Information Technology Reference
In-Depth Information
able working with their own style of language and notation. For example,
someone writing a business policy may be happier expressing goals in a declar-
ative way, saying, perhaps, that a target level of production should always be
achieved, while someone documenting a process may naturally think in im-
perative terms, expressing what has to be done as a sequence of instructions.
As another example, a business process may involve extended activities
with real-time deadlines and have measures of the fraction of work completed,
while a computational task may concentrate on sequences of events that are
considered indivisible (or atomic), with deadlines expressed by equivalent
events, with no model of continuous time at all. Working with sequence, but
without continuous time like this simplifies analysis, but makes the expression
of continuous properties of the system, such as quality of service, more dicult.
1.2.4
Viewpoint Correspondences
Although we achieve a powerful simplification by dividing a system spec-
ification into the views seen by different stakeholders, the specification must
continue to be a coherent description of a single target system. If we had no
links between the viewpoints, this would not happen; what was intended to be
a single design would just fall apart into five bits. It is therefore vital that the
viewpoints be linked, and this is done by establishing a set of correspondences
between them, as visualized in figure 1.3.
In current software tools that present different user views, this linkage
is often derived from the names of objects. If the same name appears in
two diagrams, they are assumed to represent two aspects of the same thing.
enterprise
viewpoint
information computational
terms
correspondence
technology
engineering
FIGURE 1.3: How correspondences link terms in different viewpoints.
 
Search WWH ::




Custom Search