A single uppercase letter crashed our entire Android application, but AI helped us fix it and build a brand-new feature across two platforms simultaneously!
In this live coding session, Marina and @peterfriese continue building a cross-platform to-do list application for iOS and Android. They start by debugging a frustrating crash—a subtle case-sensitivity issue with Firestore enum reflection in Kotlin.
Once the bug is squashed, they use Antigravity and the Gemini CLI to align our data models between Swift and Kotlin. They then tackle the complex task of adding a new lists feature.
Watch as they brainstorm the NoSQL data model, generate visual Mermaid diagrams with AI, and prompt their agents to implement the feature across both platforms in real-time. Finally, they troubleshoot some tricky iOS navigation issues and set up Firestore composite indexes to get everything working perfectly.
Chapters
00:00 Intro and catching up
03:33 The Android emulator crash from last week
05:06 Debugging the Firestore enum case-sensitivity bug
09:27 Exploring Kotlin enum overrides
15:14 Aligning iOS and Android data models
18:30 Analyzing the data models with Antigravity
20:53 Adding a 'description' field to the iOS task model
29:28 Generating a Mermaid ER diagram with AI
35:00 Designing the new "lists" data model for NoSQL
40:33 Creating the lists collection in Firestore
50:10 Using AI to plan and implement the lists feature
57:00 Pushing changes to GitHub with Gemini CLI
1:12:00 Testing
|
2026/03/03
youtube
