On Mon, 17 Feb 2025 15:37:56 GMT, John Hendrikx <jhendr...@openjdk.org> wrote:
>> This provides and uses a new implementation of `ExpressionHelper`, called >> `ListenerManager` with improved semantics. >> >> # Behavior >> >> |Listener...|ExpressionHelper|ListenerManager| >> |---|---|---| >> |Invocation Order|In order they were registered, invalidation listeners >> always before change listeners|(unchanged)| >> |Removal during Notification|All listeners present when notification started >> are notified, but excluded for any nested changes|Listeners are removed >> immediately regardless of nesting| >> |Addition during Notification|Only listeners present when notification >> started are notified, but included for any nested changes|New listeners are >> never called during the current notification regardless of nesting| >> >> ## Nested notifications: >> >> | |ExpressionHelper|ListenerManager| >> |---|---|---| >> |Type|Depth first (call stack increases for each nested level)|(same)| >> |# of Calls|Listeners * Depth (using incorrect old values)|Collapses nested >> changes, skipping non-changes| >> |Vetoing Possible?|No|Yes| >> |Old Value correctness|Only for listeners called before listeners making >> nested changes|Always| >> >> # Performance >> >> |Listener|ExpressionHelper|ListenerManager| >> |---|---|---| >> |Addition|Array based, append in empty slot, resize as needed|(same)| >> |Removal|Array based, shift array, resize as needed|(same)| >> |Addition during notification|Array is copied, removing collected >> WeakListeners in the process|Appended when notification finishes| >> |Removal during notification|As above|Entry is `null`ed (to avoid moving >> elements in array that is being iterated)| >> |Notification completion with changes|-|Null entries (and collected >> WeakListeners) are removed| >> |Notifying Invalidation Listeners|1 ns each|(same)| >> |Notifying Change Listeners|1 ns each (*)|2-3 ns each| >> >> (*) a simple for loop is close to optimal, but unfortunately does not >> provide correct old values >> >> # Memory Use >> >> Does not include alignment, and assumes a 32-bit VM or one that is using >> compressed oops. >> >> |Listener|ExpressionHelper|ListenerManager|OldValueCaching ListenerManager| >> |---|---|---|---| >> |No Listeners|none|none|none| >> |Single InvalidationListener|16 bytes overhead|none|none| >> |Single ChangeListener|20 bytes overhead|none|16 bytes overhead| >> |Multiple listeners|57 + 4 per listener (excluding unused slots)|57 + 4 per >> listener (excluding unused slots)|61 + 4 per listener (excluding unused >> slots)| >> >> # About nested changes >> >> Nested changes are simply changes... > > John Hendrikx has updated the pull request incrementally with one additional > commit since the last revision: > > Fix fix for regression :) <h2>Rules</h2> This PR was created because there is an unwritten contract for `ChangeListener`s that people expect, but that we are currently not guaranteeing. I've written these down as three rules: Rule 1: Old value received in callback X must match new value received in callback X-1 Rule 2: Old value should never Object::equals new value (collection properties break this rule) Rule 3: The received new value is the same as property::get I think these rules need to become part of the specification of `ChangeListener`. A possible further rule that I think we may want to explicitly specify somewhere is: Rule 4: A listener is never notified after removal <h3>Edge Case: Two listeners upon notification change the value, and cannot reach agreement</h3> |ExpressionHelper|This PR|Specification| |---|---|---| |This leads to a `StackOverflowError` as new nested levels of notifications keep being added until the stack is exhausted. During this process the listeners receive old values they may never have seen before violating Rule 1|The later listener will stop being called as its last received value equals the notification value if it was changed back; the listener will not be called (to preserve Rule 2) and so is unaware that the value has changed to something it didn't set|Specify that conflicting changes lead to an exception| Opinion: this PR should probably not silently allow this but instead throw an exception ------------- PR Comment: https://git.openjdk.org/jfx/pull/1081#issuecomment-2671686463 PR Comment: https://git.openjdk.org/jfx/pull/1081#issuecomment-2671687002