Ubuntu Touch – Early Impressions

I recently decided to flash my Nexus 7 from Android to Ubuntu Touch to see how the new mobile O/S is shaping up.

I followed this WIKI in order to get it going. Apart from getting no prompt regarding a “host key”, it worked without a hitch. However, upon reboot, it was clear that I’d chosen a bad time to try the preview – very little actually worked and many of the core apps wouldn’t even launch. When I reflashed the device a second time, I chose the “Stable” option, instead of the “Devel” option and had a (slightly) better experience. Many features are still broken, but at least most worked. Enough to get a feel for the experience.

But I have to start this article with a rant.

Ubuntu One Integration (lack thereof)

When I started using Android about a year after it came out, it wasn’t very good in a lot of ways. I used it mainly because it wasn’t made by Apple and it was “open source” (argue all you want about how it’s developed, it’s still open source).

Since then, many great features have been added – smoother interfaces, Holo theme, backup/restore, cloud saving game progress and various other tweaks.

The problem with nearly all of these advancements is that they were introduced so far into the Android lifecycle that it had already created hundreds of thousands of apps in its Play Store and those apps were written before many of these features were available! Now the big names, sure, they’ll often update their apps to take advantage of the new features. But still hundreds of thousands of others won’t.

So if the foundation hasn’t been laid BEFORE the platform becomes available for developers, it’s going to be “death by painful iteration” to play catch up. Perhaps Ubuntu Touch won’t suffer from this as badly, as an update to UT simply requires you update the software like you would your desktop.

But I’m not convinced that this is the right way to do things. There was an opportunity here to address the key areas that plague Android and give Android users a solid reason to try something new :

  1. U1 integration. Similar to Google Sign-In, but take it further. I want my devices listed on U1, with their current status (on/off), location, software revision and installed apps.
  2. Backup/Restore. I want to sign into a brand new phone that I’ve never used before and be given the option to restore it to a state exactly mirroring the state of my previous phone. And I mean “exactly” – game save progress, app preferences, everything. Google’s attempt to do so is pitiful – offering to re-install apps and remembering WIFI passwords. Apple’s iCloud is a better option, but fails to include app preference or game progress in many popular titles. There’s no reason why this can’t be done with the free 5Gb Ubuntu One cloud storage and it would generate a heap of new customers if the next item in my list were available…
  3. Cloud save. I want the pictures I take on my UT phone to sync to my U1 cloud. Videos too. I want my listed apps on there, ideally with one-click delivery like Google Play. But I also want my game progress saved up there – saves, high-scores, the lot. That might encourage me to actually play games on my phone – I gave up when I poured about 20 hours into a tower defence game on my first Android, then realised that it was gone forever when I got a new tablet.
  4. Two way Sync. I want the music I buy on Ubuntu Music to become instantly available on my UT device.

And those are just my gripes with Android! There must be a thousand other advantages to hooking your phone to the cloud from the start. And delivering APIs around that experience so that every developer uses them, from day one.

Right on to the rest of the article which focusses on the look and feel of what’s there today.

Missing Features

The first problem I ran into is keeping track of what’s open! Swiping left from the right edge doesn’t show cards, but instead “Alt-Tabs” through open apps. But without an indication of when you’ve reached the end of the list, you can keep swiping left until you’re blue in the face! Additionally, the launcher (swipe right) shows you the app you’re in, but unlike the desktop edition, doesn’t manage to indicate what’s running on the other icons. Even if you use the launcher to start three of four apps, it will only show you the one you’re in when you reveal it by swiping right.

I finally managed to run into the “App” scope and noticed the list of running apps at the top of that scope. Closing them is tedious, as you have to long-press first, but it seem robust.

The next gripe I ran into was the keyboard. I found it lacking in basic features :

  • A way to exit the keyboard (discovered swipe down eventually)
  • Keyboard auto-suggestions, auto-fill.
  • Long press on top row to chose numbers

But also it’s a bit laggy. Try typing quickly and this will inevitably lose characters. Combined with the lack of auto-correction, this made any kind of data input pretty tedious.

Home Screen customisation

I could not find any way to manipulate the home screen at all. I had to do everything from the Apps scope instead. Indeed, not only could I not customise it, but I found that many apps wouldn’t run when I clicked on them from the home scope, but they would run if I did so from the Apps scope.

I eventually found that I could long press on launcher items to add/remove them there, but this was made possible, in my opinion, by my desktop experience. I doubt that new users would figure this out – preferring no doubt to long-press on the icon in the app scope (which does nothing) instead of launching it first, then pinning it by long-pressing on the launcher icon.


Finally, these are just things that don’t feel right to me. I have no doubt that many of these are personal, and many others will be tidied up before release, or in other releases as time progresses.

  • Wooly scrolling on web pages
  • Doesn’t change portrait/landscape (but did briefly immediately after flashing)
  • No indication of shut down sequence
  • No security options.
  • What are the scope filters for? Enabling/Disabling has no effect.
  • No distinctions between a native app and a web app.
  • No feedback of launching an app (and nothing in the launcher to indicate that it’s running).
  • No feedback on a successfully joined WIFI network. Actually no feedback on WIFI at all, other than a tick box on/off and a password prompt. Did it work? Is it still connecting? Who knows.
  • No back button. Some apps have a back button when you swipe up, but web apps don’t, so if it’s a poorly designed web app, it’s easy to become “trapped” in a dead end screen and your only recourse is to swipe right, go to the Apps screen, long press the running app, kill it, then launch it again (and remember to not end up in that dead end again).
  • No native messaging client (like Hangouts or iMessage).

So will it improve in time to make a good first impression? In all likelihood, no. But does it need to? Releasing beautifully perfect software has rarely been Canonical’s modus operandi. They didn’t wait for perfection with Unity before “inflicting” it upon 11.04 users. Today, Unity is an absolute joy to use and continues to improve from release to release. Canonical is all about “getting it out there” so it can be prodded, poked, bug fixed, then improved slowly and steadily.

Sadly, as I note in my opening paragraph, what’s worked on the desktop may not work on the mobile. The mobile ecosystem has targeted apps which need a far clearer understanding (assumptions even) of the underlying operating system. So if you release 75% of an operating system, there’s less incentive for developers to make beautiful and integrated apps. Or worse, the developer will invent their own way of doing the missing 25%, leading to horrible fragmentation between apps. How do you share on Ubuntu Touch? In Apple land, sharing is a bagatel. In Android, each app you install will set up sharing “helpers”, a core part of the Android O/S. Ubuntu Touch runs the risk of going the clunky Apple route here, and that’s just one example of where the unknown 25% will manifest.

That’s assuming they attract developers at all. We have two massive mobile ecosystems already, we arguable don’t need another. But the recent App Showdown saw some nice contributions, and even if the Ubuntu Touch store isn’t hitting hundreds of thousands in two years time, I’ll settle for using it if it just covers the basics.

I’m very hopeful it will, because its promise, its potential, is huge. But that missing 25% is incredibly worrying. I sincerely hope there’s a big “Shuttleworth Reveal” before release that appease my concerns.