If you own an Android-based smartwatch, you’ll know that, up until very recently, there was only one way of getting an app onto your watch: installing an app on your smartphone and waiting for your wrist to vibrate! But that will soon be over.
In addition to some usability improvements and some massive changes to the user interface, the update to Android Wear 2.0 has some fundamental changes for smartwatch app developers in store. Until now, an Android Wear app could only be installed on the watch with the help of a companion smartphone app. Basically, what you had to do was pack the Wear app into the smartphone app.
After installing the app on the smartphone, the Wear app was installed on the watch at some point immediately. The user could not decide or influence whether that installation was taking place. As Ian lake writes inside a thread in the Android Wear developers community, users showed little enthusiasm for this feature. Which, in my opinion, is hardly surprising. Scrolling through an endless list of apps (of which only a select few, if any, are used regularly) on the very small display of your smartwatch is no fun. But that was the system in place and you needed to use it in order to maintain a working Android Wear app. Network communication with the watch was possible under certain circumstances, but in order to reliably implement the decided functions, a companion smartphone app was definitely necessary.
So, if you can no longer install apps on your watch via a smartphone app, how does it work? Manually, of course. You’ve been doing it on your smartphone all along using the PlayStore. Since the Android Wear 2.0 developer preview version 3, the PlayStore can also be accessed from the watch and is the first and only port of call for all your app-related needs.
Therefore, smartwatch apps can now be installed completely independently from any companion smartphone apps – they are real standalone apps. So you no longer need an app within an app in order to get your Android Wear 2.0 app into the PlayStore. The existing Multiple-APK-Support, which developers can use to provide multiple installation files for various different devices, is also to be used in this case. In order to really make the app completely independent from your smartphone, the watch must now support various features. And, of course, the update to Android Wear 2.0 offers that very support.
A very important and highly welcome change now enables network communication on the watch: WIFI enabled watches can now also access the internet via the smartphone during an active Bluetooth connection (which was not possible until now). And if the watch does not support an active WIFI connection, it automatically uses the smartphone’s mobile internet or WIFI connection. No extra line of code needed! The frameworks you already use for network communication within smartphone apps, can now be re-used for the Wear app. An additional Android app on the smartphone is no longer necessary for forwarding the requests. Which means (drumroll, please) that the watch can also be connected to an iPhone!
Another feature which is ubiquitous in the app-development world is push notifications. This means that notifications are sent from a server directly to the users’ devices. Google’s quasi-standard was previously called “Google Cloud Messaging (GCM)” and is now known as “Firebase Cloud Messaging (FCM)”. Forwarding push notifications from the smartphone to the watch is, of course, already possible. But the standalone functionality of Android Wear 2.0 apps makes it possible to send notifications directly to the watch via FCM. You don’t even need to be carrying your smartphone for it to work.
Now, all of this gives you the possibility of developing completely independent Android Wear apps. But why would you feel the need to do that? From a user’s perspective, the probably most obvious reason is this: watch apps now become usable and useful, even if the smartphone is not at hand. And also: the Wear app works independently from the smartphone’s operating system.
The developers’ point of view might sound a little less exciting: if there is an existing smartphone app, developing an additional Wear app is a much smaller step up than before. Existing components (network communication, data handling, etc.) can be re-used for the watch app without needing to change very much. A little bit of UI work and you’re good to go. How much can that cost?
So we get rid of the companion smartphone app? Not necessarily. There’s no unified answer, as it strongly depends on the kind of app you’re looking to make. A simple watchface app with only few setting variables probably won’t need a companion app, while more complex apps could basically be unusable without one.
There’s also a slight downside: since both apps can now be installed completely independently, developers or publishers will now have to make users aware of the fact that they also need to install the app on the respective other platform.
But, in any case, smartphone apps whose only purpose is delivering a Wear app, as well as Wear apps that serve no purpose without a companion app will have a hard time of staying on the users’ devices for long.