`docs/CONTRACT-mcp.md` goes to mcp-v3 with a change table: two tools, two resources, three `session_start` arguments, and the one rule that is not a name at all -- a connected viewer is the single exception to the idle-clock rule, and it cannot be invoked, so it is documented in §6 rather than offered in §3. v3 is additive; every v2 name still means what it meant. README.md leads the stream with the shape of it rather than the options: one port for the machine, a phrase per session, a browser `<img>` as the entire client, and the four things worth knowing before switching one on -- that it is per session and off by default, that the phrase survives a restart, that capture runs only while somebody is connected, and that watching counts as activity. The phrase is called what it is: the only protection, with no TLS and no second factor. ARCHITECTURE.md gains the fourth unit as a seventh box -- `hhstream-live`, hanging off the broker by `live.sock` with one port and a browser at the end of it -- and the counts move to 23 and 10. Two TODOs, both written by `checks/vm-stream.nix`: * `a-stream-that-gave-up-still-reports-attached.md`. The publisher stops a stream after ten consecutive capture failures and the stop is sticky on purpose, but `Control._describe` does not pass `stopped` on, so `_stream_public` reports `attached: true` for a capture that has permanently given up. That is the second place `attached` answers "is a capture running right now" wrongly, and a field that is right only when nothing is wrong is not worth reading. * `capture-binary-and-the-unit-path.md` is now half fixed, and says so. The module passes grim's store path and this check reads real frames -- which immediately found the sibling fault, the unit's missing `XDG_RUNTIME_DIR`. What is still open is the other half: the app resolves its capture command by name and never checks it, so any deployment that is not this module still discovers a broken capture ten failures later. The plan's status section records Stage 2 and Stage 3 as built, the two decisions §5 left open (that `start` takes `--stream`, and that the default rate is the daemon's), and what Stage 2 deliberately did not do -- §6's restart re-attachment, which has its own TODO. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_011ZeH65zd6pFTgBSu52W9PE
3.8 KiB
The capture command needs an absolute path under a unit
The issue
hhstream-live runs its capture command (grim -c -t jpeg -q 80 -o <output> -)
by name, from the PATH it inherits. Under systemd that path is
/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin — the system
profile is deliberately not on it — so the capture fails, ten times, and the
stream gives up:
WARNING hhstream.live: stream on hh-chromium-yhvpms: giving up after 10
consecutive capture failures: capture command ['grim', '-c', '-t', 'jpeg',
'-q', '80', '-o', 'hh-chromium-yhvpms', '-'] could not be run:
[Errno 2] No such file or directory: 'grim'
Measured on 2026-10-11 on the deployment machine, the first time the app was
run as a unit rather than from a login shell. A login shell has
/run/current-system/sw/bin on its PATH, which is why every run before this
one — the stage-1 work, the standalone serve --stream runs, the whole test
suite — never saw it.
What makes it worth a file rather than a line: the failure is quiet in the
worst way. The unit starts, the port opens, the landing page answers, and
status.json reports a stream that is simply "stopped": true with
"failures": 10. A viewer sees an empty frame area and nothing else says why.
Only the unit's log names the reason, one line above the giving-up entry.
Half of this is now fixed. The module passes grim's absolute store path
(idea 1), checks/vm-stream.nix starts the unit and reads real JPEGs off the
stream — the check this file asked for — and that check immediately found the
sibling fault: the unit had the compositor's group but not its environment,
so grim exited 1 with "XDG_RUNTIME_DIR is invalid or not set", ten times, with
the same silence. Fixed too (execLive). What is still open is idea 2: the app
resolves its capture command by name and never checks it, so a standalone
hhstream-live serve and any deployment shape that is not this module still
discover a missing or unusable capture binary ten failures later. The quiet
half — a stream that gave up while the broker still reports it attached — has
its own file now:
a-stream-that-gave-up-still-reports-attached.
Ideas
- An absolute binary, as the module already does elsewhere (recommended).
sessions.appsrequires an absolute store path for exactly this reason — "a unit's defaultPATHdoes not include the system profile", which is why the allowlist spellsfootas${pkgs.foot}/bin/foot. The stream's unit should not inheritgrimby name either: pass${pkgs.grim}/bin/grimas the capture command from the module (or setPATHon the unit, which is what made it work on the deployment machine). The absolute path is also the stronger control: it names one exact binary instead of whateverPATHresolves to at the moment the unit starts. - Resolve early and fail loudly. Have the app resolve the binary itself at
start-up and refuse to serve with a named error (
capture_binary_missing) if it is not there. A unit that dies at start is a five-second diagnosis; a unit that serves an empty frame is an afternoon. - Both. The module passes the absolute path and the app checks it once at start-up: the first is the fix, the second is what catches the next deployment shape nobody has thought of yet.
Why it matters
Stage 3's unit is the point at which running under systemd stops being an
experiment and becomes the only path every deployment takes — and a stream
that answers, reports a healthy port, and shows nothing is a worse failure than
one that does not start. Whichever idea is taken, the check for it belongs in
the same stage as the unit: a vm-* check that starts the unit and reads one
frame would have caught this on the day it was written.