The plan's two status lines both predated the gate's second run. The header
still read "approved to start Phase 0a" nine phases in, and the DAG note still
claimed the re-run had not happened and that phase 6 waits on a passing results
file. That file has existed since 2026-07-26 —
screenshare-audio-exclusion-phase5-results.md records GATE PASSED on run 2 with
all 13 §5.1 rows, including rows 4 and 5 at the real tagging sites.
Both lines now state what actually blocks phase 6: the 0c -> 0d -> 6 edge, with
0c step 2 still open (S1/S2 merged, S3a built but unmerged, S3b/S4/S5 not
started) and 0d unbuilt. The superseded 2026-07-25 note is kept for the trail
rather than deleted.
Docs only; no code or gate changes.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`target_delay` grows +1 per disruption to MAX_DELAY_FRAMES (240 ms) but only
shrinks after 250 consecutive clean frames — 5 s of unbroken audio. Two of the
five `clean_run` resets fire on every natural pause in speech (jitter.rs:201
benign underrun, jitter.rs:181 re-prime), and the sender stops transmitting
outright while the gate is closed (core/mod.rs:2041). The AIMD decrease half is
therefore unreachable under conversational voice: one early jitter burst pins
the extra latency for the rest of the session.
Found by code review; not yet reproduced live. Pairs with field-test debt #5.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Rounds 17c and 4 of the repair review, recorded together because they resolve to
one decision: `pactl`'s text output cannot carry the guarantees repair claims, so
observation and unloading now go through libpulse introspection over a single
verified-local connection. The dependency was taken with the user's sign-off after
vetting (details beside the dep in pixelpass Cargo.toml).
Three lessons that generalise beyond this phase:
- A *prescription* can fail reachability just as a finding can. "Use `pactl -f json
list modules`" is sound reasoning against an API that does not exist — those
records carry no module index, and `unload-module` accepts only an index.
- Auditing my own fixes paid a third time: two of the four fixes applied in round
17a were themselves defective, including a correlation scheme that is unsound
whenever module names repeat.
- The live field test caught a bug unit tests structurally cannot reach, and it was
phase 0b's bug one layer down: fields drop in declaration order, the Pulse
context's teardown frees IO events owned by the mainloop, and declaring the
mainloop first turned a fully successful repair into SIGABRT and exit 134.
Also recorded: the newline defect needed no adversary and was confirmed on the live
server, and the remaining namespace hole is left open with its trade stated — an
owner token would close it but would make orphans from older builds uncleanable.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Two reviews in one round. The repair planner returned changes-requested (no P1s,
four reachable P2s, all applied in pixelpass `9145b2a`); the 0c actor design
returned four blocking issues, all accepted, plus a concession on epoch.
Recorded because three of them generalise:
- A prescribed fix was not implementable as written. "Use `pactl -f json list
modules`" is sound reasoning against an API that does not exist: on pactl 17
those records carry no module index, and `unload-module` takes only an index.
The reachability rule now applies to prescriptions, not just findings.
- The actor's bounded join would have disarmed `Drop` by moving the thread handle
into `spawn_blocking` — the same defect shape as round 15's, a defence disarmed
exactly when needed.
- Epoch was over-specified and my vacuity instinct was right: serial equality is
the entire identity guarantee, so epoch is diagnostic and explicitly not a gate.
Measured on the live graph rather than argued: recorded module arguments are
byte-exact with `@DEFAULT_SINK@` unresolved (both load-bearing for exact-form
matching), and two sinks may share one `node.name` with capture attaching to the
OLDER one in 3 of 3 trials — so a surviving wedged owner silently steals the next
session's capture instead of merely risking a collision.
0c step 2 is sliced into S2–S5 so each lands reviewed. Nothing reopens D6: the
connection-owned-sink design is unchanged, and a material part of the growth is
pre-existing debt 0c forced into the light.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Codex's re-review of the branch returned "approve with follow-ups" — no
blocking findings, five P3s. All five are applied here rather than carried as
debt, since each is a few lines.
The one with user-visible consequences: `stop_host` returned a bare "was
sharing" bool, so the single case where the availability-first policy gives up
(SIGKILL queued, reap never confirmed) still sent `ScreenShareStopped` with
nothing else. The UI would say sharing had ended while pixelpass might still be
alive and fanning out — a claim the user cannot see through. `shutdown` now
returns `StopOutcome`, `stop_host` returns `Option<StopOutcome>`, and an
unconfirmed *user-initiated* stop raises a UI error naming the stray process.
Session and viewer teardown discard the outcome deliberately: nobody is waiting
on an answer there, and the residual risk is already logged.
Also: the three failure diagnoses in `shutdown` (the signal never left, the
child ignored it, the wait itself broke) were collapsed into one log line and
are now distinct — they mean different things to whoever reads the log.
The second cancellation gate is the one worth keeping. The review pointed out
that all cancellation coverage sat in the *graceful* wait, so a mutant that
disarmed the wrapper between the two waits would survive. It was right, with a
wrinkle: the naive mutant does not compile, because the child is borrowed from
`self` for the whole function — the borrow checker is doing real work here. The
restructured form (`self.child.take()` once cooperation has failed) does
compile, and the pre-existing mid-wait test passes it.
`cancelling_shutdown_after_the_kill_leaves_the_fallback_armed` kills it.
