Plight of the Pixelbook


I’ve been using my Pixelbook (base model) for five months. Here’s a brief rundown of what’s wrong and right with it, and why I decided to get one.

My trusty MacBook Air

I’d had my old MacBook Air for seven years. That’s absolutely incredible. At the time the geniuses at the Apple store swore that I was making a bad decision, instead I should invest for more processor power, more storage, a disk drive, and the ports on a Pro model. I declined.

The reasons for wanting an Air were simple. It was sleek, it had no physical moving parts, flash storage, and I was only using it for web development which used to be a lot less intensive on the CPU. And it had a higher resolution screen.

It served me very well for seven years.

Towards the end I was just using it as a thin client to access my beast of a gaming PC. That worked well over RDP. It could also manage WebStorm and static sites.

Now I do a lot of back-end work and experimental stuff. I want to learn Go and Dart. I’m playing with Kotlin and Android development. Websites require a lot of pre-processing these days, WebPack is a bit more a beast than Grunt and Gulp.

Why I chose the Pixelbook

I make it no secret that I’m a fan of what Google do, even if it can be sporadic and fleeting. Android is a great platform, Chrome is a pleasure. My targets are firmly set on what Google is doing with its Fuschia project. In the next couple of years it’s rumoured to be a light platform for smart devices. I would love to see it take over the entire hardware stack on my phone, laptop and PC.

The Pixelbook is a bet on Google bringing its power towards developers. They’re already firmly in bed with JetBrains, using the IntelliJ platform and treating Kotlin as a first class language for Android.

I’m a big fan of the design. The keyboard and touchpad are a pleasure, the touchscreen is incredible, and it’s not shy on power.

I like software. I don’t tend to care for hardware as long as it’s out of my way and not throttling what I need to do. Google’s software and software design is awesome.

These are of course my opinions.

I’ve never fancied iOS, MacOS was and is great, but Apple have left developers behind with their direction. They no longer produce laptops for serious professionals (again, in my opinion). I am not happy with the idea of a touch bar versus the idea of a touch friendly OS.

I am happy working on Windows, and I did consider a Surface Book. If it wasn’t for betting on Google I’d have been more than happy with one.

Android support

Out of the box Android apps just work. There are obvious issues when it comes to interacting. Android is a touch first OS. Many app developers do not build with a keyboard and mouse or even a landscape UI in mind.

This would obviously be glorious for developing an Android apps though, the ability to touch and work with them virtually natively on the same device.

As a consumer it’s good for games and gives the ecosystem a little boost.

One downside is there are many services were the web experience will be better in ways. Such examples are Google’s service, Maps, Gmail, Calendars etc. The Android versions pale in comparison, but this is often the case with many apps were the mobile service is treated so differently, and understandable so.

Profile switching

I do not understand what Google expects me to do on ChromeOS when it comes to managing multiple accounts. On Windows, Chrome is great with this, a different account per window. This means I can have my personal account going at the same time as my work account, and neither will be impacted by the other.

The current expectation is to sign in to a different desktop sessions. Google does allow profile switching within a single browser session, but the support is atrocious. Drive for example, will fail if you try to switch between accounts.

I want to have different accounts for Chrome in a same sessions.

Android support falls over here, logging into another Chrome session prevents the container working in the other.

Remote Desktop

One hope I had was that whilst Chrome OS is maturing and until at least Linux desktop support comes, I would be able to remote into a sessions on a Windows machine. This does work under the official Android app, but it’s not amazing. Issues with resolution, scaling, and touchpad support make it difficult.

Once Linux support lands this will not be an issue, because I can at least remote in using an RDP app for Linux, of which there are many.

Another hope was that Chrome itself would have Microsoft RDP support, or at least there would be an app. There is in an unofficial form, but I haven’t tried it.

Once Chrome OS is mature, I won’t need another machine to depend on.

Linux support

Project Crostini enables Linux desktop support for Chrome OS, running in a container. It’s currently in the Beta channel, but that hasn’t at the time of writing been updated for the Pixelbook yet (this is personally infuriating).

I fired up the Dev channel briefly to have a play. It enables a lot for developers.

Having had that taste, running IntelliJ almost natively, having a proper environment with servers and tools, was awesome. I didn’t try running Docker containers, as it felt a bit silly to try, but I imagine Google will have some official answer to that soon.

The Dev channel was very unstable and I couldn’t manage on it. When the next beta drops, I will be able to develop on this machine.

In the meantime there’s a very good SSH app for Chrome that does wonders.

Despite using Windows in the office, I miss having a Unix environment, another advantage of the Mac. Windows does have support for its own first-class Linux containers, but I tend to prefer Docker in those instances

Gaming on a Chromebook

So far any web game runs well, but when you think of the vast ecosystem for PC and console gaming, the web doesn’t exactly scream at you. Android games are similar, there is not the quality, and the ones on the Play store don’t tend to support landscape or bigger screens, as discussed.

A few things do work. Retroarch works very well via the Android app.

As for gaming on Linux, again not a huge ecosystem, Steam does boot, but with current support lacking for hardware acceleration and even any sound, this isn’t yet viable. Give it a year, I imagine gaming for Linux support will be there.

Valve have a native Android Steam Link app, and this works, and works well with one drawback. As far as I can see, there is no support for the keyboard and touchpad, quite a large issue. I haven’t tried with a controller, but I own a physical Steam link connected to a TV and a wired connection (via powerline) to the host machine, so there isn’t much in it for me playing on my Pixelbook.

Hope for the future

We’re nearing Google’s hardware event in October, and whilst I’m more interested in their software, it isn’t unlike Google to keep to its bounds.

I would hope to hear more about Fuschia, and more about a physical Google device under Chrome with at least beta Linux support.

This year Google announced Linux support, and it’s very nearly ready for beta. There are rumours of native Windows support, probably through dual-booting, and whilst this isn’t viable on the light storage space of my model, it may be a worthy feature whilst Chrome OS matures.

For now I’m happy with a really good touch laptop, that can’t do much for a developer but makes browsing the web and consuming media an absolute joy. It feels like a middle finger to the Apple idea that a keyboard and pointer OS cannot be used for touch too. I certainly have no problems to report there.

Here’s watching Google. Let’s see what they can do for us devs.