How does Kaspa distinguish propagation delay from consensus tolerance?

Kaspa's design treats propagation delay and consensus tolerance as two fundamentally different things — one is a physical fact of the internet, the other is a tunable setting in the protocol. Propagation delay is a network phenomenon: it is simply how long it takes data to travel from one computer to another, and no software can make light move faster. Consensus tolerance is a protocol parameter — a deliberate design choice that sets how much of that network delay the consensus algorithm can safely absorb before blocks start conflicting. Understanding the difference matters because it means Kaspa's engineers could measure real-world internet slowness and then build a protocol that mathematically accounts for it, rather than assuming an unrealistically fast network.