Armando Haeberer
Oblog Software SA
Alameda Antonio Sergio 7 - 1A, 2795-023 Linda-A-Velha, Portugal
Fax. +351 214144125
haeberer@oblog.pt
Abstract
Verification and validation testing of software artefacts and specifications are issues of extreme importance in software engineering. As such, these issues deserve a lot of attention and there exists a copious literature about them. Nevertheless, there are not many attempts at developing testing methodologies based on the foundations of formal specification construction. Despite the obvious connection of specification-based functional testing with the testing of scientific theories, these attempts were usually made a on purely logico-mathematical basis instead of an epistemological one. As a result, some attempts have some serious handicaps. After analysing these handicaps, an alternative approach for specification-based functional verification testing of software artefacts is analyzed. This approach is based on Clark Glymour's bootstrap strategy for relating theory and evidence in Natural Science. Unless some approaches, this strategy is applicable to the testing of non-deterministic programs; moreover, it can be used for validation testing, either for software artefacts or for specifications themselves.
Last Updated: May 29, 2000 by Elisabetta Ferrando