…Native apps start 12% faster) (#49449)
Summary:
Okay the title is a bit clickbaity, but this is actually true. (on Android)
We (Janic, Szymon, Ruby and Me) discovered something interesting. React Native uses `mmap` for mapping the JS bundle to RAM, to avoid having to load the entire thing instantly at app startup.
Ruby doubted that this was true - so we investigated.
Apparently on Android, resources are **compressed**. And if the JS bundle is stored compressed, it has to be uncompressed before it can be loaded into RAM, hence not allowing it to be mmapp'ed! (see [`_CompressedAsset::getBuffer`](https://cs.android.com/android/platform/superproject/+/master:frameworks/base/libs/androidfw/Asset.cpp;l=903?q=Asset.cpp))
So with this PR, we now add `.bundle` files to `noCompress` in the react-native gradle plugin, which disables compression for the JS bundle.
We discovered while improving the performance of one of our clients: **Discord**.
In our tests, **this improved the TTI of the Discord app by 400ms!! (or 12%)** 🤯🚀
NOTE: Yes, the .apk will now be bigger. But; Google Play compresses it anyways, so the **download size** of your .apk will likely not increase by much. It will be bigger on disk though.
## Changelog:
[ANDROID] [CHANGED] Add option to disable bundle compression to improve startup time
<!-- Help reviewers and the release process by writing your own changelog entry.
Pick one each for the category and type tags:
[ANDROID|GENERAL|IOS|INTERNAL] [BREAKING|ADDED|CHANGED|DEPRECATED|REMOVED|FIXED|SECURITY] - Message
For more details, see:
https://reactnative.dev/contributing/changelogs-in-pull-requests
Pull Request resolved: #49449
Test Plan:
### 1. Verify compression is disabled
Build two apps, one with this patch and one without. When I did this using the RN community template, the one without this patch was 47,6 MB, and the one with this patch was 48 MB in size. So the .apk got bigger, which is what we expected
### 2. Verify app startup is faster
Use tools like react-native-performance or custom markers to measure TTI. In our tests, we shaved off 400ms from the startup time, which was about 12% of Discord's total TTI. (on a low-end Android device)
In Expensify, we improved the TTI by 14-20% with this change (source: Expensify/App#56930)
Reviewed By: javache, cipolleschi
Differential Revision: D69742221
Pulled By: cortinico
fbshipit-source-id: bd59d77662bd30a3acdbb2e9f8d8f23db922c3f2
facebook/react-native#49449
Apparently on Android, resources are compressed. And if the JS bundle is stored compressed, it has to be uncompressed before it can be loaded into RAM, hence not allowing it to be mmapp'ed! (see _CompressedAsset::getBuffer)
So with this PR, we now add .bundle files to noCompress in the react-native gradle plugin, which disables compression for the JS bundle.
This PR improves TTI by 180-260ms (14-20%)
Explanation of Change
Fixed Issues
$
PROPOSAL:
Tests
Offline tests
QA Steps
// TODO: These must be filled out, or the issue title must include "[No QA]."
PR Author Checklist
### Fixed Issuessection aboveTestssectionOffline stepssectionQA stepssectiontoggleReportand notonIconClick)src/languages/*files and using the translation methodSTYLE.md) were followedAvatar, I verified the components usingAvatarare working as expected)StyleUtils.getBackgroundAndBorderStyle(theme.componentBG))Avataris modified, I verified thatAvataris working as expected in all cases)Designlabel and/or tagged@Expensify/designso the design team can review the changes.ScrollViewcomponent to make it scrollable when more elements are added to the page.mainbranch was merged into this PR after a review, I tested again and verified the outcome was still expected according to theTeststeps.Screenshots/Videos
Android: Native
Android: mWeb Chrome
iOS: Native
iOS: mWeb Safari
MacOS: Chrome / Safari
MacOS: Desktop