feat(fetch): count bytes on the wire as well as decoded

fetch-stats reported only `bytes`, len(response.content), and that read as
the download volume. ESPN gzips every scoreboard, so it overstated real
traffic about 14x: a college football Saturday is 865 KB decoded, 63 KB on
the wire. Every counter set now carries `wire_bytes`, read from urllib3's
count of raw bytes taken off the socket (decoded size when there is no
urllib3 response behind it).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

CHANGELOG.md

@@ -19,6 +19,56 @@ accepts both, but the store flags the old spelling as deprecated
 
 ## Unreleased
 
+### Fixed
+
+- The web preview and `/api/v3/display/current` no longer stay black for a
+  whole screen that draws its card once and then holds it. The snapshot is
+  written from `update_display()` at most once per write interval, so a frame
+  pushed inside that interval was skipped and left for the next
+  `update_display()` -- which such a screen never makes. Soccer's
+  recent/upcoming cards skip redundant redraws, and the first one after an
+  on-demand start lands a few milliseconds after the start's clear wrote a
+  black frame: on ledpi the preview showed 0 lit pixels for the whole 15 s
+  while the panel showed the card. `DisplayManager` now remembers a skipped
+  changed frame, and the render loop writes it (`write_owed_snapshot()`)
+  once the interval has passed. The cadence is unchanged, and nothing extra
+  runs when no frame is owed.
+
+### ESPN date-range fetches: fewer requests, fewer at once
+
+A soccer board (8 leagues, ESPN rejecting `dates=` ranges) logged ~90
+`NameResolutionError` lines and an `update() timed out` at every start on a
+Pi: each league's fortnight-either-side window was 29 day requests, fetched
+by several managers at once, ~40 in flight. Measured against live ESPN with
+soccer-scoreboard 2.39.2, alternating runs: **~450 requests per start, peak
+~45 in flight, ~75 DNS lookups -> 46 requests, peak 13, ~30 lookups**.
+
+- `fetch_espn_date_chunks()` asks for a window's partial edge month whole
+  when the window covers `ESPN_MONTH_COVER_MIN_DAYS` (7) or more of its days,
+  and trims the answer to the window's days by each event's US Eastern start
+  date -- the day ESPN's `dates=YYYYMMDD` means (417 of 417 live soccer
+  events matched). A 29-day window spanning two months is 2 requests instead
+  of 29. Short windows (a live poll's 1-2 days) stay day by day. A trimmed
+  month that comes back at the 500-event cap re-asks only the window's days.
+  An event with no readable date is kept. New: `espn_request_chunks()`.
+- Chunk requests share one process-wide cap of `ESPN_CHUNK_WORKERS` (6) in
+  flight, across every window being fetched, instead of six per window.
+- A new process starts as if a range had just been rejected, so it no longer
+  spends one doomed 400 per window at every start (eleven at once from a
+  soccer board); the range is still retried `RANGE_RETRY_SECONDS` in.
+
+### Fetch stats: bytes on the wire, not just decoded
+
+`GET /api/v3/plugins/fetch-stats` reported only `bytes`, the decoded body
+size, and that read as the download volume. ESPN gzips every scoreboard, so
+it overstated what crossed the network about 14x: a college football
+Saturday's scoreboard is 865 KB decoded and 63 KB on the wire, and ledpi's
+"643 MB in 6 hours" of football was ~47 MB of actual traffic. Every counter
+set (totals, per plugin, per host) now has `wire_bytes` too, read from
+urllib3's count of the raw bytes it took off the socket. A response with no
+urllib3 response behind it is counted at its decoded size. `bytes` keeps its
+meaning.
+
 ### Cheap per-frame and per-fetch savings
 
 - `BaseOddsManager.get_odds()` no longer pretty-prints every odds response
@@ -505,6 +555,54 @@ policies are unchanged.
   stored `ttl` was stretched the same way. A memory hit is now also checked against
   the record's own timestamp, and a stale one falls through to disk, which
   returns a newer write if there is one.
+- An on-demand request that names a `*_live` mode now shows that mode. On
+  ledpi, `{"plugin_id": "football-scoreboard", "mode": "ncaa_fb_live"}` with
+  15 college games on answered 200 and showed `nfl_recent`. The session's
+  mode list kept a live mode only when the plugin's `has_live_content()`
+  said so. That method answers the live-priority question, and the sports
+  plugins answer it for favourite teams only. A mode the request names
+  (not one resolved from a bare plugin id) now leads the session, with the
+  plugin's other modes after it. If it has nothing to draw, the session
+  moves on to the next of those modes, like any empty on-demand mode. The
+  name is saved with the session (`named_mode` in
+  `display_on_demand_config`), so a restart resumes on it.