Mutation-verified, both new gates: reporting an unconfirmed stop as `Reaped`
fails exactly `a_failed_wait_is_not_treated_as_a_confirmed_reap`; disarming
between the waits fails the new cancellation test (and the failed-wait test,
which also asserts armedness) while leaving the old mid-wait test green — which
is the proof the new test is not redundant. The logging split is diagnostics
only and has no gate; said plainly rather than dressed up as covered.
Docs: the "four mutations" line is now an explicit table naming each target and
its test, with 0c's pair counted under 0c; and the aggregate teardown latency is
recorded as a deferred item with a trigger (a fourth routine child, or a
measured teardown over 5 s) instead of an unwritten known cost.
638 lib tests, clippy clean, fmt clean.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The 0c mechanism probe was run on this host before any structural work, because
one unverified assumption could have invalidated the whole approach: whether a
hand-created `adapter` node is visible to pipewire-pulse under the name the
capture path depends on. It is. Five gates green, including the two that
mattered — `<node.name>.monitor` is exposed as a Pulse source, and SIGKILL of
the owning connection removes both Pulse-visible names with zero graph residue.
No null-sink module is involved at any point. Every O1 stop condition for 0c is
retired, and the default sink never moved, so the probe is safe on a live
desktop.
The probe also settled the native-sink scope question: it applies to every mode
that owns a capture sink, not only `DesktopExcluding`. That makes the `--repair`
rework load-bearing rather than defensive — repair derives dead PIDs only from
`module-null-sink` entries, so once the sink is native its loopbacks become
undiscoverable orphans.
§10 gains rounds 14 and 15: the 0b matrix drops to four mutations because the
best-effort wake arm is unreachable by construction (the loop owns a sender, and
the biased select would win anyway), and teardown is hoisted to one unconditional
post-loop site instead of being duplicated across one live arm and one dead one.
Round 15 records the two blocking implementation-review findings and the vacuous
gate of my own that the review's test-double critique exposed.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Design v3.7 §6.1.1 gains the round-13 box (the rule, the ordering that is
load-bearing in both directions, and why bridging deliberately still uses the
full union); the phase-5 results file records the close with the measurement
the deferral was waiting for; the impl plan's phase-6 gate note drops F11-1.
pixelpass c78eb2d is the implementation.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The §5.1 dry-run audit gate passes. All 13 rows completed with a non-empty
eligible half in every one, and O5 is re-measured on the fixed graph: worst
recompute 67 µs across every run, 10 µs mean under deliberate churn, busy
fraction 0.0006, readiness 1-2 ms with 18 binds against a 2000 ms budget.
Round 10's finding, and it is the third of exactly the same shape: the
pipewire-pulse PID derivation required a SINGLE repeated pipewire.sec.pid.
WirePlumber repeats one too (two Clients, both sec_pid 1747), so the derivation
returned None permanently on a stock desktop, key 4's suppression never fired,
and every Pulse-emulated node fused into one owner. Row 1's CLEAN control
forwarder and Firefox were both excluded. Fixed in pixelpass 91c4ded by deleting
the heuristic: probe every distinct sec_pid and let /proc/<pid>/comm decide.
Three measured rounds now, all at the observation boundary, none in the
architecture -- and all three were fail-closed and silent, caught only because
§5.1 requires asserting what must remain ELIGIBLE. An exclusion-only checklist
would have passed every one of these builds.
Rows 4-6 are closed through peerspeak's REAL tagging sites (call, mpv, notify,
plus clip) with a hand-launched mpv staying eligible, so the cross-repo contract
is proven end to end on live nodes. Row 10 covers the full sticky lifecycle
including retirement; row 11 is provably non-vacuous (the recycled
node.link-group came back byte-identical and did not inherit taint).
Recorded and NOT claimed as passes: substitutions in rows 8, 9 and 13, and two
reporting-only findings (the audit's sticky flag is uninformative; a bridge's
named key is lost when a leg reappears under a new serial).
Phase 6 remains blocked by F11-1, phases 0b/0c/0d and the Stereo Mix design
call -- this file removes one gate, not all of them.
Phase 3r landed §6.7 and the phase-5 audit was re-run against it immediately.
It found a second measured defect within minutes: a real hardware sink
carrying `unresolved-ancestry` permanently, from one link observed while its
output node was still unbound during enumeration. Round 8 made that
systematic rather than rare, because every node is now withheld until its
bind resolves.
New §6.8: sticky taint is a claim about history, and uncertainty is not
history. Retiring by reason code would not be enough — an unresolved node
propagates `TaintedUpstream`, which is indistinguishable from real
contamination once recorded — so the split is by provenance: the engine runs
its fixpoint twice, and only the evidence-only pass may feed sticky state.
Decisions are unchanged and still fail closed.
Also recorded in §6.8, both from Codex's round-9 review and both pre-existing:
hardware playback-to-capture paths ("Stereo Mix") defeat the `session_device`
classifier in a way the driver denylist cannot detect — a real echo path
needing a design call — and the 2 s readiness budget has no calibration
argument beyond one measurement on one idle desktop.
Impl plan: phase 3r marked built and merged with its gate results, including
the extra Device-side live gate and why row 1 alone could not cover it.
