Abstract

This paper presents reasoning implying that the outcome of the execution of an <SMALL> E</SMALL>-type program or <SMALL> E</SMALL>-type software system (<SMALL> software</SMALL> for short) of whatever class are not absolutely predictable. It is intrinsically uncertain. Some of the sources of that uncertainty are identified and it is argued that the phenomenon qualifies as a Principle of Software Uncertainty. The latter represents an example of an assertion in a Theory of Software Evolution which is ripe for development based on empirical generalisations identified in previous research, most recently in the FEAST projects. The paper briefly discusses some practical implications of uncertainty, and the other concepts presented, on evolution technology and software processes. Though much of what is presented here has previously been discussed, its presentation as a cohesive whole provides a new perspective.

Description

sdasda

Links and resources

Tags

community

  • @neilernst
  • @dblp
@neilernst's tags highlighted