A few weeks ago I gave a talk summarising the schedule for Google’s big developer conference Google I/O, which runs over three days in an outdoor concert venue across the road from its headquarters in Mountain View, California. I called out the tech giant for not including any content on Kotlin, a much-hyped alternative to Java for Android development written by JetBrains, the team behind IntelliJ and Android Studio.
How wrong I was.
First up was the Google Keynote with Google’s CEO Sundar Pichai, and he was clearly focussed on AI and machine learning with the statement that Google was shifting from mobile first to AI first. He also showed off the new Cloud TPU chips, which provide monstrous amounts of processing power (180 teraflops) in order to speed up machine learning systems such as voice recognition.
Pichai also announced Google Lens, which uses image processing to provide information and actions relating to the content of the image. But instead of being a standalone app like its predecessor Google Goggles, it will be rolled into Google Photos and Google Assistant.
Several relatively cool announcements followed including new features for Google Home such as proactive notifications that use the device’s lights to get your attention when timely information is available — traffic delays for example. Home will also be launched in several new countries later this year including Australia! At the end of the keynote it was announced that everyone would get a Google Home and $700 in Google Cloud credits to encourage us to build voice apps.
We also heard about the next version of Android’s OS, Android O, which will focus on two things: fluid experiences such as picture in picture, new notification styles and autofill; and vitals, aimed at conserving battery life.
Better battery life sounds awesome but one of the ways this is achieved in Android O is by restricting background location polling. This has huge implications for lots of apps that rely on real-time location data. Fortunately there were Google engineers on hand in the Office Hours tent to answer the questions we had about the impact on our applications and users.
The good news is that Google’s Awareness API is exempt from the background restrictions. When the restrictions are enforced, background location polling will occur roughly two to three times per hour.
Toward the end of the keynote Android’s Product Management Director, Stephanie Saad Cuthbertson, took to the stage and dropped the biggest bombshell of the day: Official support for Kotlin as a programming language for Android.
While this announcement made complete sense— Kotlin is much loved in the Android community and already deployed in some production apps — Google caught everyone off guard with its ability to keep the deal top secret. Several Kotlin-related sessions were immediately added to the schedule including one with open source hero Jake Wharton.
The What’s new in Android session ran through some of the more intricate changes in Android O like the ability to assign fonts in XML layouts, auto-sizing text views, and — most importantly — backwards compatible emojis!
Another announcement with huge implications for Android developers was the decision by Google to provide official, documented guidelines for Android application architecture and components to assist with the often painful tasks of lifecycle management and data persistence. The introductory session on these architecture components was packed to the rafters with people nodding and cheering as the details were laid out. We’ll get more information during the follow up architecture sessions tomorrow.
The night finished with dinner, drinks and quirky entertainment including a karaoke bike, bands and stand-up comedy. Check out more photos from day one below.
Tags: Android, Android app development, AndroidDev, Google, Google IO 2017
This post was written by Luke Simpson