FP Tolerance in iOS Paper SafeAreaView debouncing #41614
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
facebook-github-bot
added
CLA Signed
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D51539091 |
Summary: Fixes facebook#41545 SafeAreaView works by adding padding in order to shift content out of the safe area. This may change the layout dimensions of the SafeAreaView, in turn effecting its safe area insets. This can cause layout results to change, which in turn changes the inset value. Because of this, there is a tolerance, where safe area inset changes do not trigger a new update. Yoga is instructed to round layout dimensions to the closest physical pixel, so a very small difference in layout may result being off by about a pixel. Right now the tolerance is exactly one physical pixel, and if there is FP error here, we may not pass the test, and start oscillating with different layout values. After changing affected ShadowNode order to always be root-first, the first call to set the frame of the `SafeAreaView` happens when a non-zero-sized RootView is present, which I think may lead to a safe area inset update communicated that wasn't before? Or other cosmic butterflies. Layout rounds to one physical pixel in difference, and our tolerance is `0.00001` dips off (not helped that 1/3 screen scale cannot be represented as decimal, even without FP error). This adds a small tolerance beyond just the pixel boundary, matching the logic in Fabric, which seems to resolve the issue. Changelog: [iOS][Fixed] - FP Tolerance in iOS Paper SafeAreaView debouncing Reviewed By: philIip Differential Revision: D51539091
NickGerleman
force-pushed
the
export-D51539091
branch
from
e0ca1dc to
f7c3ef1
Compare
Contributor
|
This pull request was exported from Phabricator. Differential Revision: D51539091 |
|
This pull request was successfully merged by @NickGerleman in c9c5651. When will my fix make it into a release? | Upcoming Releases |
huntie
pushed a commit
that referenced
this pull request
Nov 24, 2023
Summary: Pull Request resolved: #41614 Fixes #41545 SafeAreaView works by adding padding in order to shift content out of the safe area. This may change the layout dimensions of the SafeAreaView, in turn effecting its safe area insets. This can cause layout results to change, which in turn changes the inset value. Because of this, there is a tolerance, where safe area inset changes do not trigger a new update. Yoga is instructed to round layout dimensions to the closest physical pixel, so a very small difference in layout may result being off by about a pixel. Right now the tolerance is exactly one physical pixel, and if there is FP error here, we may not pass the test, and start oscillating with different layout values. After changing affected ShadowNode order to always be root-first, the first call to set the frame of the `SafeAreaView` happens when a non-zero-sized RootView is present, which I think may lead to a safe area inset update communicated that wasn't before? Or other cosmic butterflies. Layout rounds to one physical pixel in difference, and our tolerance is `0.00001` dips off (not helped that 1/3 screen scale cannot be represented as decimal, even without FP error). This adds a small tolerance beyond just the pixel boundary, matching the logic in Fabric, which seems to resolve the issue. Changelog: [iOS][Fixed] - FP Tolerance in iOS Paper SafeAreaView debouncing Reviewed By: philIip Differential Revision: D51539091 fbshipit-source-id: 88bddc38c7cd8d93feef5f12da64b124af22f46d
Othinn
pushed a commit
to Othinn/react-native
that referenced
this pull request
Jan 9, 2024
Summary: Pull Request resolved: facebook#41614 Fixes facebook#41545 SafeAreaView works by adding padding in order to shift content out of the safe area. This may change the layout dimensions of the SafeAreaView, in turn effecting its safe area insets. This can cause layout results to change, which in turn changes the inset value. Because of this, there is a tolerance, where safe area inset changes do not trigger a new update. Yoga is instructed to round layout dimensions to the closest physical pixel, so a very small difference in layout may result being off by about a pixel. Right now the tolerance is exactly one physical pixel, and if there is FP error here, we may not pass the test, and start oscillating with different layout values. After changing affected ShadowNode order to always be root-first, the first call to set the frame of the `SafeAreaView` happens when a non-zero-sized RootView is present, which I think may lead to a safe area inset update communicated that wasn't before? Or other cosmic butterflies. Layout rounds to one physical pixel in difference, and our tolerance is `0.00001` dips off (not helped that 1/3 screen scale cannot be represented as decimal, even without FP error). This adds a small tolerance beyond just the pixel boundary, matching the logic in Fabric, which seems to resolve the issue. Changelog: [iOS][Fixed] - FP Tolerance in iOS Paper SafeAreaView debouncing Reviewed By: philIip Differential Revision: D51539091 fbshipit-source-id: 88bddc38c7cd8d93feef5f12da64b124af22f46d
This was referenced Feb 29, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
Fixes #41545
SafeAreaView works by adding padding in order to shift content out of the safe area. This may change the layout dimensions of the SafeAreaView, in turn effecting its safe area insets.
This can cause layout results to change, which in turn changes the inset value. Because of this, there is a tolerance, where safe area inset changes do not trigger a new update.
Yoga is instructed to round layout dimensions to the closest physical pixel, so a very small difference in layout may result being off by about a pixel. Right now the tolerance is exactly one physical pixel, and if there is FP error here, we may not pass the test, and start oscillating with different layout values.
After changing affected ShadowNode order to always be root-first, the first call to set the frame of the
SafeAreaViewhappens when a non-zero-sized RootView is present, which I think may lead to a safe area inset update communicated that wasn't before? Or other cosmic butterflies. Layout rounds to one physical pixel in difference, and our tolerance is0.00001dips off (not helped that 1/3 screen scale cannot be represented as decimal, even without FP error).This adds a small tolerance beyond just the pixel boundary, matching the logic in Fabric, which seems to resolve the issue.
Changelog:
[iOS][Fixed] - FP Tolerance in iOS Paper SafeAreaView debouncing
Reviewed By: philIip
Differential Revision: D51539091