The phase-5 dry-run gate failed on its first live run: the engine built to
v3.4 could not see its own primary taint root (echo, AEC off) while excluding
every stream on the machine (silence). One cause — the PipeWire registry
`global` event carries only a filtered subset of an object's properties, and
eight the design depends on are never announced.
Design doc (v3.4 → v3.5):
- NEW §6.7 — the observation boundary. The global is an index, not a source of
truth: bind every Node and Device, `info` props are the sole source, live
prop tracking, one readiness obligation per unbound node, fail closed.
Four user design calls recorded.
- §5.1 — a second, registry-visible tag carrier (`node.name` prefix) alongside
`peerspeak.owned`, so the primary root does not rest on one mechanism.
- §6.4 — node/device props are not an optimisation to skip, they are
unavailable from the global; the round-6 Link lesson was right and applied
to exactly one object type.
- §6.1.0, §6.1.4 — the two corrections the impl plan owed v3.5: a
time-dependent "hazard is LIVE" claim, and an unreachable nominated test
case (twice over).
- §9.1 measured facts, §12 rig discipline (pw-dump binds; the registry does
not), §14 readiness.
Impl plan:
- NEW phase 3r with a four-part exit gate, the first the direct inverse of the
finding. Ports deliberately not bound in v1, with a revisit trigger.
- Phase 1 pins the second carrier literal as a cross-repo contract.
- Phase 5 marked GATE FAILED; matrix and O5 re-run after 3r and 1.
- Risk register: the over-exclusion row fired and worked; new row for the
observation boundary class.
Architecture is unchanged and vindicated: fed correct properties, the engine
decided correctly in every fixture. The §5.1 exact-partition requirement is
what caught this — every exclusion was defensible and the eligible half was
empty.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Impl plan §5's required results file. Phase 6 does not start.
F1 (fatal): the PipeWire registry `global` event delivers only a filtered
subset of node properties, and eight of the properties the phase-3 adapter
reads are not among them — peerspeak.owned, pulse.module.id, node.link-group,
application.process.id, node.passthrough, device.api, factory.name,
alsa.driver_name (plus port.exclusive on Ports). They are silently absent, so
the primary taint root never fires, the AEC identity can never validate, and
session_device is universally false. Measured on PipeWire 1.6.8 /
WirePlumber 0.5.15, with the full announced key set for all five object types
recorded. Links and Clients are unaffected; pulse-PID derivation works.
F2: with F1 in force no node has a strong owner key, so any tainted capture
stream is an unbounded tainted reader and phase 2's fail-closed backstop
excludes every Stream/Output/Audio on the machine. Fail-closed, so silence
rather than echo — but entirely non-functional, and non-functional in a way an
exclusion-only checklist would have scored as passing. The eligible half of
the §5.1 partition is what caught it, exactly as the plan argued it would.
The fix direction is measured and recorded: binding each Node and reading its
info props recovers every missing property, which is the pattern phase 3
already built for Links. factory.id is not a shortcut — it resolves to
"adapter", not api.alsa.pcm.sink.
O5 is closed with ~4 orders of magnitude of headroom: 308 graph events in
6.5s under churn, every recompute under 50us (max 15us), busy fraction 0.0004.
Caveat recorded — measured on the degraded graph, and the F1 fix adds
per-node bind I/O this run did not measure.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Turns the converged v3.4 design into ordered phases with falsifiable exit
gates. Three adversarial review rounds with Codex (gpt-5.6-sol, xhigh);
findings adjudicated rather than accepted wholesale, with reachability
verified against source on both sides.
Structural decisions:
- Phase 0d closes BOTH unsafe paths into the capture (source string and
capture-sink inputs) before any machinery that could take them exists.
- Phase 5 dry-run audit mode is a hard gate: the taint engine runs against
the live graph, creating no links, asserting exact eligible/excluded
partitions with reason codes.
- Link manager is deliberately last among the pixelpass components.
Two measured corrections owed back to v3.4 (plan §11): §6.1.0's "hazard is
LIVE right now" has already flipped and must not be gated on, and §6.1.4
nominates an unreachable test case (as did my first replacement for it).
Design approval only. No code, nothing approved for merge.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Checked ~/.config/peerspeak/config.json: output_device and input_device are
both pinned to the Arctis, so echo_cancel::enable always passes sink_master
explicitly and the AEC binds to real hardware regardless of Sunshine owning
the default sink. Not live for this user.
Kept as a low-priority general defect: on "system default", the master args
are omitted (echo_cancel.rs:89-94) and module-echo-cancel binds to whatever
the default is, which on a box like this one is a null sink. Hardening would
be to resolve and validate the default before load. Own task, not this
feature.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Round 7. Codex ratifies: v3.3 is ready to become the implementation plan.
Seven rounds, every blocker fixed or refuted with evidence. Design approval
only — nothing approved for merge, no code written.
Two subtle catches from the ratification round, both applied:
- The owner-key union had a wording trap that would have preserved the exact
bug it was written to fix. "Resolves" must mean "yields a MATCH between the
two legs", not "first property present on the node" — client.id IS present
on both gst-launch legs but differs, so a first-present implementation stops
at key 3, sees a mismatch, concludes "different owners" and leaks. Now
specified as try-in-order-until-equal, with a dedicated test.
