Concurrent test systems using the structured distributed-programming paradigm

Abstract

Traditionally developers in the automatic testing community have had no viable paradigm for handling concurrent systems. Test suites that could be processed in one hour using concurrency are being processed in 24 hours sequentially. The structured distributed programming paradigm (SDPP) is a non-traditional paradigm that allows developers to conceive viable distributed/concurrent/parallel systems with the same ease and assurance that they have been conceiving regular single-process sequential systems for the past 30 years. Structured distributed programming (SDP) is primitive, recursive, and applicable at all levels from application systems to operating systems to hardware systems to network systems. SDP is a mindset like object-oriented programming (OOP) is a mindset. As with OOP, the concepts of SDP are primitive and supportable by primitive syntax/semantics in programming languages. At compile time, an OOP sequential system is completely visualizable as a tree-structured graph. At compile time, an SDP system is completely visualizable as dual graphs, one a tree-structured graph of time and the other a cyclic graph of space. Since SDP systems are completely visualizable at compile time, test systems can be guaranteed, at compile time, to never race or deadlock; to always complete, cooperate and integrate with their environment; and to be maintainable, modifiable, reliable, and scalable to any degree.