+- A restart during an on-demand session whose plugin then fails to load no
+  longer leaves a session with no modes. On ledpi, `clock-simple` failed
+  config validation after a crash. The display logged `No valid display
+  modes found for on-demand plugin 'clock-simple' after restoration` and
+  kept reporting the session as active until its first pass ended it as
+  `idle`. The cached request stayed behind for the next restart. The session
+  now ends at startup with status `error` and error `restore-failed`, which
+  `/display/on-demand/status` reports, and the cached request is dropped. The
+  same applies when the plugin system itself fails to start.
+- `POST /api/v3/config/schedule` and `/config/dim-schedule` accept a
+  disabled per-day schedule with every day off. That is the shape
+  `config.template.json` ships, so posting back what GET returned on a fresh
+  install answered 400 "At least one day must be enabled". An enabled per-day
+  schedule still needs a day on. A day that is off now keeps the times it
+  was posted with (the schedule picker sends them). Before, saving dropped
+  them, so turning the day back on showed the defaults.
+- `POST /api/v3/config/main` answers `restart_required: true` only when the
+  save changed a setting the running display does not apply by itself.
+  Brightness (`brightness.set` and the config watcher), the per-mode
+  durations and plugin sections are applied live. A brightness-only save,
+  such as the MQTT bridge's slider, or a save that changed nothing, no longer
+  shows the restart banner. Hardware, rotation order, timezone and every
+  other setting still ask for the restart.
+- `GET /api/v3/health` reports `degraded` when the display service is
+  stopped. Before, only the sub-checks changed, and the overall status stayed
+  `healthy` for as long as the last preview frame was under 60 s old.
+  `checks.display_loop.status` is now `stopped` when three things agree:
+  systemd says the service is not active, the control socket does not
+  answer, and there is no live heartbeat. Where the platform has no socket
+  (Windows) or it is switched off, nothing changes.
+- `GET /api/v3/display/current-status` no longer reports the stopped
+  display's last state (`is_display_active: true`) from the cache for up to
+  120 s. When the control socket does not answer and the render loop's
+  heartbeat is absent, stale, or from a process that is gone (#726's rules),
+  the answer is unknown, with every field `null`. A display that still beats
+  without a socket, Windows and a socket switched off read the cache as
+  before. New `web_interface.display_state.display_gone()`.
 - The garbage-collection timer (`GcMonitor`, above) no longer prints
   `Exception ignored while calling GC callback ... 'NoneType' object has no
   attribute 'perf_counter'` when the display service or a test run exits.

docs/REST_API_REFERENCE.md

@@ -159,14 +159,16 @@ there an unchecked checkbox — which the browser omits — is saved as
 }
 ```
 
-`restart_required` is always true here: display hardware, rotation,
-durations and general settings take effect when the display restarts, and
-the web UI shows its restart banner on the flag. (Plugin sections saved
-through this route reach the running plugin live, like
-`POST /plugins/config`.)
+`restart_required` is true when the save changed a setting that takes
+effect when the display restarts: display hardware, rotation order,
+timezone, general settings and the rest. The web UI shows its restart banner
+on the flag. It is false when the save changed only what the running display
+applies by itself, or nothing: `brightness`, the per-mode durations
+(`duration__<mode>`, `display.display_durations`) and plugin sections, which
+reach the running plugin live, like `POST /plugins/config`.
 
-A saved `brightness` is the exception: it reaches the panel without a
-restart. The route also sends it to the running display over the control
+A saved `brightness` reaches the panel without a restart. The route also
+sends it to the running display over the control
 socket (`brightness.set`), which puts it on the panel at once, and the
 response adds `"brightness_transport": "socket"`. Otherwise it is
 `"config"`, with `brightness_socket_error` giving the reason, and the
@@ -246,7 +248,10 @@ Replace the schedule configuration.
 ```
 
 A day whose `<day>_enabled` key is absent counts as enabled, with default
-times `07:00`-`23:00`. At least one day must be enabled.
+times `07:00`-`23:00`. An enabled schedule needs at least one day enabled; a
+disabled one (`"enabled": false`) may have every day off, as
+`config.template.json` ships it. A day that is off keeps the times sent for
+it, when they are valid `HH:MM`.