- Sticky taint must be lifetime-aware, not keyed on raw ids. client.id, node
ids, module indices, link-groups and PIDs all recycle on this stack, so a
bare key would hand an unrelated future app permanent inherited taint.
Stored against live owner components, cleared only when all members vanish.
Also added: how pixelpass learns the pipewire-pulse PID itself (consistent
pipewire.sec.pid across Pulse clients, validated against /proc/<pid>/comm),
with the failure modes in both directions — safe only because unresolved
ancestry is fail-closed, which is the invariant the section rests on.
NEW LIVE FINDING (§6.1.0), the strongest reachability evidence yet and one
Codex's sandbox could not have seen: the user's CURRENT DEFAULT SINK is
sink-sunshine-stereo, a support.null-audio-sink. Every hardware sink is
SUSPENDED; the only RUNNING sink is Sunshine's virtual one, with Firefox
playing into it and sunshine reading its monitor. The hazardous forwarder
topology is live in the default audio path full time, with no EasyEffects
involved. It also means the rejected hardware-sink-only shortcut would have
captured NOTHING on this machine. Flagged separately, explicitly UNVERIFIED:
what module-echo-cancel binds to when the default sink is an app-owned null
sink.
§12 expanded with a graph-engine test surface (node-local tests cannot catch
C2/C3-class defects). §14 rewritten: convergence table, agreed v1 scope, and
what is deliberately out.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Round 6. Codex disagreed with two of my three round-5 claims and was right
about both; I had independently refuted one of them with a sharper test.
C2 REFUTED (by my own measurement): client.id is NOT an owner bridge. One
gst-launch process doing capture+playback produced TWO client objects (209
input, 210 output), no link-group, same application.process.id 20172. So
client.id bridges a *connection*, not an owner, and GStreamer — the same
framework pixelpass uses — splits them by default. Replaced with a
conservative union, strongest first: node.link-group, owned pulse.module.id,
client.id, node application.process.id, else fail closed.
With a trap Codex did not flag: application.process.id is pipewire-pulse's
PID for module-created streams, so bridging on it would fuse every Pulse
module's legs into one owner and mass-exclude tunnel/RTP/loopback audio the
user may legitimately want shared. Never bridge on that key when it equals
the pipewire-pulse PID; keys 1-2 already cover those precisely. PID thus
returns to the design in the CORRELATION role while remaining unusable in
the IDENTITY role — and in that role a wrong answer fails closed.
C3 CONCEDED: taint must be STICKY. Current-topology taint forgets buffered
audio — an app that reads a tainted monitor, buffers, then closes its input
leg would be relinked while still emitting peerspeak audio from the buffer,
and no graph event marks the drain. Taint now persists per owner until its
nodes disappear. Added §6.1.4 quantifying the arrival-side window (~10.6-21.3
ms quantum plus scheduling) and noting it is zero when taint roots already
exist, which is the common case.
C1 SUSTAINED with Codex's caveat: node-granular traversal is free for the
monitor boundary, but over-taints Audio/Duplex nodes. Fail-closed, accepted
for v1, documented as a known contradiction of the "Firefox with a mic stays
shareable" promise on duplex devices.
S1: endpoint props demoted to an optimization; bind-LinkInfo fallback is the
correctness path. S2: readiness epoch + revalidate before each link creation.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Round 5. Codex and I converged independently on the same conclusion — a
Link-only ancestry walk does not catch the leak — it from the crate/header/
WirePlumber sources, me from the live graph. Its sandbox could not reach the
daemon (pw-dump: Operation not permitted), so the measurements are mine.
Reproduced the EasyEffects topology with module-null-sink + module-loopback
(same shape, no EasyEffects needed). Result: there is NO Link object between
a forwarder's input leg and its output leg. Walking upstream from the leaking
node over Links alone finds no inbound links at all — a dead end that reads
as "clean". The legs are related only by shared node.link-group / client.id /
pulse.module.id.
So the signal graph needs three edge types:
1. Link edges — measured: registry Links carry all four endpoint props.
2. Sink-monitor — measured FREE at node granularity: the monitor connection
IS a real Link whose output node is the sink itself. Codex held that this
must be modelled explicitly; that is true only for a port-granular walk.
Taint walks at node granularity, links are created per port.
3. Owner bridge — node.link-group when present, else client.id (measured
shared across the forwarder's legs, distinct per app). Only modules set
link-group, so client.id is what covers ordinary apps.
New §6.1.1: bridge taint must be CONDITIONAL on the input leg being tainted.
"Client has both legs ⇒ exclude" would exclude every app using a microphone.
Firefox in a Meet call stays shareable; Firefox sharing desktop audio does not.
Also: §6.5 rejects the cheap "hardware-sink-only" predicate with a measurement
— the forwarder's output leg links directly to alsa_output, so the shortcut
passes the leak and excludes the innocent app, backwards on both halves.
§6.3 barrier corrected: core sync/done is a previous-work roundtrip, not graph
quiescence. §6.4 adds crate version, endpoint fast path + bind fallback, and
full-recompute cost. §5.2 correction 5 rewritten: application.process.id lives
on the Node and is the app's own PID; pipewire.sec.pid lives on the Client and
is pipewire-pulse's for every Pulse client. That resolves four rounds of
contradictory PID claims.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Round 4 (review-2026-07-21-design-v3-round4.md) returned 5 findings, 3 of
them blocking. All claims re-verified against source before acceptance.
