Hi everyone… My access to the “Forums” for replies was just restored. I could read all your notes, but some bug was not letting me log-in. I love discussing the risks of the Cloud approach and have been anxious to chip in again…
Well: I’m very glad to see that others have the same concerns about the SmartCloud. Some others of us in this Developer community may be writing about this under other Topic threads as well.
More importantly, I hope that people from Physical Graph Corp. / SmartThings are taking note and discussing out concerns internally as they work on next phases and think of the mid and long-term strategy.
Out of the gate, SmartThings has done some stuff really well. Basic activation, SmartCloud connection and adding officially supported Things (provided you have iOS!) is quick and easy. The REST-API is pretty sensible at first glance. The IDE has received kudos. The iOS App is reasonable intuitive and graphically pleasing. And more.
But there are plenty of design flaws that so far have not received much attention. For example, there is no way to add Things to overlapping “Sets” (not “Groups” which are limited in scope…). To me, some concept of “Sets” in the Object Model is necessary to facilitate the installation of “Scene” Apps (multiple lights and other Things that react together: The Set of Home Theater Things, The Set of Security Things: And note that certain lights in the Home Theater may be used for both home theater mood and security lighting! Well… We could spend hours hashing out Object Model changes. Ironically, the ability to define Scenes is present in the most basic of Z-Wave Controllers (handheld programmable remotes from GE/Jasco, for example). This irony makes me worried about the “thinking” at SmartThings…
<ramble>
Back to the topic at hand: You can read my previous and scattered posts regarding risks of the SmartCloud requirement.
Yet I will take this opportunity to emphasize that I think the greatest risk is the permanent failure or dissolution of the SmartCloud due to bankruptcy or buyout. Periodic failures are annoying (e.g., Netflix down, gmail down, Amazon Web Services outages), but they are nothing compared to the complete loss of all your work automating your home, configuring and/or coding Apps, etc…
We all know of the substantial number of products/services dropped by Google over the years. We all know the current concerns as to what will happen to Tubmlr and Instagram now that they have been purchased. We know of the restrictions Twitter (and others) have been placing on third-party users of their API.
There are competitive enterprises on the horizon (The Ubi, Ninja Blocks, IP based devices, …?) that are substantially more Open in their DNA. Or even 100% “open” (Ninja Blocks say “hack-me” and have published hardware and software, right?).
Perhaps we can take solace in knowing that as SmartThings “lack of openness” becomes evident to more and more designers; the competition will feel more confident in their approaches. Openness, however, has both functionality risks (loss of control of complexity, perhaps?) and business risks (no IP protection?). SmartThings is certainly more open than Crestron, ELAN, RTI, and others, no? A step in the right direction…?
</ramble>