 
 **Response**:
 ```json
@@ -343,7 +348,11 @@ control socket ([IPC_CONTROL_SOCKET.md](IPC_CONTROL_SOCKET.md)), and `cache`
 when it came from the `display_current_state` cache key (no socket: the
 display is stopped or older, or this is Windows). A display whose render
 loop has not refreshed its state for 120 seconds is reported with every
-field `null`, either way.
+field `null`, either way. So is a stopped display: when the socket does not
+answer and the render loop's heartbeat
+(`/run/ledmatrix/display-heartbeat.json`) is absent, stale or from a process
+that is gone, the cache's last entry is not used. A display still beating
+without a socket, Windows, or a socket switched off reads the cache.
 
 ### List Display Modes
 
@@ -2281,7 +2290,11 @@ display snapshot. `data.status` is `healthy` or `degraded`, with
 (with `heartbeat_age_seconds`), `stalled` (no heartbeat for 60s: the panel is
 frozen even if the service is active; the status turns `degraded`), or
 `not_reported` when the display writes none (not started yet, the dev server,
-Windows), which does not affect the status. Its `source` is `socket` when the
+Windows), which does not affect the status, or `stopped` (with `source:
+"service"`) when the display service is not active, the control socket does
+not answer and there is no live heartbeat; the status then turns
+`degraded`. A platform with no control socket (Windows) or a socket switched
+off never reports `stopped`. Its `source` is `socket` when the
 age came from the display's state stream over the control socket (measured
 in memory by the display) and `heartbeat_file` when it came from
 `/run/ledmatrix/display-heartbeat.json`.
@@ -2344,8 +2357,9 @@ Replace the dim schedule. `dim_brightness` is 0-100 (default 30). In
 `per-day` mode the days can be sent either as the `days` object that GET
 returns, or as the web form's flat fields (`monday_enabled`,
 `monday_start`, `monday_end`, ...). A day that is not sent counts as
-enabled with default times `20:00`-`07:00`; at least one day must be
-enabled.
+enabled with default times `20:00`-`07:00`. As for the schedule above, an
+enabled dim schedule needs at least one day enabled and a disabled one may
+have every day off.
 
 ---
 

docs/RUN_LOOP_REDESIGN.md

@@ -181,13 +181,14 @@ that the harness patches in today.
   - dynamic duration (cycle complete, plugin cap, global cap)
   - live priority taking over and handing back; live round-robin
   - on-demand start/stop/expiry; pinned on-demand; a session resumed after
-    a restart
+    a restart, and one that cannot resume (its plugin did not load); a
+    request naming a live mode the plugin's live check would drop
   - schedule off and dim, with an on-demand override during downtime
   - WiFi notice; sync follower
   - Vegas, with and without `live_in_ticker`
 - Each trace row is `[start, mode, duration, exit_reason, frames,
   force_clear]`. The exit reason is the event that decided what came next.
-- All 16 tests run in under a second. The goldens were generated from
+- All 18 tests run in under a second. The goldens were generated from
   main's `run()` before any code moved.
 - Vegas uses `FakeVegas`, which implements only the contract the controller
   depends on: `run_iteration()` returns True after its duration and False

src/common/README.md

@@ -109,8 +109,11 @@ and the plugin test harness all use it. Most plugins get BDF text through
 [`espn_dates.py`](espn_dates.py). ESPN's site API rejects `dates=` ranges
 and truncates results when `limit` is above 500. `fetch_espn_scoreboard()`
 splits a range into month and day requests ESPN accepts and merges the
-results; `espn_date_chunks()`, `fetch_espn_date_chunks()`,
-`clamp_espn_limit()` and `merge_scoreboard_payloads()` are the pieces.
+results; `espn_date_chunks()`, `espn_request_chunks()`,
+`fetch_espn_date_chunks()`, `clamp_espn_limit()` and
+`merge_scoreboard_payloads()` are the pieces. A window's partial edge months
+are asked whole and trimmed to its days (US Eastern), and chunk requests share
+one process-wide cap of `ESPN_CHUNK_WORKERS` in flight.
 Every request goes through [`fetch_service`](#fetch_service), the chunks
 counted against the plugin that asked. Scoreboard plugins also bundle a copy
 for older cores.

