useNestedSettingsUpdate: prevent unneeded syncing of falsy templateLock values #46357
Conversation
This file contains 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
|
Size Change: +12 B (0%) Total Size: 1.32 MB
ℹ️ View Unchanged
|
youknowriad
reviewed
Dec 8, 2022
| parentLock: | ||
| select( blockEditorStore ).getTemplateLock( rootClientId ), | ||
| blockListSettings: getBlockListSettings( clientId ), | ||
| parentLock: getTemplateLock( rootClientId ) ?? false, |
There was a problem hiding this comment.
I wonder if this should be in the selector itself? I remember I had that change in one PR that I closed at some point.
There was a problem hiding this comment.
I see you mentioned this alternative in the description of the PR. Let's just do it. I remember @jorgefilipecosta confirmed that both undefined and false are equivalent here.
There was a problem hiding this comment.
Performance tests show some significant improvements on this PR. The firstPaint, firstContentfulPaint, loaded and domContentLoaded metrics all go down.
I'd be doubtful about this though. I think it's probably just instability on the PR but we'll confirm it over time in the graph.
There was a problem hiding this comment.
OK I pushed the alternative implementation where getTemplateLock always returns string | false, never undefined.
I'd be doubtful about this though. I think it's probably just instability on the PR but we'll confirm it over time in the graph.
Still, an undeniable fact is that the PR removes hundreds of cascading updateBlockListSettings dispatches that are running after the initial render.
…from getTemplateLock
jsnajdr
force-pushed
the
optimize/template-lock-nested-sync
branch
from
edefaac to
fab34f3
Compare
jsnajdr
requested review from
ajitbohra,
ryanwelcher,
juanmaguitar and
fabiankaegy
as code owners
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
When working on #46350 I investigated how often does
useNestedSettingsUpdatedispatch theupdateBlockListSettingsaction and how these updates are batched.Testing on the
large-post.htmltesting post, I see a big initial batch of 579 updates, one for each block, followed by a long series of approx 100 smaller batches, about 4 updates each. This is weird, where are all these small updates coming from?It turns out they all update a
templateLocksetting fromundefinedvalue tofalse, and the affected block is always alistorlist-item. Indeed, thelarge-post.htmlpost hass approx 100 list with about 4 items each. That looks like wasted time, because in practice there's no behavior difference betweenundefinedandfalse.These updates happen because
templateLockis propagated down the tree from the root to leafs. Every block'stemplateLockis either copied from parent, or sourced from the block itself, typically from a block attribute. And thelistblock enforces atemplateLock: falsesetting on all its children. Therefore, anupdateBlockListSettingson alistchanges thelist'stemplateLockfromundefinedtofalse, and that triggers an update on all children, because theiruseNestedSettingsUpdatedepends on theparentLockvalue.I'm fixing this by defaulting
parentLocktofalse. That way, none of the redundant updates will be performed. Only the meaningful updates would be done, e.g., these that update the initialfalsevalue to'all'or'contentOnly'.And alternative fix would be to patch the
getTemplateLockselector to returnfalseas the default value instead ofundefined. We don't need it to return two distinct falsy values. What do folks think about this?Performance tests show some significant improvements on this PR. The
firstPaint,firstContentfulPaint,loadedanddomContentLoadedmetrics all go down.