Formal Specification of Open Standards and the Case of RSS v2.0

Abstract

Open standardization seems to be very popular among software developers as it makes the standard's adoption by the software engineering community easier and smoother. Formal specification methods, on the other hand, while very promising, are being adopted by protocol engineers very slowly; the industry seems to have little motivation to move into this, almost unknown, territory.In this paper the authors present the i) idea of applying formal methods (formal specification techniques) to open standards' specifications, and ii) an example of a formal specification of open standards, RSS v2.0 in particular. The authors support and provide evidence for the advantages of the open standards formal specification over natural language documentations (the current way that open standards' specifications are released). Formal specifications are more concise and consistent than the ones written in natural languages, while significantly less ambiguous and more complete with respect to the original documentation. Furthermore, they are executable in many cases and highly reusable as most formal methods automated support tools allow for module inheritance. The merging of formal specification methods and open standards allows for i) a more concrete standard design; ii) an improved understanding of the environment under design; iii) an enforced certain level of precision into the specification, and also iv) provides software engineers with extended property checking/verification capabilities, especially if they choose to use any of the algebraic specification languages. The authors showcase how the RSS v2.0 standard can formally be specified using the CafeOBJ formal specification language and demonstrate why the particular formal specification is beneficial.