RORK LABJP
PLAY — Google Play's target API level 36 requirement took effect yesterday, August 31. From today, new apps and updates must target Android 16VISIBILITY — Apps still on API 35 stay listed but disappear for users on newer Android versions. No error is raised; new installs simply fade, which makes the change easy to missEXTENSION — If you missed the deadline, an extension through November 1, 2026 can be requested in Play Console — best filed alongside a concrete migration planAPPLE — On the Apple side, the event lands September 9 and iOS 27 is reported to ship September 14. Testing generated apps on iOS 27 hardware before release week is time well spentEXPO — Expo released expo-paste-input on August 28, a native module that brings image, GIF, and sticker paste to React Native TextInputEAS — EAS Observe reached general availability on August 20, putting crash and performance monitoring on the same EAS platform as builds and updatesPLAY — Google Play's target API level 36 requirement took effect yesterday, August 31. From today, new apps and updates must target Android 16VISIBILITY — Apps still on API 35 stay listed but disappear for users on newer Android versions. No error is raised; new installs simply fade, which makes the change easy to missEXTENSION — If you missed the deadline, an extension through November 1, 2026 can be requested in Play Console — best filed alongside a concrete migration planAPPLE — On the Apple side, the event lands September 9 and iOS 27 is reported to ship September 14. Testing generated apps on iOS 27 hardware before release week is time well spentEXPO — Expo released expo-paste-input on August 28, a native module that brings image, GIF, and sticker paste to React Native TextInputEAS — EAS Observe reached general availability on August 20, putting crash and performance monitoring on the same EAS platform as builds and updates
Back to Blog

Five Days, Two Apps: What I Learned Building Non-Stop with Rork Max Over Golden Week

Rork MaxIndie DevelopmentGolden WeekApp DevelopmentDeveloper Diary2026

Japan's Golden Week just wrapped up — and I spent most of it building.

No trips, no sightseeing. Just five days of focused development using Rork Max. Part of the reason was practical (avoiding the crowds), but honestly, I had a nagging feeling that I hadn't fully explored what Rork Max could do. A long, uninterrupted stretch felt like the right time to settle that.

I ended up shipping two apps and getting one to TestFlight. That's the highlight reel. The full story has some messier parts too, and I think those are worth sharing.

The First Decision: Pick One Theme

One trap I've fallen into before is trying to "explore everything" during a creative break. I once spent an entire week dabbling in three different projects and shipped none of them. Painful memory.

This time, I gave myself one constraint: stick to a single theme. I chose interactive generative art with ambient sound — partly because that's territory I know well (I've been doing art and music work for years), and partly because having a clear personal vision tends to make in-the-moment decisions faster.

The concept was simple: a canvas where the shapes you draw with your finger influence the sound playing in real time. The meditation/wellness app market is crowded, I know. But building something you'd genuinely use makes the development process feel less like work.

What Rork Max Got Right

Here's the thing about Rork Max that I keep coming back to: when your concept is clear, the generation speed is genuinely fast.

I opened my first session with a prompt describing an interactive canvas using react-native-skia with touch path tracking and reactive sound. Within twenty minutes, I had something moving on screen. The Path + gesture handling structure it generated was honestly cleaner than what I might have written freehand.

The other thing that's improved — at least in how I use the tool — is the granularity of follow-up instructions. I've started being much more specific: "ease out the color transition curve," "delay the sound trigger by 100ms after the finger stops." Precise, small instructions lead to smaller diffs, which means fewer unintended side effects. That pattern made a real difference over five consecutive days.

Where Things Got Complicated

The synchronization between audio playback and Skia animations was where I lost the most time.

Using expo-av while keeping Skia animations in sync turned into a platform-specific nightmare — timing drift on certain Android devices, stuttering on older iOS hardware. I fed the issue back to Rork Max through several rounds of iteration, and it eventually rewrote the timing logic using useRef and requestAnimationFrame. That fixed it on one device and broke something else.

What finally resolved it was me reading the generated code, identifying the actual problem (startTime calculation during rerender), and giving a surgical instruction. The lesson here, which I keep relearning: the quality of AI-assisted development scales with how well you can read and interrogate the output. It's not about removing the developer from the loop — it's about changing what the developer does in that loop.

App Two: A Speed Experiment

After spending a day and a half on the generative art app, I deliberately switched modes for the second build: speed over completeness.

I built a minimalist Japanese proverb app — one a day, clean typography, no fluff. The design brief to Rork Max was explicit: "prioritize high-quality UI, minimize features, leave room for later additions." The generated code was noticeably leaner. Less to debug. Faster to iterate.

The shape of your prompt mirrors the shape of your app. Complicated prompts produce complicated code. Simple, well-scoped prompts produce code that's easier to own and extend.

After Five Days

I'm sitting here on the last day of Golden Week with two working apps. One goes to App Store review this week.

Rork Max made this possible. That's not a small thing — a year ago, this kind of output in five days would have required a team or a lot of cutting corners. But the honest reflection is that the tool wasn't the only variable. What to build, how to describe it, where to trust the output and where to push back — those decisions still belong to the person at the keyboard.

I'll keep writing about actual development experience here at Rork Lab. Not just "here's how feature X works," but what it actually felt like to build with it. If you're curious about anything specific from this sprint, leave a comment or reach out.

Thanks for reading, and I hope your Golden Week treated you well too.