Archive for October 11th, 2012

Phenomenology of Software Diagnostics: A First Sketch

Thursday, October 11th, 2012

Influenced by stages of Husserlian phenomenological investigation I propose the following stages for the investigation of phenomena as it appears in software execution artifacts such as memory dumps, traces and logs:

1. Bracketing the outside source code as reduction to patterns of phenomena independent from causal software engineering explanations.

2. Constructing the computational world for the given incident (the so called horizon of computation).

3. Comparing with “computed-in” experience of past computational worlds from which all universal patterns of computational structural and behavioral phenomena emerged.

PS. According to the above software diagnostics is a phenomenological science of patterns. Most probably this sketch will be revised soon. In the mean time here’s a funny coincidence. The first step in a phenomenological method is the so called epoché. I provide my own interpretation of this - “exception processing of crash” hypothetical episteme. Similar to EPOCH metaphysical grand conjecture that our World is just one enormous exception processing handler after Big Bang.

- Dmitry Vostokov @ DumpAnalysis.org + TraceAnalysis.org -