A critical examination of software science

Four correlates of complex behavioral networks: Differentiation, behavior, connectivity, and compartmentalization: Carving networks at their joints

“…In addition to leading to conclusions about invariantlength structures, another main, and connected, generalization we may make is that, because of the invariant-length structures, behavioral selected networks become increasingly differentiated as they enlarge. Others have studied the increase of differentiation in behavioral networks such as organisms [18 -29], social insects [1, 30 -32], businesses [1,[33][34][35][36][37][38][39][40][41], computer software [42][43][44][45][46][47][48], electronic circuits [1], as well on hierarchy more generally [49 -52]. We see here in Figure 4 that increasing differentiation in larger networks appears to be common (possibly universal?)…”

“…See Ref. 1 for more information about the other plots, except for software, which is new here, and was obtained from data in [42][43][44][45][46][47][48]. (Note that Halstead's theories [42] are not well-founded, but using his data is appropriate.)…”

Software Quality Metrics

“…In certain languages it is difficult to determine whether some tokens should be operators or operands (Lassez et al 1981). The meaning may depend on how the token is used at execution time, or on information given in the declaratior section.…”

“…SQM RELATIONSHIP.-Maintainability TOOLS: LOGISCOPE, Verilog USA Inc., Alexandria, VA PC-Metric, SET Laboratories, Inc., Mulino, OR F " RENCES: Albrecht and Gaffney 1983;Arthur 1983;Bulut 1974;Carver 1986;Coulter 1983;Elshoff 1976;Fitzsimmons 1978;Funrami 1976;Gordon 1976Halstead 1972Halstead , 1977Halstead , 1979Halstead and Zislis 1973;jones 1978;Kafura and Reddy 1987;Lassez 1981;Li and Cheung 1987;; Shen, Conte, and Dunsmore…”

A quantitative charcterization of control flow context: software measures for programming environments

“…First, the method of determining operators and operands was never well-defined. Lassez et al [31] argued that in some languages deciding whether a token is an operator or an operand is impossible. He gave as an example the use of a function identifier both as an invocation of the function and as a parameter to another procedure in a parameter list.…”