Could jamstack be the future of web applications architecture?: an empirical study

The ubiquity of internet connections has made web applications one of the most widespread means of contents delivery. Indeed, very often they are preferred to applications run locally due to their flexibility, portability, maintainability, and so forth. However, the growth of web contents, and correspondingly of users, has raised critical issues that can be reduced to the required communications between clients and servers and to the consequent overload problems.

Between 2015 and 2016, a small group of developers proposed a novel architecture for web applications based on three pillars: JavaScript, APIs, and Markup, named as Jamstack. In particular, Jamstack was founded on the idea of leveraging pre-compiled pages that could be delivered from content delivery networks, and consequently avoiding continuous requests to servers. Since then, the interest of the web community in Jamstack has been continuously growing and nowadays can be considered as an established web development architecture.

In this paper, we report on the planning, execution, and results of a mixed-methods empirical study on the maturity of Jamstack, and its adoption. More in general the study is intended to provide a structured and comprehensive assessment of Jamstack in its peak of inflated expectations technology adoption phase. We started with a systematic literature review and from an initial set of 77 studies, we selected 6 primary studies, which we analysed according to a data extraction, analysis, and synthesis process. We used the results of the systematic literature review for building an online survey, which we distributed to 44 practitioners in the web development domain. We extracted, analysed and synthesised the data from the survey and complemented it with qualitative insight from 4 online, semi-structured, in-depth interviews.