Biggest correction: eligibility is a GRAPH property, not a node property.
Exclusion does not propagate downstream — a filter-chain/loopback/combine-sink
re-emits the mix as a fresh untagged Stream/Output/Audio that passes both the
peerspeak.owned and pulse.module.id checks, re-injecting the whole call into
the share. Reachability confirmed: easyeffects IS installed on this machine
(it merely wasn't running during the fan-out spike, which is why the spike
missed it). §6 rewritten around transitive upstream reachability, tracking
Node/Port/Link globals, with a registry sync barrier and revalidation
immediately before each link creation.
Also applied:
- §5.3 is now a bounded validation state machine, not a one-shot check.
wait_for_nodes only waits for the virtual source/sink, never the playback
hazard leg, and pixelpass capture spawns lazily on first viewer, so the
one-shot check raced in both directions. Revocation redefined as loss of
the module identity, not transient absence of one leg.
- §7.2: reordering ActiveSession fields is NOT sufficient — kill_on_drop
sends SIGKILL without waiting, so AEC can still unload while pixelpass
lives. Fix is explicit shutdown().await at both channel-close breaks,
field order as defence in depth, plus a fake-resource ordering test.
- §5.1 relabelled implementation sites; none of them tag anything today.
- Stop Share citation corrected to :699/:3480.
- D1-D7 resolved; readiness section added.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
v1/v2 described a move-based design that Option C superseded on 2026-07-20,
and the AEC playback-leg identity gate has since passed. Roughly two thirds
of v2 documented problems Option C does not have, so this is a rewrite rather
than a patch (v1/v2 remain at 88ad5a0 / 10203e1).
Folds in: the four AEC gate results, the five corrections that constrain them
(observed correlation not a contract; exact-equality only; index/link-group
reuse and node-id recycling; group prefix = hazard detection not ownership;
application.process.id == pipewire-pulse for module-created streams), the
verified implicit-drop ordering defect in ActiveSession, fail-closed
validation/revocation, the IPC shape, and the split-out prerequisites.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Ran the direct-link spike on the live graph (PipeWire 1.6.8,
WirePlumber 0.5.15). Fan-out carries full-level audio for paplay, mpv
and VLC while the application keeps its existing speaker link;
WirePlumber does not reap foreign links across default-sink switch,
suspend/resume or 100s steady state; and non-lingering links are
destroyed automatically when their owning connection is SIGKILLed.
The decisive result is that destroying the capture sink mid-share left
the application playing to its speakers undisturbed, so capture-side
failure degrades to "not captured" rather than breaking the user's
audio. That is the property the move-based design had to work hard to
approximate.
Records what the spike does not prove: fidelity beyond signal presence,
daemon restart, quantum perturbation, and exclusive/passthrough streams.
The capture null sink is still pactl-owned, so Stop Share continues to
leak a module every time and the graceful-stop work is still owed.
Eligibility becomes a broad guarded selector rather than a narrow
allowlist, since copying no longer risks disturbing the source.
Option A and its attendant cleanup, restore and output-switch machinery
are retained for the record but are no longer the plan. A v3 rewrite is
owed once the AEC playback-leg identity is settled.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Rewrite after Codex's adversarial review of v1 found four release
blockers, all independently verified against source.
v1's premise was wrong: whole-desktop capture bypasses Routing::start
entirely (pipeline.rs:121), so this needs a new capture mode rather than
an inverted predicate.
v2 replaces PID-based identity with ownership by inherited tag, and makes
the router an allowlist so unrecognized infrastructure is left alone
rather than optimistically moved. Graceful stop becomes a prerequisite:
Stop Share is currently SIGKILL, so cleanup never runs on the normal path.
Records live measurements taken 2026-07-20: PULSE_PROP tagging reaches
the graph for paplay, mpv and VLC, and application.process.id is the
client's own PID, not pipewire-pulse's — correcting a claim both the
review and v1 relied on.
Adds Option C (fan out a second owned link instead of moving streams),
which deletes most of the cleanup, latency and multi-host problems the
move-based design has to solve. Not yet implemented; gated on a
feasibility spike.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes the final phase of docs/chat-hardening-plan.md. Two problems: a
locally echoed message always looked sent even when the core had no active
session or the gossip broadcast failed; and a fast burst could broadcast
'successfully' yet be silently dropped by every receiver's per-author rate
bucket (8 burst, then 1/s) with no sender feedback.
Send status: CoreCommand::SendChat/SendChatFile carry a local-only id (never
on the wire); the core replies with UiEvent::ChatSendResult after the gossip
broadcast succeeds or fails, and a no-active-session is now an explicit
failure rather than a silent no-op. gossip send_chat, which previously
returned Ok on a missing sender/topic or an encode failure, now returns Err.
ChatEntry gains local_send: Option<LocalSend>; failed sends render a red
'Not sent — {reason} [Retry]' line, Broadcast/Pending render nothing
(there are no delivery receipts, so silence is the honest success state).
Sender-side pacing (new src/app/sendqueue.rs): sends past the burst queue
locally as 'queued…' and trickle out at the receivers' sustained rate, so
nothing is lost and typing is never blocked (user chose queue-and-trickle
over input throttling). The pacer reuses the gossip gate's own TokenBucket +
per-author constants (now pub(crate)) so the two sides of the policy can't
drift. A 250ms drain subscription runs only while the queue is non-empty.
