fix(live): a dead output stops the stream even after a good frame

The failure cap read `consecutive_failures >= FAILURE_LIMIT and captured ==
0`, so it could only ever fire on a cold start. An output removed while
someone was watching -- grim answering "unknown output" to every grab from
then on -- kept the loop retrying twice a second for as long as the viewer
held the page open. That is the busy loop the cap exists to stop, so the
count is now of failures in a row and nothing else. Giving up stays sticky:
a removed output does not come back, and re-enabling the session is how it
returns.

While here, the capture thread is named for its output and the give-up is
logged by output. The phrase is the stream's only credential, and
faulthandler, `threading.excepthook` and any sampling profiler all print
thread names.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01URhyHGQnmKH9zBXxoznfQY
This commit is contained in:
HermenixandClaude Opus 5 committed 2026-10-06 16:19:32 +00:00
1 parent 9744d35945
commit 52e3786bfb
3 files changed
+92 -10

No files matched your search

+12 -2
View File
@@ -25,11 +25,19 @@ PHRASE = "amber-otter-lantern-harbor-7k2qp"
class StubCapture:
"""A `Capture` under the test's control: counts grabs, can be told to fail."""
"""A `Capture` under the test's control: counts grabs, can be told to fail.
def __init__(self, frames=None, fail_first: int = 0, delay: float = 0.0):
`fail_first` is a capture that is broken from the start (a cold start that
never works); `fail_after` is one that works and then stops -- the output
was removed while someone was watching, which is grim's "unknown output" on
every grab from then on.
"""
def __init__(self, frames=None, fail_first: int = 0, delay: float = 0.0,
fail_after: int | None = None):
self.frames = list(frames) if frames else [jpeg_bytes(i) for i in range(1, 50)]
self.fail_first = fail_first
self.fail_after = fail_after
self.delay = delay
self.calls = 0
self._lock = threading.Lock()
@@ -42,6 +50,8 @@ class StubCapture:
time.sleep(self.delay)
if call <= self.fail_first:
raise RuntimeError(f"stub capture failing on purpose (call {call})")
if self.fail_after is not None and call > self.fail_after:
raise RuntimeError(f"stub capture: unknown output (call {call})")
return self.frames[(call - 1 - self.fail_first) % len(self.frames)]