As you may know, there are LOTS of applications that live on the thing called a blockchain, on a decentralized network, on a distributed network. But, hat is that REALLY? Centralized, decentralized and distributed network models by Paul Baran (1964).
What!? 1964? Yep!
Take a look.
These models were made back in 1964.
PROS / CONS
Points of Failure / Maintenance : Centralized systems are easy to maintain as there is only a single point of failure. Decentralized have more but still finite. Distributed systems are the most difficult to maintain.
Fault Tolerance / Stability : Centralized can be highly unstable. Kill the leader and there will be chaos. Kill the leader for a decentralized system and you will have many decentralized systems. Distributed systems are very stable and a single failure doesn’t do much harm.
Scalability / Max Population : Centralized — low scalability, Decentralized — Moderate, Distributed — Infinite.
Ease of development / Creation : Centralized systems can be created really fast, you pick up a framework and apply it everywhere. For Decentralized and Distributed, you have to first work out the lower level details like resource sharing (trade) and communications (transport).
Evolution / Diversity : Since centralized systems follow a single framework, they don’t have diversity and evolve slowly. But for Decentralized and Distributed systems, once the basic infrastructure is in place, evolution is tremendous.
Take a minute, think about that. Oh by-the-way in Q3 2018, Lightning is coimg out, get ready!
Marco