Retry re-dispatches the retained payload; re-serving the same attachment id
replaces the ServeStore entry rather than double-counting bytes. The pacer
and monotonic send-id counter survive a room reset (receivers' buckets
persist; ids never alias a late result); queue and retry payloads are cleared.
582 lib tests (+11: 4 pacer/queue seam, 7 app-level transition/retry/reset);
all-targets green, clippy -D warnings clean, fmt clean, smoke launch OK. No
wire change (GOSSIP_PROTO stays 5). Tests-green-only — the two owed
two-machine field-test items are logged in the plan.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Also re-triggers CI: run 165 on 1d038be died to rust-lld crashes from disk
exhaustion on the runner host (12G free vs ~12G cold-build transient), not a
code failure; 18G of local build artifacts have been swept (30G free now).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Phase 4 of docs/chat-hardening-plan.md — URL and rendering resilience.
Closes the chat-body half of S14 (bidi override strip).
- sanitize: new is_safe_web_url shared link policy (url crate, promoted to a
direct dependency): http/https scheme + non-empty host + no userinfo;
candidates failing it stay plain text (their whole whitespace run, interior
not re-scanned). Scheme detection is now ASCII-case-insensitive.
- sanitize: linkify() -> link_ranges()/segments(): validated byte ranges
computed once, exact-roundtrip slicing, at most CHAT_MSG_MAX_LINKS (8)
clickable links per message; the rest stays selectable plain text.
- sanitize_chat: strips bidi overrides/isolates (U+202A-202E, U+2066-2069)
from message bodies while keeping ZWJ/ZWNJ/LRM/RLM (S14 chat-body half).
- app: ChatEntry caches its link ranges (filled in push_chat), so redraws
slice instead of rescanning/re-validating; only link spans allocate.
- app: chat history now also bounded by 512 KiB total sanitized text
(CHAT_HISTORY_MAX_TEXT_BYTES) alongside the 300-entry cap; the attachment
byte cache is deliberately untouched by history eviction (own budgets).
- app: AppMessage::OpenUrl re-checks the same parsed policy (defence in
depth) instead of prefix checks - non-web schemes can never reach the
opener even if the handler is invoked directly.
571 lib tests green (+3 net); clippy -D warnings + fmt clean.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Phase 3 of docs/chat-hardening-plan.md — attachments can no longer turn
into unbounded memory, bandwidth, decoder, or task pressure (S15 closed;
S14's filename half closed).
Cache and image cost (3A): AttachmentCache now carries encoded- and
decoded-byte budgets (96 MiB / 64 MiB) on top of the count cap, with
per-entry weights, replacement accounting, and oldest-first eviction; an
individually over-budget fetch services any pending Save/Play from the
bytes in hand and is exposed as Evicted instead of retained.
validate_image_bytes prechecks header dimensions (per-side AND a new
14 MP total-pixel limit) before any decode; the renderer only ever
receives a ≤1600 px downscaled RGBA preview whose w*h*4 cost counts
against the decoded budget — originals stay encoded-only for Save.
sanitize_filename strips the bidi/zero-width spoofing set (RTL-override
extension spoof).
Download policy and state (3B): images auto-fetch only when roster-
authored AND declared ≤4 MiB, gated by a new deterministic
AutoFetchBudget (per-author and session request+byte token buckets,
check-then-take, bounded author map) alongside the existing dedup and
four-permit bound. Attachment state is now explicit — absence/Loading/
Ready/Failed/Evicted — driven by a new AttachmentFetchStarted event, so
skipped or evicted images render a "Load image" button instead of an
indefinite "loading…", and repeated clicks can never spawn duplicate
fetch tasks.
Exact transfers and serve store (3C): fetch_blob requires the received
length to equal the declared size (short = local error, overlong =
bounded-read reject, empty keeps meaning "sender no longer has it");
the file picker's unbounded read is replaced by a metadata-prechecked
cap+1 bounded reader; one Arc<Vec<u8>> now backs the UI cache, command
queue, and serve store; served_files is a count- and byte-budgeted FIFO
ServeStore (16 entries / 128 MiB).
37 new tests (568 lib total) including a real two-endpoint loopback
exercising exact/short/overlong/unknown-id transfers. Plan checkboxes
ticked and constant deviations decision-logged. Tests-green-only: the
plan's two-machine field-test section remains open.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Chat-hardening plan Phase 2 — only current authenticated room members can
create chat UI work, impersonation via the wire name is structurally closed,
and no member can monopolize the event channel:
- core: new ChatRoster (bounded id -> sanitized-name map, shared) replaces the
event task's bare HashSet; upserted on PeerJoined/PeerUpdated, removed on
graceful PeerLeft AND terminal grace-expiry eviction (both timer paths).
Non-roster chat is dropped before attachment handling; the rendered author
label is the roster-bound name — the sender-claimed wire name is never read.
- gossip: ChatIngressGate after verify_gossip, before any sanitize work or
event send: early known-author gate (live + mid-reconnect peers), exact-
replay suppression keyed on the deterministic Ed25519 signature (1024-entry
cap + freshness-window TTL, zero new deps vs the plan's BLAKE3 option), then
per-author (8 burst, 1/s) and room-wide (32 burst, 8/s) token buckets.
