State Machines and Hopeful Dreams
summary
Compares Mavericks, Workflow, and planned Decompose by actually building a Compose-first hydration tracker backed by Firebase with three features (daily, weekly streaks, settings). Mavericks uses mutable data-class states, derived properties, setState/withState and is easy to adopt; Workflow models a tree of state machines with props/output, workers for side effects, and steeper mental shift but richer testing support. The app works with both approaches; Mavericks is quicker to bootstrap, Workflow enforces a more opinionated, navigable architecture. Decompose is mentioned but not evaluated due to time.
AI-written summary · about
chapters (11)
- 0:00Introduction and welcome
- 3:57Why this talk
- 5:20MVI architecture overview
- 9:10App demonstration
- 14:28Mavericks deep dive
- 22:36Mavericks feature walkthrough
- 28:39Workflow introduction
- 34:59Workflow view layer
- 37:46Workflow feature walkthrough
- 40:36Comparison criteria analysis
- 43:44Should you use them?
related talks
19mThe Art of Functional State Management for iOS Apps
44mState in Jetpack Compose
43mGood Intentions II: Enemy of the State
42mCompose: Beyond Android development
41mStateMachine Architecture Based on MVI: Combining KMM, Jetpack Compose and SwiftUI
28mThe Art of Functional State Management
11mInvest in Real State: A workshop for State in Compose
34mBeyond the UI: Compose as a Foundation for Multiplatform Apps