Re: What is an algorithm?

Recently I read "Principles of eventual consistency" by Sebastian Burckhardt. This book is about distributed systems. Briefly speaking, the author gives examples of distributed systems, such as: complex web of clients and servers, "mobile + cloud revolution", "mobile client and server", etc.
This book is dedicated to the problem that occurs in distributed systems and which is called consistency of data. The consistency of data is related to "Replicated data". In his book, Phil Bernstein wrote that " A replicated database is distributed database in which multiple copies of some data items are stored at multiple sites."

In above mentioned book S. Burckhardt defined "History" and "States". He defined History as an event graph. For precise definition of History look at Definition 3.1 at his above-mentioned book.

I have defined history as "history of events" or as a "history of states". I use the "History of events" when I want to know which events occurred. Since the event is always changing a state of an entity, then the "History of events" is the same as the "History of states of entities or relationships." Note that this is the basic definition. For example, in Anchor modeling, it is not clear, which history they maintain.

Another term that I have mentioned here, is "states". As I wrote, I am working states of entities and relationships. These are the most general categories.
I want to say that some important concepts as "states" and "history of states" now are normally used, as most ordinary things. However these are not ordinary things.

Anyway, the above-mentioned book by S. Burckhardt is a good book about complex matters. In this book was used a good mathematical notation. I think that my solution which uses states, also would be a good solution for distributed systems.

Vladimir Odrljin Received on Tue Jun 23 2015 - 21:49:12 CEST