src/common/espn_dates.py

@@ -26,10 +26,30 @@ A month can hold more than 500 events (college baseball's March does), and
 ESPN answers that with exactly ``limit`` events and no hint that more exist. A
 month chunk that comes back full is therefore re-asked day by day.
 
+A window's *partial* edge months are asked for whole, too, once the window
+covers ``ESPN_MONTH_COVER_MIN_DAYS`` or more of their days, and the answer is
+trimmed back to the window's days. A scoreboard's default fortnight either side
+of today (29 days, two partial months) was 29 day requests per league; it is
+now 2. Trimming needs ESPN's "game day", which is the event's start in US
+Eastern time -- checked against the live API on 2026-10-03: 417 of 417 soccer
+events across five leagues and three months (one of them spanning the end of
+daylight saving) came back from exactly the day query their Eastern date
+names. A short window (a live poll's one or two days) stays day by day, so it
+never downloads a whole month to read a day of it.
+
+Chunk requests share one process-wide budget of ``ESPN_CHUNK_WORKERS`` in
+flight, however many windows are being fetched at once. Each window used to get
+its own six, so a scoreboard starting eight leagues -- each with a recent and
+an upcoming manager -- had ~40 requests in flight, every one beyond a session's
+pool a new connection and a new DNS lookup. On a Pi whose resolver could not
+keep up, that was ~90 ``NameResolutionError`` lines within a minute of every
+start.
+
 Once a range has been rejected, later ranges skip straight to chunks for
 ``RANGE_RETRY_SECONDS`` instead of spending a doomed request first -- live
 scoreboards ask every 30 seconds. After that the range is tried again, so the
-workaround retires itself if ESPN reverts.
+workaround retires itself if ESPN reverts. A process starts inside that
+period, as if a range had just been rejected.
 
 ONE CACHE KEY PER SCOREBOARD
 ----------------------------
@@ -55,7 +75,7 @@ import re
 import threading
 import time
 from concurrent.futures import ThreadPoolExecutor
-from datetime import date, datetime, timedelta
+from datetime import date, datetime, timedelta, tzinfo
 from functools import partial
 from typing import Any, Callable, Dict, Iterable, List, Optional, Tuple, cast
 
@@ -100,16 +120,57 @@ RANGE_RETRY_SECONDS = 6 * 60 * 60
 # pool_maxsize of 10 so the shared Session never has to discard connections.
 ESPN_CHUNK_WORKERS = 6
 
+#: An edge month the window covers at least this many days of is asked for
+#: whole and trimmed, instead of one request per day (see module docstring).
+#: Below it the days are cheaper than the month: a whole month is two to
+#: three times the bytes of the half of it a fortnight window holds.
+ESPN_MONTH_COVER_MIN_DAYS = 7
+
+# Every chunk request in the process holds one of these while it is in flight
+# -- the cap is per process, not per window (see module docstring).
+_chunk_slots = threading.BoundedSemaphore(ESPN_CHUNK_WORKERS)
+
+
+def _eastern_zone() -> Optional[tzinfo]:
+    """US Eastern, the zone ESPN's ``dates=YYYYMMDD`` means, or None when
+    this Python has no time zone data (no edge month is trimmed then)."""
+    zone: Optional[tzinfo] = None
+    try:
+        from zoneinfo import ZoneInfo
+        zone = ZoneInfo("America/New_York")
+    except Exception:  # noqa: BLE001 - no zoneinfo module or no tz database
+        zone = None
+    if zone is not None:
+        return zone
+    try:
+        import pytz
+        return cast(tzinfo, pytz.timezone("America/New_York"))
+    except Exception:  # noqa: BLE001
+        return None
+
+
+_EASTERN = _eastern_zone()
+
+# What _fetch_one_chunk returns for a month that came back at the cap.
+_CAPPED: Any = object()
+
 _range_lock = threading.Lock()
-_ranges_rejected_until = 0.0
+# A process starts out assuming ranges are still rejected, as they have been
+# since 2026-09-15, and tries one again RANGE_RETRY_SECONDS in. Starting
+# from "unknown" cost one doomed range request per window at every start --
+# eleven 400s at once from a soccer board, each fetching before any had
+# answered -- to learn what every start learns.