Replays are detected before tokens are consumed; a room-bucket reject
refunds the author token; rejection logging is squelched per author.
- The inner Chat.ts is now ignored entirely; RoomEvent carries the signed
envelope timestamp.
550 lib tests (+18), reconnect_eviction +1 (grace keeps chat authority,
terminal eviction revokes it), clippy --all-targets -D warnings clean.
Tests-green-only: the plan's two-machine field-test section remains open.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Chat-hardening plan Phase 1. The chat body policy (2,000-char + 8 KiB
ceilings, single-pass control/whitespace normalization) moves from the UI
layer into src/sanitize.rs and is now enforced at every trust boundary:
cap_chat_input bounds the live input (oversized paste), the gossip sign
point re-sanitizes so non-UI callers can't bypass policy, and gossip
ingress rejects oversized raw text before sanitizing (admit_chat_text)
and drops messages with neither visible text nor an attachment. The
incoming chat author label now uses the strict name sanitizer until
Phase 2 roster-binds it. +8 tests (532 lib green).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
GPT-5.6's 5-phase plan for the chat identity/replay/rate-limit cluster
(2026-07-16 review findings 5-8): roster-bound display names, replay
dedup, quiet rate limiting, bidi-aware sanitization, attachment size
checks. Self-describes as temporary — delete when the work completes.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Answer "am I actually P2P right now?" per peer. A 1 Hz session task
snapshots the selected QUIC path of every live audio connection
(IrohTransport::connection_stats), core::connstats::derive turns
consecutive snapshots into RTT/loss/bitrate (path switches and counter
resets invalidate the rate window), and the peer card shows a
Direct/Relay badge with a hover tooltip for address, loss, and up/down
bitrate. No new dependencies, no wire change.
Loopback-integration-tested against real iroh endpoints; not yet
field-verified on a 2-machine call (FEATURES.md row marked 🧪).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Codex-authored refresh of docs/WINDOWS.md and packaging/windows/{README,INSTALL}.md
from the 2026-07-01 Windows session; scan.rs qualifies super::running_executables()
to drop an unused glob import. PKGBUILD pkgver reflects the last Arch build
(auto-regenerated by makepkg's pkgver()).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Implemented by Codex (gpt-5.5); reviewed and committed by Claude.
W8 — chimes were played by shelling out to pw-play/paplay/aplay, which don't
exist on Windows, so every chime silently no-op'd there. spawn_player is now
cfg-split: Linux/unix keeps the existing player list; Windows plays the WAV via
PowerShell's System.Media.SoundPlayer (PlaySync on the existing detached thread).
Dependency-free, same fire-and-forget / silent-on-failure contract. Custom chime
paths are single-quote-escaped for the PowerShell command (helper + unit test).
Also adds docs/WINDOWS.md: a build/run/status guide (native MSVC + cross-compile
to -gnu, first-run firewall/UDP note, %APPDATA% paths, and the honest known-gaps
table — echo-cancel/screenshare/resampling/device-id/buffer-pacing).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wire the Discoverable presence posture to n0 DNS publish/lookup, the last
core piece of W7 (friends-first contacts). When a friend moves networks and
their saved address goes stale, they flip Discoverable to publish their
current address; everyone else resolves it by node id. Asymmetric: only the
mover publishes.
- src/discovery.rs (pure seam, +3 tests): lookup_plan(network_mode, want_publish)
-> LookupPlan { resolver, publisher }. Relay-capable modes always resolve and
publish only when Discoverable; DirectOnly (the explicit no-server posture)
gets neither, overriding the toggle. DISCOVERY_TIMEBOX = 30 min.
- apply_discovery (core edge): clears + reinstalls the bound endpoint's
address-lookup services at runtime (no endpoint rebuild). memory-lookup always;
n0 PkarrResolver + DnsAddressLookup when resolver; PkarrPublisher when publisher.
Toggling publish off drops the publisher (republish task ends; TTL-30s record
expires). build_net_stack now binds uniformly with Minimal + per-mode relay and
installs discovery via apply_discovery (drops the per-mode presets::N0 build).
- Toggle + time-box: SetPresenceMode re-applies discovery and arms/cancels a
discovery_deadline; a select! branch fires at the deadline -> revert to Normal,
stop publishing, and emit UiEvent::PresenceModeReverted so the GUI mirrors and
persists it. Re-selecting Discoverable restarts the clock.
Decisions (user, 2026-06-16): 30-min auto-revert (not sticky); resolver always
on in relay-capable modes so a stationary friend in Normal can look up a mover.
266 lib tests green, clippy clean (--all-targets). Runtime smoke-tested: the new
Minimal+apply_discovery path binds and runs with no error/panic for both Normal
and Discoverable startup postures. Cross-network publish->lookup and the live
30-min revert still want a 2-machine field test (P7).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a purely-local, most-recent-first recents list so users can hop back
into a room they were just in — meaningful now that rooms carry cosmetic
labels.
- src/recents.rs (new): `Recent {name, ticket, joined_at}`, `push_recent`
(de-dupes by room `topic_id`, refresh-and-move-to-front, caps at
RECENTS_MAX=12), `remove_recent`, `relative_time` ("5m ago"). 6 tests.
