mirror of
https://github.com/permissionlesstech/bitchat.git
synced 2026-07-25 22:25:20 +00:00
* Consolidate duplicate BLE links after restore; suppress duplicate fragment streams; onChange coalescing
Field evidence (two-phone test after BLE state-restoration relaunches):
both phones held 2-3 simultaneous same-role connections to each other —
one side received every packet from three distinct centrals all bound to
the same peer — so every PTT voice frame arrived 3x, and a 41KB voice
file went out as TWO complete independent 89-fragment streams (different
assembly ids), both fully reassembled and the duplicate only dropped at
the very end by messageID dedup. 2-3x airtime/battery on all traffic
plus doubled reassembly memory.
Root causes and fixes:
- Duplicate links stayed unbound forever: announces (the only packet
that binds a link to a peer) went through the per-peer duplicate-link
collapse, so a peer's second/third link never received the announce it
needed to become bound — and unbound links pass the collapse untouched,
so every broadcast sprayed down all of them. Direct announces now
bypass the collapse and reach every live link (relayed announces keep
it); once bound, the existing collapse dedups all traffic.
- Same-role link retirement: a verified direct announce now consolidates
our central-role connections to that peer — keep the link the announce
arrived on (or the most recently bound one), cancel other connected
links bound to the same peer. One connection per role per peer is the
normal dual-role topology; only same-role duplicates are touched, only
links already announce-bound are retired (never pre-announce links),
at most one retirement per peer per rebind-cooldown window, and the
peer keeps a live link either way. Directness stays forgeable (TTL is
unsigned), so a replay could nominate the survivor — bounded by the
cooldown and by DM routing's existing canDeliverSecurely gate. A
rotation rebind now also cancels other stale links still bound to the
rotated-away ID, so the ghost identity retires promptly.
- Deterministic preferred-link collapse: when several bound links to one
peer are candidates, collapse now keeps the peer's most recently bound
link (the reverse-mapped one) instead of dictionary order; links
without a discovered characteristic are excluded from fanout (they
cannot be written to, and could silently eat a peer's collapsed copy).
links(to:) now reports all bound peripheral links, and removing one
duplicate no longer clobbers the reverse map of the survivor.
- Duplicate fragment streams: a transferId-less resend (gossip-sync
replay, spool) of file content already being fragmented out to a
covering audience is dropped at the outbound scheduler (broadcast
covers everyone; directed covers its recipient). App-initiated sends
carry an explicit transferId the progress UI tracks and always run.
A peer that asks after the stream completes still gets a resend.
- didUnsubscribeFrom no longer flaps a peer that is still live on other
links (the far side retiring its duplicate arrives as an unsubscribe);
didDisconnectPeripheral bookkeeping is skipped for self-retired links
by removing the store entry before cancelling.
Also: guard the DeliveryStatus onChange state write in TextMessageView/
MediaMessageView (unconditional per-row writes under a message storm
tripped SwiftUI's "tried to update multiple times per frame" warning).
The restore-path bgRemaining=∞ log was already fixed on main (#1425
review follow-up 07b5ac31: init-time seed + sampler-routed restore
captures); verified, no change needed.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
* Review fixes: multi-link disconnect guard, characteristic-aware retirement
Adversarial review of the duplicate-link PR (merge-with-nits):
- didDisconnectPeripheral gets the same multi-link guard as the
unsubscribe path: when a duplicate link drops naturally while the peer
stays live on another (dual-role central link, or a second bound link
during the post-restore consolidation window), peer-disconnect
bookkeeping (markDisconnected + disconnect notify) no longer runs — a
UI blip until the next announce re-marked the peer connected. The
reverse map was just repaired onto a connected survivor, so
directLinkState is the accurate probe. Scan restart and connect-slot
refill stay unguarded: they respond to the physical drop regardless of
remaining logical links. (No unit test: driving the CBCentralManager
delegate requires CBPeripheral instances, which cannot be constructed
in tests; the policy pieces backing the guard are covered.)
