• CRDTs: conflict-free replicated data types provide update operations that are designed for merging when devices synchronise

  • DDD: domain-driven design is a method for naming the business entities and their operation such that domain experts (not programmers) can understand them; this leads to better specifications, architecture, and implementation

  • distributed event-log: each node in the system writes events to their own section of an overall fully distributed log; events can be freely replicated between nodes (since they are facts) and totally ordered based on node ID and logical timestamps

  • edge computing: using the CPU, memory, and persistent storage of the edge device to run business logic — the contrary of “thin client” (which only renders the UI) or “cloud gateway” (which only performs local communication tasks at the bidding of a cloud service that does the data processing)

  • event sourcing: the application does not store state, instead it stores a stream of events and computes the needed state at runtime by applying the events according to the local business logic

  • peer-to-peer: interaction between neighbouring devices using some communication technology (like wifi, bluetooth, ultra-wide band); can also be extended to cover direct interactions between edge devices via the Internet