State management approaches in Flutter

Abstract

Flutter has tens of different ways to solve the issue of state management, which creates an ambiguity in the selection of the approach. Flutter is a promising technology which can possibly reduce the costs of cross-platform application development. The main objective of the thesis was to categorize the state management approaches and find a way to select the most suitable one for the most common use case scenarios. Instead of focusing on the particular use cases, a set of decision-making criteria was determined, and the approaches were analyzed and compared based on these criteria.

Quantitative research methods were used in the theoretical part in order to introduce the mobile development market, compare Flutter to the competing technologies, prove that the problem of state management approach selection ambiguity indeed exists, and determine which approaches would be worth studying by measuring their popularity among Flutter developers. These methods included, for instance, analyzing and comparing the statistical data concerning the mobile development market, number of GitHub stars, various pub.dev scoring metrics and so on. Qualitative methods (e.g. critical analysis of existing literature) were used to categorize the approaches, determine the comparison criteria, analyze and compare the selected approaches and make a conclusion.

The study resulted in several new contributions to the existing literature. First, all of the at least a little bit popular state management approaches were collected in one place. Second, these approaches were categorized by the common attributes defined in the thesis. Third, the most popular and widely used representative of each group was studied in detail by implementing a real application with shared preferences, local persistence and remote API requests. These included setState, InheritedWidget, Provider, GetX, BLoC, MobX, Redux. Finally, the approaches were analyzed based on the criteria defined in the thesis, and the comparison table was created. This table allowed finding the most suitable approach for the needs of each particular project or developer.

The resulting comparison table and the implementation part explaining how to use each state management approach give an answer to the question: how to select the most suitable state management approach? It also gives the starting point beginning to use the selected approach. The goals that were set for this thesis were fully achieved; in several places the thesis provides even more detailed answers than was initially expected.