“…Consider the sequence of root schemas: root.1.0.xsd, root.2.0.xsd, ... We write a simple temporal bundle for these and invoke the Squash utility, which produces a single temporal document, tv snapshot.xml which is then referenced by multiple <?xml version="1.0" encoding="UTF-8"?> <temporalBundle xmlns="http://www.cs.arizona.edu/tau/tauXSchema/TBSchema" xmlns:tv="http://www.cs.arizona.edu/tau/tauXSchema/TVSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.cs.arizona.edu/tau/tauXSchema/TBSchema"> <format plugin="XMLSchema" granularity="date"/> <bundleSequence defaultTemporalAnnotation="defaultTA.xml" defaultPhysicalAnnotation="defaultPA.xml"> <schemaAnnotation snapshotSchema="root.xsd" temporalAnnotation="temp_anno.xml" physicalAnnotation="phy_anno.xml"> </schemaAnnotation> </bundleSequence> </temporalBundle> This rather involved state of affairs, with time-varying documents and timevarying schemas, is illustrated with a T Diagram in Figure 7. In this notation, first described over forty years ago [4], the input of a translator is given on the left arm of the "T" (for example, for SchemaMapper in the upper right-hand-side of the figure, the input is the logical schema document, bundle.xml), the name of the translator is given at the base of the "T" (here, "Schema Mapper"), and the output of the translator is given on the right arm of the "T" (here, a representational schema, rep.xsd). The name of these diagrams was to the best of our knowledge given by McKeeman, Horning, and Wortman in their classic compiler book [17].…”