The basic properties of software have not yet rigorously studied in software engineering. The size or functional complexity of software is one of the examples.
The conventional model of software size is a symbolic size (Ss) in the unit of line of code (LOC). However, no matter one divides a given size of software Ss into how many pieces (Ssi, 1< i
Hi,
Software complexity and size are not the same, as Peter Breuer above has indicated. I believe complexity is a broader attribute which size is possibly a contributing sub-attribute. In software engineering, we have evolved from Maurice Halstead's metrics where we counted operators and operands and the number of occurrences of these to McCabe's complexity count of number of binary predicates to Henry-Kafura's information flow where we counted the number of calls and number of passing parameters to more recent definitions of and counting of "cohesion" and "couplings."
In all of these work we were always striving to "simplify" complexity to improve on and lessening human effort. We have spent a lot of time exploring and explaining the principle of divide and conquer since functional decomposition, modular programming and structured programming days of H. Mills, E. Dijkstra ,L. Constantine and D. Parnas. So, while we are still working on this concept and do not have a universally agreed to principle of divide & conquer, i do not believe we have "failed" in this endeavor. On the contrary, I believe we have made great strides in this area.
Your thoughts?
- Badges
- Science topic
- Similar topics
- Engineering
- Computer Engineering