- PeerSpeakTicket::topic_of — the stable room identity used as the de-dup
key (host addr + label change between members/sessions; topic doesn't).
- AppConfig.recents (`#[serde(default)]`, back-compat) — local UI state,
never sent over the wire.
- Recorded on RoomJoined (label via label_of); rendered as a "Recent
rooms" block in connect_card (each entry → JoinRecent, ✕ → RemoveRecent),
shown only when non-empty.
Rejoin is best-effort by design: the stored ticket only admits us while
the room is still live and reachable (reliability is P6 discovery + the
member-issued ticket floor, not this list).
263 lib tests green, clippy --all-targets clean. Recents UI
screenshot-verified (seeded config → ages + Untitled-room fallback render).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The friends-only presence listener (W7) must answer pings while the app is
open, whether or not we're in a call — but a node id has exactly one live
endpoint instance (proven by the dual-endpoint spike: two endpoints sharing a
SecretKey collide, all inbound connections land on one and the other's ALPN
fails the QUIC handshake). So the listener can't get its own endpoint; the
whole app must share one persistent endpoint. Today the core rebuilds the
endpoint+gossip+router on every Join and tears them down on leave, so there's
nothing alive between calls.
B1 hoists those durable pieces to the app lifetime (no new behavior):
- New persistent `NetStack` (endpoint + gossip + Router) built once at startup
under the RelayNoDiscovery default (relay reachability, no DNS beacon);
`online()` is backgrounded so launch isn't blocked.
- New persistent `AudioRouter` (src/network/iroh_impl.rs) replaces the
per-session `AudioProtocol`: it's registered once on the single Router and
delegates each inbound audio connection to whatever session `Shared` is bound
(`bind` on join, `clear` on leave), dropping links when idle. `IrohTransport::
new` now returns just `Self`.
- Join reuses `net.endpoint`/`net.gossip` and only subscribes its gossip topic +
binds the audio router; Leave clears the router but keeps the endpoint up.
- `SetNetworkMode`/`RegenerateIdentity` rebuild the stack immediately when idle,
else defer to the next Leave/Join (preserves "applies on next join"), and the
existing session is always torn down before any rebuild closes the endpoint.
Tests/loopback updated for the new transport API. 256 lib + 6 reconnect + 4
loopback + 2 ignored real-endpoint tests green, clippy --all-targets clean,
release builds. NOT yet 2-machine field-verified — that regression is the gate
before this merges to main.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Section 3 claimed playout is pinned to exactly 1024 frames; the code now
follows the graph's Buffer::requested() quantum (pipewire_impl.rs:229-237),
with 1024 only as a fallback -- the doc described the pre-fix behavior that
caused crackle. Flagged by the 2026-06-15 Codex/GPT-5.5 review (backlog A20).
The broader ARCHITECTURE refresh (missing modules, persistent identity, signed
gossip, friends/W7, recording modes) remains under A20.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Flagged by the 2026-06-15 Codex/GPT-5.5 review (backlog A20): the header
contradicted the per-phase statuses showing P1-P3 done + P4 partial. Now reads
'IN PROGRESS' with the current phase rollup.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Desktop<->dopedart on c5df7a3: persistent identity stable across restarts,
two-way audio regression clean, member-issued ticket lets a room outlive its
creator (desktop rejoins via dopedart's ticket after leaving), friends-add
persists. dopedart resynced to c5df7a3. Clears the P3 field-test.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Record the persistent-identity foundation as landed (d157d78) and add the
user's requirement: surface a persistent UI warning (not just a log) when the
key can't be read/written, since running on an ephemeral fallback silently
breaks friend recognition next launch. To build with the regenerate UI slice.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Rewrite the plan around the model converged in the 2026-06-15 design
session. Core shift: the friends list (stable node IDs) is the durable
anchor; rooms become ephemeral cosmetic labels, not addressable places.
Locked: persistent identity default-on + Settings regenerate; friends-only
idle listener (answers pings only from friends via remote_id, from a saved
address, no presence beacon); presence axis invisible/normal/discoverable
with discovery opt-in default-off and asymmetric (only the friend who moves
networks publishes); silent gossip-driven address auto-heal; and a universal
floor of hand-shared member-issued tickets that always connect.
Supersedes the heavyweight control-plane cut and the room-centric durable-room
cut; both retained at the bottom with reasons. ~3-4 sessions, build order
P1 identity -> P7 security/field-test. iroh does not force rolling IDs.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Pivot the contacts plan away from the heavyweight control-plane design
(always-on listener + persistent reachable identity + n0 DNS presence
beacon) toward a serverless, no-beacon spine: member-issued, multi-bootstrap
tickets. Any current member can mint a fresh ticket from their own live
address + the room's existing topic_id (both already in hand; join() already
takes extra_bootstrap: Vec), so rooms outlive their creator for ~no cost.
Layered optional add-ons: one-click invite sharing (peerspeak:// deep links
+ QR + .peerspeak files, delegating delivery to Signal/email/OS), persistent
identity (save-and-return), name-derived rooms, favorite-room bookmarks, and
a privacy-bounded silent occupancy peek. The old heavyweight path is retained
at the bottom of the doc as 'decided against', with the reasoning.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
molluskandClaude Opus 5