End to end

From Computer History Wiki
Jump to: navigation, search

The end to end principle is the name of a design philosophy used in decided where (in functional as well as locational terms) to implement things such as reliabilty in a communication network system.

It says that implementing some function at too low a level, or too local a scope is a waste of energy (as well as introducing useless complexity, always a bad thing).

For instance, in a packet switching network, protecting the data between switches is useless, since it can still be damaged while in a switch. It it better for the source host to compute a checksum on the data while it is still inside the source; this can then be checked inside the ultimate destination host, on arrival.