react-native-startup-time

This module helps you to measure your app launch time. It is measured from the earliest point in time available to the native module, that is the module's initialization. When getTimeSinceStartup is called on JS side, you'll get a promise which resolves with difference between these two moments in ms. This is not very accurate, but should give you good enough base for further optimizations.

On iOS time measurement is based on this article. On Android SystemClock.uptimeMills is used.

As far as I know, there's no way to programmatically obtain time passed since the moment when user taps on app icon. For this you have to use native dev tools. On Android this module will call reportFullyDrawn so you can inspect adb logs.

If you know a better way to measure startup time (in a module), let me know or better shoot a PR.

Getting started

$ yarn add react-native-startup-time

$3

This module supports autolinking so if you use RN 0.60+ then no additional action is required.

Otherwise, run

$ react-native link react-native-startup-time

$3

Usage

Render startup time badge somewhere on your first screen:

`jsx
import { StartupTime } from 'react-native-startup-time';

ready={true / optional, defaults to true /}
style={styles.startupTime / optional/}
/>
`

Or use imperative call:

`jsx
import { getTimeSinceStartup } from 'react-native-startup-time';

// when you app is ready:
getTimeSinceStartup().then((time) => {
console.log(Time since startup: ${time} ms);
});
`

_The following sections are applicable to Android only_

$3

This makes sure Android doesn't resolve the getTimeSinceStartup promise more than once per app execution. More information in PR #10.

Since v1.4.0 this strategy is enabled by default, if you're migrating from a previous version and you just want things to keep working as they are, follow the steps below.

#### Disabling Single-Sampling:

Be aware, depending on which lifecycle hook you've attached your call to getTimeSinceStartup() you might receive redundant invocations, e.g. when the app is brought from bg to fg. Because the app isn't really starting up, the measured time can be unrealistic; such unrealistic samples adulterate your data.

To disable single-sampling strategy, create your package using constructor with parameter false:
`java
// Define package
new RNStartupTimePackage(false)
``