4.2 Software Engineering (SE) 

Software Engineering was originally defined as (Naur and Randell, 1969)
"the establishment and use of sound engineering principles in order to obtain economically software that is reliable and works efficiently on real machines."
The area of software process modeling -- abbreviated SPM (or just PM in an SE context) -- is a subdomain of software engineering, focusing on the development and use of models of software production processes. The process concept is often used in a broad sense, incorporating artifacts, roles and tools, in addition to activities (Conradi, Fernström and Fuggetta, 1994:13).

The relevance of SPM to enterprise modeling is clear when considering the software producing organization to be the enterprise.

Purposes of SPM are many and include (from Curtis et al., 1992:77; Armenise et al., 1992:455; and Lonchamp, 1994:337) expressing for humans, analysis, guidance, enforcement, planning, measuring and monitoring, improving, communication, configuring, reuse of process fragments, verification, simulation, coordination and process execution. The stated purposes are clearly overlapping and internally dependent, but they indicate that the SPM community intend to cover all three categories of modeling.

However, even if human understanding is claimed to be supported, what it takes to make humans understand software processes and process models is not discussed to any significant extent in the surveyed literature.

The type of work being modeled in SPM is obviously in the information part of the work matrix of Figure 3.4: Both the software artifacts that are created and the development tools that are used to produce the software are essentially information. Concerning risk, Bollinger and McGowan (1991:35) claim that

"...the assembly-line processes on which the SEI model is based are dominated by replication risks; while the vast majority of software processes are instead dominated by design risks."
Hence, development of software is associated with design risk. Still, the SE community envision software production as an engineering discipline (Booch, 1986:28), so SPM must handle both replication and design risk.

A characteristic feature of SPM is that both the actual production process and the meta-process are modeled (Conradi, Fernström and Fuggetta, 1994:15). By meta-process is meant the process of building and maintaining process models. A main reason for focussing on management of the meta-process is that the production process may change frequently and the process models must evolve accordingly.