I’ve just installed a doorbell. “Well, whoop-de-do” I hear you mutter. But bear with me – there’s a sort of a point to this.

It’s a Thing – a Skybell – and it’s connected to the Internet, ostensibly to potentially do useful stuff. Does that though automatically make it part of the buzz of the twenty-teens, the Internet of Things?

Functionally, it should ring on any and all of our iOS devices. This is good, as Weasel Towers is fairly sizeable and we can’t guarantee to hear a standard doorbell. We also might be at a neighbour’s or out walking locally. We are however likely to be within earshot of an iDevice at most times. Skybell also includes a video link and intercom that can also be activated remotely, so we can, should we be so inclined, either interrogate visitors before getting off our backsides to answer the door or talk remotely to confused couriers about where they should deposit our latest bulk order of kitty treats.

I did use the term “potentially” there, in conjunction with “useful stuff” – the Skybell is in fact a rather bad internet of anything device, in that it’s part of a closed and vertical ecosystem of nothing-else, it relies entirely on a connection to its maker’s servers and, above all, it doesn’t really do the job it’s intended for: It is slow and it is random – yes, it will ring our devices, but anything from ten seconds to several hours after the button has been pushed. The video link takes up to a minute to establish, even from within the house and the intercom would be greeted as a long-lost sibling by any dyspraxic Dalek.

Now, we do have a local issue that clearly falls outside Skybell’s design considerations: thanks to the abysmal communications infrastructure of the rural UK, we’re on a satellite connection, so a ring on our doorbell is relayed through our home network, then 72,000km+ to the Skybell servers which, if they deign to respond (which they don’t always do), then send the signal back another 72,000km+ to our network, which then alerts the Apps on our phones and tablets. Or not. If I then choose to activate the video feed to see who’s at the door, the process is repeated, this time at video bandwidth.

All that however, shouldn’t account for more than 3–4 seconds latency, which does not explain the functional inconsistency of the Thing – most often, by the time we get to the door or have established the video link, our caller has either expired of hypothermia or has emigrated to a new and fulfilling life in the off-world colonies. And, if we’re out and about, local mobile data networks being the sad joke that they are, we may get notified at any point in the next day or two, or not at all.

There is also the small issue of cognitive dissonance, especially with couriers or besieging hordes of peasantry: most people have the hang of door intercoms, but not of door intercoms that are currently being answered from the other side of the planet. By a dyspraxic Dalek.

So far then, we’ve seen a good idea compromised by the design and performance of its service infrastructure. But the killer is that, as household device, it simply doesn’t interoperate with anything else. At the very least, we’d look for integration with any current internet-enabled door lock and thereafter, with security systems, CCTV and anything else we care to throw at it. But not a bit of it: Skybell offer neither direct integration nor an API (Application Programming Interface) to allow any third party to integrate it with other connected devices.

And this is something that a large proportion of vendors (ironically, usually start-ups) offering connected devices, from fit bands to kettles, don’t get: that the functionality and attraction of their offerings are massively enhanced by allowing them to casually interconnect with whatever else is out there, regardless of whether they’ve thought of a use case for it: someone will and people will then buy more of their stuff because of it.

Which is what the Internet of Things is about for me – the casual ability to mash up and build value from multiple devices in arbitrary combinations, whether explicitly enabled by developers, services such as IFTTT and SmartThings and emerging standards such as OIC, Thread, Hypercat or Alljoyn (in fact, we’ll soon need a metastandard for the interconnection of IoT standards…) or through new systems for automated self-discovery and association (Small disclaimer here: I’m co-founder of a startup focussed on the latter).

I do wonder just how long it will be before vendors wake up to the existential advantages of opening up their devices to the world. Until they do, we’re left with lots of little Internets of Thing and not a true Internet of Things. Which is deeply boring, arbitrarily restrictive and simply not what the world needs to be about.