I've experimentally seen that the notifications(named:) API of NotificationCenter appears to buffer observed notifications internally. In local testing it appears to be limited to 8 messages. I've been unable to find any documentation of this fact, and the behavior seems like it could lead to software bugs if code is not expecting notifications to potentially be dropped. Is this behavior expected and documented somewhere? Here is a sample program demonstrating the behavioral difference between the Combine and AsyncSequence-based notification observations: @Test nonisolated func testNotificationRace() async throws { let testName = Notification.Name("TestNotification") let notificationCount = 100 var observedAsyncIDs = [Int]() var observedCombineIDs = [Int]() let subscribe = Task { @MainActor in print("setting up observer...") let token = NotificationCenter.default.publisher(for: testName) .sink { value in let id = value.userInfo?["id"] as! Int observedCombineIDs.append(id) print("🚜 observed note with id: \(id)") } defer { extendLifetime(token) } for await note in NotificationCenter.default.notifications(named: testName) { let id: Int = note.userInfo?["id"] as! Int print("🚰 observed note with id: \(id)") observedAsyncIDs.append(id) if id == notificationCount { break } } } let post = Task { @MainActor in for i in 1...notificationCount { NotificationCenter.default.post( name: testName, object: nil, userInfo: ["id": i] ) } } _ = await (post.value, subscribe.value) #expect(observedAsyncIDs.count == notificationCount) // 🛑 Expectation failed: (observedAsyncIDs.count → 8) == (notificationCount → 100) #expect(observedCombineIDs == Array(1...notificationCount)) print("done") }

In the header for workloop.h there is this note: A dispatch workloop is a "subclass" of dispatch_queue_t which can be passed to all APIs accepting a dispatch queue, except for functions from the dispatch_sync() family. dispatch_async_and_wait() must be used for workloop objects. Functions from the dispatch_sync() family on queues targeting a workloop are still permitted but discouraged for performance reasons. I have a couple questions related to this. First, I'd like to better understand what the alluded-to 'performance reasons' are that cause this pattern to be discouraged in the 'queues targeting a workloop' scenario. From further interrogation of the headers, I've found these explicit callouts regarding differences in the dispatch_sync and dispatch_async_and_wait API: dispatch_sync: Work items submitted to a queue with dispatch_sync() do not observe certain queue attributes of that queue when invoked (such as autorelease frequency and QOS class). dispatch_async_and_wait: Work items submitted to a queue with dispatch_async_and_wait() observe all queue attributes of that queue when invoked (inluding [sic] autorelease frequency or QOS class). Additionally, dispatch_async_and_wait has a section of the headers devoted to 'Differences with dispatch_sync()', though I can't say I entirely follow the distinctions it attempts to draw. Based on that, my best guess is that the 'performance reasons' are something about either QoS not being properly respected/observed or some thread context switching differences that can degrade performance, but I would appreciate insight from someone with more domain knowledge. My second question is a bit more general – taking a step back, why exactly do these two API exist? It's not clear to me from the existing documentation I've found why I would/should prefer dispatch_sync over dispatch_async_and_wait (other than the aforementioned callout noting the former is unsupported on workloops). What is the motivation for preserving both these API vs deprecating dispatch_sync in favor of dispatch_async_and_wait (or functionally subsuming one with the other)? Credit to Luna for originally posing/inspiring these questions.