Callstack

✨ Just wrapped up my talk “The Missing Middle Ground: Fast & Real Testing for React Native” at React Alicante 🎤🇪🇸 I introduced a new library we’ve been working on at Callstack: React Native Harness. 👉 Created by my colleague Szymon Chmal (amazing work!) 👉 It makes it possible to run Jest-style tests directly on real devices/simulators, with no mocks by default 👉 With just a few imports, your familiar Jest tests suddenly get full access to the native runtime - it feels automagical ✨ This library fills a gap between unit tests (fast, but mocked) and end-to-end tests (real, but slow). It’s designed for testing native modules, spec compliance (like fetch or WebSocket), and more - all with developer-friendly DX. 📺 The recording of my talk will be published soon. 📦 We’re releasing the library on October 14 - stay tuned! Big thanks to the React Alicante organizers, and to the React Native community for the energy and feedback 🙌

Yesterday Amazon announced Vega OS, their new Linux-based platform with first-class support for React Native. This Thursday, we're going live to walk you through it and build the first Vega OS app together. 📅 October 2, 5PM CET https://lnkd.in/dGSNNge6

📢 Introducing Vega OS and Vega Developer Tools! Vega OS allows for fast app launches and smooth interfaces without sacrificing the Fire TV experience customers know and love. With Vega Developer Tools, you can build for Vega OS your own way - from React Native to web apps. Read the blog➡️ https://lnkd.in/gMvqTqhP Start building today➡️ https://lnkd.in/gFwNYSAm

In React Native, you may be working with both the built-in Animated API and Reanimated. Each uses its own container for values (Animated.Value vs SharedValue), but they don’t interop. That becomes a real challenge when third-party components expect one format while your app uses the other. This article covers: - Why interoperability is tricky - When you actually need it - How to bridge values natively to keep animations smooth 👉 [link in next post]

🚀 New in React Native AI: the MLC provider. Run any model locally with React Native AI and AI SDK ⚡️ How it works? - MLC is an optimized runtime for iOS, Android, and desktop GPUs - You download model weights onto the device - The engine loads and sets up everything needed for inference Unlike the Apple provider, MLC doesn't rely on system-wide models. Instead, it fetches weights for each model you choose. That can take a moment, but it unlocks a lot of flexibility. The MLC provider is part of our growing suite of tools for running AI on mobile devices. ⬇️

If you've ever hit the "Unable to load script" error in the Android emulator, you know how frustrating it can be. We put together a step-by-step list that covers the most common fixes: - Metro Bundler not running - Port reverse not applied - Network and security settings - Emulator data and firewalls It's meant as a quick reference the next time this red box appears. 👉 Full article in the comments

One of the coolest details in iOS 26 is Liquid Glass: UI elements that morph into a single surface when placed together. We've brought this effect to React Native with our `@callstack/liquid-glass package` ✨ If you want to understand how it works and how to use it effectively in your apps, we put together a guide covering: - What Liquid Glass is and how Apple designed it - How to implement it in React Native - Best practices to keep your UI clear and accessible 👉 Link in comments