IT EventsBook

Discussions

compose_casey
compose_casey
@compose_casey
Active Member
Joined: Aug 3, 2026
Last seen: Sep 25, 2026
Topics: 1 / Replies: 11
Reply
RE: SwiftUI vs Jetpack Compose state management, why are state bugs still hard?

The process death test least_privilege mentioned is the one nobody runs until production complains. We added it to CI as a literal test case now - kil...

2 days ago
Forum
Reply
RE: Kotlin Multiplatform vs React Native for native iOS and Android UI

We went KMP too and the 'black box' worry was real for about two months, then mostly resolved itself once iOS devs realized they could just... read Ko...

2 days ago
Forum
Reply
RE: best AI coding agent in 2026, Claude Code vs Codex vs Cursor vs Copilot

Jumping in - we ran exactly this experiment for a month each on two tools and the numbers surprised us. Time-to-first-working-diff was basically ident...

2 days ago
Forum
Reply
RE: Kotlin Multiplatform vs React Native for native iOS and Android UI

react Native is better if one team needs to ship the same screens quickly. KMP is better when native UI and platform teams are already part of the pla...

3 weeks ago
Forum
Reply
RE: On device LLM for mobile apps, privacy win or battery killer?

we download on demand over Wi Fi with clear size and storage information, then allow deletion from settings. Version the model separately from the app...

3 weeks ago
Forum
Reply
RE: SwiftUI vs Jetpack Compose state management, why are state bugs still hard?

The duplicated events are mostly after process recreation. We persist data, then replay the same success event and navigate twice

3 weeks ago
Forum
Reply
RE: On device LLM for mobile apps, privacy win or battery killer?

we use a capability check and offer local processing only on devices that meet memory and acceleration requirements. Cloud remains an opt in fallback....

4 weeks ago
Forum
Reply
RE: Spec driven development with AI coding agents, useful or waterfall in Markdown

That is still a review problem. Add open questions and confidence, and let implementation findings update the spec. The document is a shared working m...

4 weeks ago
Forum
Reply
RE: AI coding agent in a Docker dev container, how much access is too much?

A little. Testcontainers can point to a controlled remote daemon, or the CI environment can run the privileged suite after review. The local agent doe...

4 weeks ago
Forum
Reply
RE: Spec driven development with AI coding agents, useful or waterfall in Markdown

Specs help most where the agent cannot ask the product owner every five minutes. They also expose disagreement between humans before code exists.

4 weeks ago
Forum
Reply
RE: AI coding agent in a Docker dev container, how much access is too much?

Separate build access from deployment access. The agent gets source, test data and short lived credentials for a disposable environment. It doesnt get...

4 weeks ago
Forum
Scroll to Top