Erlang Master Class 2: Video 3 - Handling errors

These Master Classes will show you how Erlang can be used in practice to solve larger problems. The examples provide 'capstones' for different aspects of Erlang: functional programming, concurrent programming and larger-scale programming with OTP.

Erlang is best known for its “share nothing” concurrency model that supports separate lightweight processes that communicate by message passing. This master class shows how sequential code – from the expression example – can be made concurrent. Building on this is a discussion of patterns for systems built on communicating processes. Erlang is built to be fault tolerant, and its “let if fail” philosophy is supported by its approach to process errors. Key to the successful use of Erlang in practice is the recognition that many of the components that programmers naturally build – for example servers – follow a similar pattern, and this is illustrated by making a more generic version of the calculator, serving as a bridge to the final master class.

Joe Armstrong is one of the “gang of three” Erlang inventors, and a champion of the Erlang approach to programming large complex systems. He has worked for Ericsson, as well as in startups, and is also professor at KTH, Stockholm. He is the author of Programming Erlang (2ed).

[Qaboom]

Erlang: the only programming language that acknowledges the inevitability of death :)

[sakiburrahman]

Architectural overview to handle the errors. Just 'let it crash' and it's the problem-solving approach like as natural way.