gui-dod (per-PR: scroll-60fps + unhappy-path) and gui-dod-nightly (rss-flat,
schedule/workflow_dispatch) are live in ci.yml, driving GUI's newly-landed
gui/tests/dod/run.sh + gui-dod-harness. No Xvfb step: run.sh already runs
QT_QPA_PLATFORM=offscreen itself.
Each gate forced red once before being trusted (tests/integration/README.md
has the transcripts): VELOX_DOD_FRAME_BUDGET_MS=0.01 for scroll-60fps,
VELOX_DOD_RSS_SLACK_KIB=-999999999 for rss-flat, and — since run.sh always
starts a working mockd — a direct gui-dod-harness invocation against an
unreachable socket for unhappy-path, which hit the harness's own 75s
watchdog exactly as documented.
Recorded GUI's live finding (gui/docs/proto-requests-m1.md) that mockd
--drop-connection is a no-op over the UDS transport, so unhappy-path's
drop-connection phase can't yet exercise a real drop — coordinating with
PROTO on the fix rather than working around it locally. gui-dod stays
required regardless: its other two phases and the crash/hang/watchdog paths
still catch real regressions.
Added gui-dod to BRANCH_PROTECTION.md's required-checks table.
ADR 0019: the live-veloxd conformance runner (run.sh step 3b, already
unconditional inside the already-required conformance job) stays required
as PROTO's xfail list shrinks (18 entries now, down from 34; 57/57 fixtures
passing on main). No CI change needed — it was already inside a required
check; this records the decision not to carve out an exception for it.
Co-Authored-By: Claude Sonnet 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01RBPR7iM3YPyxrjWsVtZDPJ
Adds tests/integration/nightly_run.py, wired as the nightly-integration job in
ci.yml (schedule + workflow_dispatch). Real veloxd + tools/testserver, 50
concurrent downloads mixing flaky-reset/throttled/no-range/plain, asserting:
every completed file's SHA-256 against testserver's own /sha256/ route (never
trusting veloxd's own success claim), every task reaching a terminal state
inside the timeout, and veloxd's own open-FD count settling back to baseline.
veloxd runs isolated (XDG_RUNTIME_DIR/XDG_DATA_HOME/XDG_CONFIG_HOME under a
fresh mkdtemp — not the session scratch dir, whose path overflows AF_UNIX's
sun_path). saveTo.allowedRoots is seeded directly into velox.db after a
migrations-only warm-up start, since settings.set returns -32603 'not
implemented in this build' on the veloxd this job builds (verified live).
Every task gets its own filename override on download.add: tasks sharing
(mode, size) share a URL, and without distinct filenames they raced each
other's rename on the first real run (48/50 'passed' with io_errors and
checksum mismatches on the collided tasks) before this fix.
Every assertion was forced red once on purpose and the transcript recorded in
tests/integration/README.md, per this repo's history of green checks that
didn't look where the bug was.
Every spawned child (veloxd, testserver.py) gets PR_SET_PDEATHSIG plus its own
process group, so a hard-killed harness can't strand a daemon the way a prior
run did (4h40m under systemd --user, because SIGKILL never reaches a
finally: block). Proven by kill -9'ing a running harness mid-download and
confirming both children exit with it.
Co-Authored-By: Claude Sonnet 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_01RBPR7iM3YPyxrjWsVtZDPJ
R3: the three GUI DoD gates (10k rows at 60fps, flat RSS over 10 min,
--slow/--flaky/--drop-connection recovery) have nowhere to run. GUI owns the
harness, PKG/QA owns the job. tests/integration/README.md records the wiring
contract — driver invocation, exit-code and --json semantics — and carries
the pre-drafted per-PR and nightly job stanzas with TODO(GUI) markers for the
harness path. Wire for real when GUI files the follow-up.
Co-Authored-By: Claude Sonnet 5 <[email protected]>
Claude-Session: https://claude.ai/code/session_0143aKiohmDiyefJBwHDJJqw
samiandClaude Sonnet 5