- Retirement is now characteristic-aware: the policy snapshot carries
characteristic presence, and keptPeripheralUUID selects anchors only
among writable links while any exist — consolidation must not keep a
link mid-service-rediscovery (didModifyServices cleared its
characteristic) and cancel the writable duplicate, stranding outbound
traffic on the central link until rediscovery finishes. When neither
anchor is writable but a writable duplicate exists, consolidation
defers to a later announce instead of guessing. The reverse-map
survivor repair in removePeripheral prefers writable links for the
same reason. Three new policy tests cover charless-vs-writable.
Deferred with code comments per review: central-side collapse keeps the
oldest subscription (no recency signal; remote consolidates within its
cooldown), and the extra preferred-bindings bleQueue hop in
sendOnAllLinks (fold into a combined snapshot if profiling flags it).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---------
Co-authored-by: jack <jackjackbits@users.noreply.github.com>
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
74 lines
3.2 KiB
Swift
74 lines
3.2 KiB
Swift
import BitFoundation
|
|
import Foundation
|
|
|
|
/// Decides which central-role connections (peripheral links we own) are
|
|
/// redundant duplicates of a peer's live link.
|
|
///
|
|
/// One connection per role per peer is the normal dual-role topology (each
|
|
/// device is both central and peripheral). After a BLE state-restoration
|
|
/// relaunch, though, the same phone can reappear under a fresh peripheral
|
|
/// UUID while the restored connection lives on — leaving several live
|
|
/// central-role connections to one peer, each carrying every packet
|
|
/// (field-verified: 2-3x airtime on all traffic). Only same-role duplicates
|
|
/// are retired; the peer's central-role subscription on our peripheral
|
|
/// manager is its own connection to manage, and it runs the same policy.
|
|
enum BLERedundantLinkPolicy {
|
|
struct PeripheralLink: Equatable {
|
|
let uuid: String
|
|
let peerID: PeerID?
|
|
let isConnected: Bool
|
|
/// Whether the link has a discovered characteristic (is writable).
|
|
/// A link mid-service-rediscovery (didModifyServices cleared it)
|
|
/// must never be kept over a writable duplicate.
|
|
let hasCharacteristic: Bool
|
|
|
|
init(uuid: String, peerID: PeerID?, isConnected: Bool, hasCharacteristic: Bool) {
|
|
self.uuid = uuid
|
|
self.peerID = peerID
|
|
self.isConnected = isConnected
|
|
self.hasCharacteristic = hasCharacteristic
|
|
}
|
|
}
|
|
|
|
/// The link to keep when a peer has several connected bound peripheral
|
|
/// links, or nil when there is nothing to consolidate. Prefers the
|
|
/// ingress link of the verified direct announce that triggered the check
|
|
/// (the strongest liveness proof available), falling back to the peer's
|
|
/// most recently bound link — but only among writable links while any
|
|
/// exist: keeping a characteristic-less link and cancelling the writable
|
|
/// duplicate would strand outbound traffic on the central link until
|
|
/// rediscovery finishes. When neither anchor is a viable candidate,
|
|
/// consolidation waits for a later announce rather than guessing.
|
|
static func keptPeripheralUUID(
|
|
ingressPeripheralUUID: String?,
|
|
mostRecentlyBoundUUID: String?,
|
|
links: [PeripheralLink],
|
|
peerID: PeerID
|
|
) -> String? {
|
|
let bound = links.filter { $0.peerID == peerID && $0.isConnected }
|
|
guard bound.count > 1 else { return nil }
|
|
|
|
let writable = bound.filter(\.hasCharacteristic)
|
|
let candidates = writable.isEmpty ? bound : writable
|
|
|
|
if let ingressPeripheralUUID, candidates.contains(where: { $0.uuid == ingressPeripheralUUID }) {
|
|
return ingressPeripheralUUID
|
|
}
|
|
if let mostRecentlyBoundUUID, candidates.contains(where: { $0.uuid == mostRecentlyBoundUUID }) {
|
|
return mostRecentlyBoundUUID
|
|
}
|
|
return nil
|
|
}
|
|
|
|
/// Connected peripheral links bound to the peer other than the kept one.
|
|
static func peripheralUUIDsToRetire(
|
|
links: [PeripheralLink],
|
|
peerID: PeerID,
|
|
keeping keptUUID: String
|
|
) -> [String] {
|
|
links
|
|
.filter { $0.peerID == peerID && $0.isConnected && $0.uuid != keptUUID }
|
|
.map(\.uuid)
|
|
}
|
|
}
|