Audax Development Research Notes - 3: software and dataset

Audax Development Research Notes - 3 (ADRN-3) Quantifying Green IT: Architectural Complexity and Runtime Power Dynamics for Web Frameworks Author: Andrea Davanzo ORCID: 0009-0000-5170-1737 License: MPL-2.0 Overview This repository contains the experimental datasets, telemetry logs, plotting utilities, and automation scripts used in Audax Development Research Notes – 3 (ADRN-3). The study investigates how software architectural complexity influences electrical power consumption across multiple programming language web frameworks. Building on the Accidental Complexity Score (ACS) introduced in ADRN-1 and the energy measurements of ADRN-2, this work extends the analysis to PHP, Ruby, Python, and JavaScript web frameworks. Electrical power draw is measured using Intel Running Average Power Limit (RAPL) telemetry under controlled hardware conditions. ADRN-3 introduces two energy-centric metrics: Software Median Power Overhead (SMPO) – Steady-state architectural overhead Software Complexity Power Jitter (SCPJ) – Runtime variability and burst dynamics Repository Structure framework/ Minimal applications for each framework baseline/ Hardware baseline reference workload f3/ Fat-Free Framework (PHP) laravel/ Laravel (PHP) sinatra/ Sinatra (Ruby) rail/ Ruby on Rails (Ruby) flask/ Flask (Python) django/ Django (Python) fastify/ Fastify (Node.js) express/ Express (Node.js) log/ Raw telemetry and event datasets plot/ Plotting utilities and generated charts plot_results.py Utility for plotting the results charts/ venv/ script/ Supporting utilities test-framework.sh Main test script stats.php Utility for print stats