+_ranges_rejected_until = time.monotonic() + RANGE_RETRY_SECONDS
 
 __all__ = [
     "ESPN_MAX_LIMIT",
     "ESPN_CHUNK_WORKERS",
+    "ESPN_MONTH_COVER_MIN_DAYS",
     "RANGE_RETRY_SECONDS",
     "clamp_espn_limit",
     "parse_espn_date_range",
     "espn_date_chunks",
+    "espn_request_chunks",
     "merge_scoreboard_payloads",
     "fetch_espn_date_chunks",
     "fetch_espn_scoreboard",
@@ -220,6 +281,79 @@ def espn_date_chunks(start: date, end: date) -> List[str]:
     return chunks
 
 
+def espn_request_chunks(
+    start: date,
+    end: date,
+    month_cover_min_days: Optional[int] = None,
+) -> List[Tuple[str, Optional[Tuple[date, date]]]]:
+    """The requests that fetch ``[start, end]``, as ``(dates, trim)`` pairs.
+
+    :func:`espn_date_chunks`, except that a partial edge month with
+    ``month_cover_min_days`` (default ``ESPN_MONTH_COVER_MIN_DAYS``) or more
+    of its days in the window becomes one ``YYYYMM`` request whose ``trim``
+    is the first and last of those days: its events that start outside them
+    (US Eastern) are dropped. ``trim`` is None for every other request.
+    Without time zone data nothing can be trimmed, so the edge days stay day
+    requests.
+    """
+    if month_cover_min_days is None:
+        month_cover_min_days = ESPN_MONTH_COVER_MIN_DAYS
+    planned: List[Tuple[str, Optional[Tuple[date, date]]]] = []
+    run: List[str] = []
+
+    def flush() -> None:
+        if (_EASTERN is not None and month_cover_min_days > 0
+                and len(run) >= month_cover_min_days):
+            planned.append((run[0][:6], (_parse_day(run[0]), _parse_day(run[-1]))))
+        else:
+            planned.extend((day, None) for day in run)
+        run.clear()
+
+    for chunk in espn_date_chunks(start, end):
+        if run and (len(chunk) != 8 or chunk[:6] != run[0][:6]):
+            flush()
+        if len(chunk) == 8:
+            run.append(chunk)
+        else:
+            planned.append((chunk, None))
+    flush()
+    return planned
+
+
+def _parse_day(text: str) -> date:
+    return date(int(text[:4]), int(text[4:6]), int(text[6:8]))
+
+
+def _eastern_day(stamp: Any) -> Optional[date]:
+    """The US Eastern date of an ESPN event ``date`` ("2026-10-10T11:30Z"),
+    or None when it cannot be read."""
+    if not isinstance(stamp, str) or _EASTERN is None:
+        return None
+    try:
+        moment = datetime.fromisoformat(stamp.strip().replace("Z", "+00:00"))
+    except ValueError:
+        return None
+    if moment.tzinfo is None:
+        return None
+    return moment.astimezone(_EASTERN).date()
+
+
+def _trim_to_days(payload: Any, first: date, last: date) -> Any:
+    """Drop the events of a month payload that start outside ``[first, last]``
+    (US Eastern). An event whose date cannot be read is kept: its day query
+    might well have returned it, and a game is never dropped on a guess.
+    """
+    if not isinstance(payload, dict) or not isinstance(payload.get("events"), list):
+        return payload
+    kept = []
+    for event in payload["events"]:
+        day = _eastern_day(event.get("date")) if isinstance(event, dict) else None
+        if day is None or first <= day <= last:
+            kept.append(event)
+    payload["events"] = kept
+    return payload
+
+
 def merge_scoreboard_payloads(payloads: List[Any]) -> Dict[str, Any]:
     """Fold chunk responses into one scoreboard payload.
 
@@ -250,37 +384,54 @@ def merge_scoreboard_payloads(payloads: List[Any]) -> Dict[str, Any]:
 def _fetch_one_chunk(
     session, url: str, params: Dict[str, Any], headers, timeout, logger, chunk: str,
     cache_max_age: Optional[float] = None,
-) -> Optional[Dict[str, Any]]:
+    trims: Optional[Dict[str, Tuple[date, date]]] = None,
+) -> Any:
     """GET a single ``dates=`` chunk, or None when it failed.
 
     One bad chunk must not sink the rest of the season, so every error is
     logged and swallowed here rather than raised to the gather below.
+
+    A month that comes back at the cap is truncated: it returns ``_CAPPED``,
+    its payload dropped here before it is ever held beside the others. A
+    month in ``trims`` loses its events outside the days given there.
+
+    The request holds one of the process-wide ``_chunk_slots`` while it runs.
