King Ludd # 11 : Was the Chumby Ahead of the Curve?

Tweet

You probably haven’t heard of the Chumby, and that’s kind of the point. It’s rapidly slipping off the radar of interesting gadgetry in the 21st century, but for a while there it was one of the most curious shining lights. The Chumby was a range of devices, with an early version appearing in mid 2006, in a world before the iPhone and before the maker scene had really kicked off. A Chumby was essentially a small touch screen attached to a small computer with a wireless network connection and some methods for hardware hackers to attach stuff to it. Generally they looked like techno-beanbags.

The creator of the Chumby is now legendary hardware hacker Andrew “Bunnie” Huang (currently with a curious open source laptop kickstarter campaign), and it reflects a few of his priorities, most notably with the open availability of schematics and PCB layouts. Different models with improved screens, CPUs and so on were made, but it was also more widely commercialised by Sony as the Sony Dash, which substituted some of the hackability for more conventional packaging. My direct experience with the Chumby platform is largely the result of having picked up a couple of the Sony Dash devices years ago when they were being sold off cheaply, presumably from having failed to make a dent in the market.

The hardware is what you’d expect: mid/high range ARM processors for the era (improving with subsequent models) and 320×240 screens giving way to 800×600 single touch point screens (this being pre-multitouch explosion). The software was also fairly predictably Linux based, but more curiously used the now fashionable to hate but not completely awful Flash Lite graphics system. Flash Lite was a subset of the more famous Flash, and created using the same tools, but was intended for use on small embedded systems, mainly phones or PDAs.

From the user point of view you would configure various “widgets” on the device, and the device would simply display these on rotation, or allow you to flick through them. A widget in this context is really a mini Flash Lite app, which can do things such as connect to the network, play sound, receive touch events, show pictures and so on. Widgets I actually used included ones to show particular flickr feeds, Facebook photos, webcams from famous places, and generic RSS feed displays. Even with all that 99% of the time it would fall back to being an alarm clock that synced the time from the network and displays the time in a very nice high resolution Helvetica Neue.

Sound familiar? This is almost exactly the value proposition of most smartwatches, with relatively few important details changed. Firstly a watch is portable, and on your wrist, while Chumby devices tended to be plugged in. Secondly with smartphones around we’ve got a lot more in the way of notifications happening, so while the Chumby was a “pull” device a lot more of the content to be on smartwatches will have been “pushed” there.

Sadly the Chumby failed, and unless it is resurrected as a smartwatch the platform does look fairly dead. One particularly nasty aspect of this is the Chumby was largely configured by a website hosted by the company formed to sell the device, and as the company went down so too did the servers, resulting in the devices losing functionality. This should be a valuable lesson to all prospective consumers moving into the wearables and connected devices markets, as if the manufacturer of your device goes under you could be left with an expensive worthless composite of unrecyclable rare earth metals.

Many companies now attacking the wearables sector see themselves as service companies first, and device makers second, meaning the devices are being made as ways to access the service. Chumby didn’t really know what it was, and I suspect thanks to the background in hardware hacking overemphasised that aspect of the business, leading to a lot of orphaned devices.

In a world where we’re looking at things like network controlled thermostats having the consumer face the risk of being left hanging with a useless device after only a few years seems unreasonable. Either the devices will end up being subsidised at the point of delivery (like how cellphones are “cheaper” on contracts) or consumers will need to be satisfied that if the company behind the device goes bankrupt their device will remain useful. Predictably all the companies involved will believe they will be the ones to get to critical mass and ensure their longevity, but those that actually make it are just going to be the first to convince consumers of it, regardless of the merits of what they’re doing.

Nigel Birkenshaw runs atomirex.