (or–gasp–Windows on ARM) and with a decent, non-kludgy keyboard and trackpad.
I know we have the technology–after all, I still have an old Chromebook, and every time I pick it up I wonder at how far we’ve strayed.
Becoming a Computing Extremophile
My workflows over the years have pretty much converged into a consistent pattern that spans three different contexts: work, consulting (I freelance now and then, mostly in the security startup and technology advisory spaces), and personal projects. The first two are almost indistinguishable (they simply happen on different machines) and rely on either a high-powered laptop or a Mac Mini M2 Pro–effectively a desktop machine.
Everything else (including occasional spillover into non-working hours and weekends) tends to begin on an iPad and escalate into other machines as needed.
Since I no longer travel to and from an office (mine or a customer’s) on a daily basis, mobility is not so much a concern as focus 1, and I know myself well enough to know I do some of my best thinking and creative work someplace/sometime quiet and comfortable (like long evenings on a couch).
Work
All my work machines (local or Remote Desktop environments) boot up to exactly two applications: Microsoft Edge and Microsoft Teams2.
I can do 90% of my typical workday without using anything else, and things typically escalate along two different paths:
- I need to do cosmetic tweaks to Office documents or open some particularly sensitive files (which I’ll just click to open in the native app). I also use Xmind extensively, so that’s usually the third app I open (when available).
- I need to do some infrastructure, coding, or analytics work that requires either a terminal or Visual Studio Code3.
The most obvious reason I can’t do any of the above directly on an iPad is that Safari is not a match for a Chromium browser (and many people will just cotton on to that), but there are many more: management policies, corporate security, and the lack of a UNIX userland (even a sandboxed one) that comes anywhere close to what I can get out of the Windows Subsystem for Linux.
The only relevant bit that truly differs when I do advisory or consulting work is that I’ll be using my own hardware and using Zoom instead of Teams, but all the rest is largely the same.
Personal Projects
The pattern is pretty similar: I start out with Safari and Slack (because that’s where most of my friends are), use Mail.app (still, because despite lacking things like properly working contextual search, it’s the only e-mail client, anywhere, that doesn’t get in my way), and build from there:
- I’ll pop open Reminders to see what I need to do for the day (so yes, I’d really like to see that properly baked into Calendar).
- Where it regards writing, I’ll draft mostly everything in iA Writer, refine it in Visual Studio Code, and then file it away or post it here.
- CAD, electronics, 3D printing, or any other sort of creative stuff will get done using native apps.
- There’s always a terminal window open because most of my personal projects involve coding of some sort (usually in weird runtimes)–again, I’ll use
vim
or VS Code depending on the kind of project4.
I can reproduce most of this workflow (except CAD and electronics) on an iPad, and from pretty much anywhere using apps like Blink Shell, but all the iPad coding experiences (except in vim
) are sub-par.
Workable, usable, sometimes even surprisingly productive, but ultimately not a match for my taking out my Lenovo Flex, booting Fedora, and just doing it locally on the same CPU that is driving my screen… like an animal?
Content Consumption
I might as well get this out of the way, since it’s something that people like to pigeonhole iPads into:
Yes, I consume a lot of content on my iPad–especially the mini, which is still the first thing I reach for in the morning and the last thing I put down at night–but I also do a lot of reading on my Kindle, and am not oblivious to the fact that there are now such things as e-readers with color screens and actually usable web browsers, let alone RSS readers that will match the (excellent) experience I have with Reeder on the iPad every day, so I’m not that attached to the iPad as a content consumption device.
Honestly, I can see an e-reader in my future because, upon reflection if I step back and accept that I’ve spent nearly 15 years trying to use the iPad for things I could possibly have done better or easier on a small ultralight laptop, I might as well switch to a device that is optimized for focused reading and browsing–and if it can do note-taking (and occasionally some drafts with a Bluetooth keyboard), all the better.
Options
The way I see it, if I want to make my life a lot simpler, I would probably be best served by a small, low-power laptop running Fedora (or your favorite Linux flavor), but the entire “low power/low cost” segment is crawling with Chromebooks or woefully underpowered Windows laptops with ancient chipsets, and it will take a long time for mainstream manufacturers to either move wholesale onto N100-like chipsets or any form whatsoever of ARM chip that can last a day without breaking the bank.
Amazingly, the iPad may remain the best hardware option for many years, which, despite my love for the device, is disappointing as Apple still treats it as fancy hardware for children.
I still take notes and write on an iPhone–another part of this draft was written during a commute–but the point here is that it’s not the norm. ↩︎
I stopped using the Windows Outlook client many years ago due to its incredibly slow performance and just plain bad UX and will never willingly go back (although I still use the Mac one for advisory work, just because I can manage multiple calendars better). ↩︎
There is a gradient here, because Azure Cloud Shell or GitHub Workspaces are often enough, and most of the high-end analytics work I do in Spark or Fabric always starts in an online notebook of some sort. ↩︎
I have of late increasingly been remoting to
borg
or any other of my hosted sandboxes for AI/ML/coding work, simply because it’s too much of a hassle to run Docker locally. ↩︎