How PHP Releases Are Adopted in the Wild?

TL;DR

This paper analyzes PHP release adoption patterns on the web using big data and Markov chain analysis, revealing uniform adoption, infrequent downgrades, and evolving trends over time.

Contribution

It introduces the application of DTMCs for empirical tracing of online software deployments, providing insights into PHP release adoption behaviors.

Findings

Adoption of PHP releases is relatively uniform across observed domains.

Downgrading of PHP releases is generally rare.

Adoption trends vary between 2016-2017 and the 2010s.

Abstract

This empirical paper examines the adoption of PHP releases in the the contemporary world wide web. Motivated by continuous software engineering practices and software traceability improvements for release engineering, the empirical analysis is based on big data collected by web crawling. According to the empirical results based on discrete time-homogeneous Markov chain (DTMC) analysis, (i)~adoption of PHP releases has been relatively uniform across the domains observed, (ii) which tend to also adopt either old or new PHP releases relatively infrequently. Although there are outliers, (iii) downgrading of PHP releases is generally rare. To some extent, (iv) the results vary between the recent history from 2016 to early 2017 and the long-run evolution in the 2010s. In addition to these empirical results, the paper contributes to the software evolution and release engineering research traditions by elaborating the applied use of DTMCs for systematic empirical tracing of online software deployments.