Just before Christmas, Canonical founder Mark Shuttleworth wrote about a new area of expansion for the firm’s Ubuntu operating system. He described it was a “snappy” version of Ubuntu and Christened it Core. And it’s designed to power the Internet of Things (IoT).
Today, at Mobile World Congress in Barcelona we got the first proper outing for Core – and it’s a lot more finished than frankly I expected.
The basis of Core, says VP for IoT Maarten Ectors, is that its essentially a lightweight version of Ubuntu Server with an easy-to-use web interface. There’s a bunch of images available, each designed for specific hardware like the Raspberry Pi, Beagleboard and so on.
So far, so wonderful for providing hackers building homebrew projects with a desktop-like environment to use them in. Ectors demoed a Beagleboard hooked up (via Bluetooth) to a fitness band. The band’s sensor was being used to determine the unique characteristics of his own heartbeat, turning it into a biometric doorpass (pictured above). Easy to set up using Core, apparently.
Where Ubuntu Core gets interesting, however, is where it’s closer to taking on the duties of a traditional server. It supports a (limited, at the moment) app store-like method for installing packages and maintenance, not unlike Microsoft’s Azure platform. Using this, Ectors showed a standard network switch running Ubuntu Core, which could be quickly configured to add extra features like a firewall, an F5 instance and even a control app for a robot arm in a mouse click. Snappy – the delivery mechanism – keeps everything automatically updated and theoretically safe too.
The most impressive demo involved a mobile base station running Ubuntu Core. This, explained Ectors, could be instantly provisioned with a Netflix app, say, for locally caching videos and reducing bandwidth consumed. Ectors argues that this kind of versatility is going to be essential for mobile operators.
“No-one in the world is ready for the amount of traffic that’s going to be generated by 4K video cameras in home security systems,” he says, explaining that an app which can filter traffic and quickly provisioned would be a cinch using Snappy. By increasing the capabilities of switches and base stations, the increased load of thousands of sensors distributed around a city – he uses the example of data from hundreds of smart parking bays – is that bit easier to cope with.
Which makes Core a strange but compelling prospect. Big enough for a network operator to trust, easy enough for a homebrew project on the Raspberry Pi.
An intriguing mix.