0001-strategy-for-data-streams.rst

Title: Strategy for asynchronous data streams in the OpenEdx Project

Date: 14 November 2023

Status

Accepted

Context

In the OpenEdx project, we are developing a mobile application using a Kotlin language for Android users. To ensure optimal support of the application, we need to make a decision regarding which asynchronous data streams will be used for the future Android app development. This document outlines the decision to support native Kotlin based StateFlow and SharedFlow.

Decision

We decide to use StateFlow and SharedFlow for asynchronous data streams management between UI and view models. All new features should use flows instead of LiveData. All LiveData occurrences in current code should be replaced with flows in future.

Why is this important?

1. Deep Kotlin integration

Flow is tightly integrated with Kotlin Coroutines. Code parts which are using Kotlin Coroutines is usual use flows, we don't need to map data to another format and could use it as is.

2. Decreasing dependencies count

If we are using language based choose, we don't need to add additional dependencies to our code.

Project Impact

This decision will impact the project in the following ways:

Improved performance and functionality. Reducing application size.

Implementation

To implement this decision, we will use flows as asynchronous data streams for future development. We will replace LiveData with flows during refactoring

Alternatives

Continuing to use LiveData as asynchronous data streams, will keep job to us to maintain LiveData library updates.