Skip to content

chore(ci): Upgrade performance test apps to React Native 0.85.1 - #6088

Merged
antonis merged 6 commits into
mainfrom
antonis/upgrade-perf-test-apps-rn-0.85.1
May 6, 2026
Merged

chore(ci): Upgrade performance test apps to React Native 0.85.1#6088
antonis merged 6 commits into
mainfrom
antonis/upgrade-perf-test-apps-rn-0.85.1

Merge branch 'main' into antonis/upgrade-perf-test-apps-rn-0.85.1

690bc89
Select commit
Loading
Failed to load commit list.
@sentry/warden / warden: find-bugs completed May 6, 2026 in 36s

1 issue

find-bugs: Found 1 issue (1 medium)

Medium

startupTimeTest diffMin lowered to -3000 effectively disables regression detection on improvements side and masks anomalies - `performance-tests/metrics-ios.yml:9`

The diffMin threshold for startupTimeTest was changed from -20 to -3000. Since diffMin defines the lower bound of an acceptable startup-time delta, setting it to -3000 ms means virtually any unexpectedly large negative delta (e.g., caused by a broken/short-circuited startup path that no longer initializes required components) will pass the test. This removes the guardrail that detects suspicious startup-time drops, which can hide real regressions where the app appears faster because it isn't doing its initialization work.


Duration: 35.2s · Tokens: 151.7k in / 556 out · Cost: $0.49

Annotations

Check warning on line 9 in performance-tests/metrics-ios.yml

See this annotation in the file changed.

@sentry-warden sentry-warden / warden: find-bugs

startupTimeTest diffMin lowered to -3000 effectively disables regression detection on improvements side and masks anomalies

The `diffMin` threshold for `startupTimeTest` was changed from -20 to -3000. Since `diffMin` defines the lower bound of an acceptable startup-time delta, setting it to -3000 ms means virtually any unexpectedly large negative delta (e.g., caused by a broken/short-circuited startup path that no longer initializes required components) will pass the test. This removes the guardrail that detects suspicious startup-time drops, which can hide real regressions where the app appears faster because it isn't doing its initialization work.