Abstract

The question of what is a well-designed temporal data model and query language is a difficult, but also an important one. The consensus temporal query language TSQL2 attempts to take advantage of the accumulated knowledge gained from designing and studying many of the earlier models and languages. In this sense, TSQL2 represents a constructive answer to this question. Others have provided analytical answers by developing criteria, formulated as completeness properties, for what is a good model and language. This paper applies important existing completeness notions to TSQL2 in order to evaluate the design of TSQL2. It is shown that TSQL2 satisfies only a subset of these completeness notions. In response to this, a minimally modified version of TSQL2, termed Applied TSQL2, is proposed; this new language satisfies the notions of temporal semi-completeness and completeness which are not satisfied by TSQL2. An outline of the formal semantics for Applied TSQL2 is given. 1 Depa...

Citations