     """
     try:
-        response = fetch_get(
-            session,
-            url,
-            params=dict(params, dates=chunk, limit=ESPN_MAX_LIMIT),
-            headers=headers,
-            timeout=timeout,
-            **_memo_kwargs(cache_max_age),
-        )
-        response.raise_for_status()
-        return cast(Optional[Dict[str, Any]], response_json(response))
+        with _chunk_slots:
+            response = fetch_get(
+                session,
+                url,
+                params=dict(params, dates=chunk, limit=ESPN_MAX_LIMIT),
+                headers=headers,
+                timeout=timeout,
+                **_memo_kwargs(cache_max_age),
+            )
+            response.raise_for_status()
+            payload = response_json(response)
     except Exception as exc:  # noqa: BLE001 - see docstring
         if logger:
             logger.warning("ESPN chunk %s failed, skipping it: %s", chunk, exc)
         return None
+    if len(chunk) == 6 and isinstance(payload, dict):
+        if len(payload.get("events") or []) >= ESPN_MAX_LIMIT:
+            return _CAPPED
+        trim = (trims or {}).get(chunk)
+        if trim is not None:
+            payload = _trim_to_days(payload, *trim)
+    return payload
 
 
 def _fetch_chunks(
     session, url: str, params: Dict[str, Any], headers, timeout, logger,
     chunks: List[str], cache_max_age: Optional[float] = None,
-) -> List[Optional[Dict[str, Any]]]:
+    trims: Optional[Dict[str, Tuple[date, date]]] = None,
+) -> List[Any]:
     """Fetch every chunk, returning payloads positionally aligned with ``chunks``.
 
     Requests go out ``ESPN_CHUNK_WORKERS`` at a time because a cold season is
-    over a hundred of them. The order they come back in is not significant --
+    over a hundred of them -- and no more than that across every window the
+    process is fetching, which ``_fetch_one_chunk``'s slot enforces. The order they come back in is not significant --
     callers keep ``chunks`` order from the returned list -- but it does mean
     the session is shared across threads, which is why this only ever issues
     GETs and never touches session state.
@@ -293,7 +444,7 @@ def _fetch_chunks(
         return []
     fetch = partial(
         _fetch_one_chunk, session, url, params, headers, timeout, logger,
-        cache_max_age=cache_max_age,
+        cache_max_age=cache_max_age, trims=trims,
     )
     if len(chunks) == 1:
         return [fetch(chunks[0])]
@@ -340,7 +491,9 @@ def fetch_espn_date_chunks(
     if span is None:
         return None
 
-    chunks = espn_date_chunks(*span)
+    planned = espn_request_chunks(*span)
+    chunks = [chunk for chunk, _ in planned]
+    trims = {chunk: trim for chunk, trim in planned if trim is not None}
     if logger:
         logger.debug(
             "Fetching ESPN date range %s as %d month/day chunks",
@@ -349,32 +502,31 @@ def fetch_espn_date_chunks(
 
     results = _fetch_chunks(
         session, url, params, headers, timeout, logger, chunks, cache_max_age,
+        trims,
     )
     attempted = len(chunks)
 
-    # A month that came back at the cap is truncated; its days replace it in
-    # place, so merged events stay in chunk order however the requests raced.
+    # A month that came back at the cap is truncated; its days (only the
+    # window's, for a trimmed edge month) replace it in place, so merged
+    # events stay in chunk order however the requests raced. Its payload was
+    # already dropped in the worker: a capped college-baseball month is ~2MB
+    # of parsed JSON, and holding four of them through ~120 day requests added
+    # ~25MB to the peak -- more than the concurrency itself. Low-memory boards
+    # (docs/LOW_MEMORY_BOARDS.md) have under 200MB of headroom.
     slots: List[Any] = results
     capped: Dict[int, List[str]] = {}
     for index, chunk in enumerate(chunks):
-        payload = slots[index]
-        if payload is None or len(chunk) != 6:
+        if slots[index] is not _CAPPED:
             continue
-        events = payload.get("events") if isinstance(payload, dict) else None
-        if len(events or []) >= ESPN_MAX_LIMIT:
-            if logger:
-                logger.info(
-                    "ESPN month %s hit the %d-event cap; re-asking it day by day",
-                    chunk, ESPN_MAX_LIMIT,
-                )
-            capped[index] = _days_of_month(chunk)
-            # Drop the truncated month now rather than after its days arrive:
-            # a capped college-baseball month is ~2MB of parsed JSON, and
-            # holding four of them through ~120 day requests added ~25MB to
-            # the peak -- more than the concurrency itself. Low-memory boards
-            # (docs/LOW_MEMORY_BOARDS.md) have under 200MB of headroom.
