On Fri, 2003-12-05 at 01:01, Brian de Alwis wrote: > This loss of context seems to happen for two reasons: they forget > what they are doing while browsing and understanding the code > (perhaps due to information overload), or are distracted from their > original task and unable to remember what they were doing (digression).
This reminds me... After having read Graesser, Millis, and Zwaan's "Discource Comprehension" (Annual Review of Psychology, vol 48, pp. 163-189, 1997) a while back, I attempted to transfer their discontinuities to software development. Below you will find an extract of my notes on cognitive complexity. I do not claim that they are accurate or sufficient or even scientific, but they may help with your focus: Expectancy ---------- Unexpected information causes interruptions which flushes information in the working memory. Deviations from expected patterns of processing creates discontinuities. Handling such discontinuities requires extra cognitive effort. Inspired by discourse catagories (Graesser et al., 1997), we can classify the discontinuities as follows: o Spatial Discontinuity. The spatial dispersion of necessary information. Functional dispersion results in more difficult tracing of control-flow, in higher coupling, and in lower cohesion. Data dispersion results in more difficult tracing of data-flow, and in lower encapsulation. o Temporal Discontinuity. The difficulty of correlating temporally separated information, or predicting the occurrence of events. o Causal Discontinuity. The difficulty of directly relating a symptom to its cause. This phenomemon intensifies as the software evolves. o Intentional Discontinuity. The difficulty of correlating the intended behaviour with the actual behaviour. A discrepancy between intended and actual behaviour normally signifies an error. o Stylistic Discontinuity. The difficulty of adjusting to changes in the coding style. A coding style is normally designed to reduce the introduction of errors or to signal intend. Deviations in coding style can lower the readability of source code. ---------------------------------------------------------------------- PPIG Discuss List ([EMAIL PROTECTED]) Discuss admin: http://limitlessmail.net/mailman/listinfo/discuss Announce admin: http://limitlessmail.net/mailman/listinfo/announce PPIG Discuss archive: http://www.mail-archive.com/discuss%40ppig.org/
