Linda is a coordination model using a shared tuple space where processes write and read tuples matching patterns, decoupling producers from consumers in both space and time. Unlike CSP, there is no direct addressing; all interaction goes through the shared associative memory.
Linda provides generative communication: produced data persists until consumed.
This example shows tuple-space operations and pattern matching:
out("task", 42) // write tuple to shared space
in("task", ?x) // remove matching tuple, bind x=42, block if none
rd("task", ?x) // read tuple without removing it