Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Disclosure: I work at Siemens on an implementation of the Web of Things not related to the Mozilla Project.

What is mostly misunderstood about the Web of Things, is that it does not define a network protocol but a document format to describe devices in terms of properties, actions and events.

Interactions are then further described by „forms“ that contain Hrefs with the protocol scheme supported by the device (e.g. coap://, opc.tcp://). The idea is not to introduce the next unifying protocol, but to unify the description of devices that support arbitrary protocols. Think of OpenAPI for IoT.

It serves us well in industrial settings where protocols like opc-ua, bacnet, coap, modbus and the like will stay the leading protocol in their respective domain for the foreseeable future. Such a descriptive approach to me seems the only option to at least write applications in a uniform manner.

Looking at efforts like MS Device Description Language (DTDL) or AWS Things Graph Data Model (TDM), it makes sense to have at least one openly standardized description language to prevent fragmentation at that level. The war on fragmentation at the protocol level is already lost IMHO (and everybody lost).



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: