What should a Thing API look like

by David Janes

In the last 15 years, technologists have converged on a simple but powerful model for writing APIs: HTTP, REST, and JSON. The net result is that developers are usually able to pick up new APIs and start getting useful results within minutes.

If we were to use the “API Model” for providing access to Things, what would this look like? This talk walks through real-world examples of what a Thing API should look like, using exactly the same model you would to access any other Internet resource. It outlines some common practical concerns, such as what IDs should be used for Things, how do you differentiate between the “input state” and the “output state” of “real” Things, and how semantics can be added to enable interoperability.




We are very sorry, but due to some technical problems we currently couldn't offer a video for this talk.