System Architecture and Complexity

8Interoperability

The concept of interoperability is at the very center of the justification for writing this work. The emergence of this new concept dates back some time: in a timeframe of scarcely 10 years in the 1980s we witnessed a progressive, though rapid, replacement of “large” mainframes and their environment of thousands of “dumb” terminals, in the sense that the user cannot program them – they can only be configured – by groups of machines of varying power that are organized to cooperate in architectures described as clients/servers. Communications networks are at the center of these new groups, an architecture that for a while was described as Network-centric to really emphasize its importance.

The processing power that until then was concentrated in the central system, with up to 8 or 16 processors, but not many more for performance reasons, was from then on able to distribute itself across all equipment containing a programmable processor, thanks to the microelectronics and VLSI (very large-scale integration revolution [of components]). Terminals were replaced by work stations and/or PCs which came into use at that time; these are individual work stations that are all programmable. The central system was split up over potentially tens/hundreds of servers that can specialize in data servers, application servers, safety and autonomic manager/back-up servers, lists of personnel and authentication servers. The power provided, as long as the software architectures allow ...