-            slots[index] = None
-    payload = events = None
+        if logger:
+            logger.info(
+                "ESPN month %s hit the %d-event cap; re-asking it day by day",
+                chunk, ESPN_MAX_LIMIT,
+            )
+        trim = trims.get(chunk)
+        capped[index] = (_days_of_month(chunk) if trim is None
+                         else espn_date_chunks(*trim))
+        slots[index] = None
 
     if capped:
         days = [day for index in sorted(capped) for day in capped[index]]

src/common/fetch_service.py

@@ -59,7 +59,10 @@ says how old with ``cache_max_age`` (``fetch_get(..., cache_max_age=ttl)``;
 Identical means what the validator store keys on: URL, query, effective
 headers and, for a session with cookies or auth, the session.
 
-**Counters.** Requests, merged requests, bytes, 304s, errors, HTTP errors,
+**Counters.** Requests, merged requests, bytes (``bytes`` decoded, as the
+caller reads them; ``wire_bytes`` as they crossed the network, which is
+what a metered connection pays for -- ESPN gzips, so the two differ ~14x),
+304s, errors, HTTP errors,
 adapter retries, throttled requests and seconds waited, plus requests
 answered without the network: ``memo_hits`` (the response cache) and
 ``cache_hits`` / ``legacy_cache_hits`` (a shared ESPN scoreboard cache entry,
@@ -201,6 +204,7 @@ _COUNTER_FIELDS = (
     "throttled",     # requests that waited for a host budget
     "overruns",      # requests that went after max_wait_seconds anyway
     "bytes",         # decoded response body bytes received
+    "wire_bytes",    # body bytes as they came off the socket (still compressed)
     "wait_seconds",  # time spent waiting for host budgets
     "memo_hits",     # answered from the response cache (max-age); nothing sent
     "cache_hits",    # scoreboard fetches answered from a shared ESPN cache entry
@@ -616,6 +620,30 @@ def _body_of(response: Any) -> Optional[bytes]:
     return content if isinstance(content, bytes) else None
 
 
+def _wire_bytes_of(response: Any, body: Optional[bytes]) -> int:
+    """How many body bytes came off the socket for ``response``: the
+    compressed size when the server sent gzip, which ESPN does for every
+    scoreboard (63 KB on the wire for an 865 KB college football Saturday).
+
+    urllib3's ``HTTPResponse.tell()`` counts the raw bytes read before
+    decoding. A response without one (a test double, an adapter that is not
+    urllib3) or one whose body was not read is counted at its decoded size,
+    or as 0, so the counter never claims less than it can prove.
+    """
+    if body is None:
+        return 0
+    raw = getattr(response, "raw", None)
+    tell = getattr(raw, "tell", None)
+    if callable(tell):
+        try:
+            read = tell()
+        except Exception:
+            read = None
+        if isinstance(read, int) and not isinstance(read, bool) and read > 0:
+            return read
+    return len(body)
+
+
 def _retries_of(response: Any) -> int:
     raw = getattr(response, "raw", None)
     retries = getattr(raw, "retries", None)
@@ -1117,6 +1145,7 @@ class FetchService:
                         http_errors=int(status is not None and status >= 400),
                         retries=_retries_of(response),
                         bytes=len(body) if body is not None else 0,
+                        wire_bytes=_wire_bytes_of(response, body),
                         throttled=int(waited > 0), overruns=int(overrun),
                         wait_seconds=waited)
         except Exception:

src/display_controller.py

@@ -368,6 +368,10 @@ class DisplayController:
         self.on_demand_last_error: Optional[str] = None
         self.on_demand_last_event: Optional[str] = None
         self.on_demand_schedule_override = False
+        # The mode the request named, when it named one (not a mode resolved
+        # from a bare plugin id). Shown even when the plugin's live checks
+        # would leave it out of the session (_on_demand_modes_for_plugin).
+        self._on_demand_named_mode: Optional[str] = None
         # Plugins that are disabled in config and loaded only because an
         # on-demand request named them. The main loop unloads each one once
         # on-demand has moved off it (_release_on_demand_plugins).
@@ -561,6 +565,17 @@ class DisplayController:
         except Exception:  # pylint: disable=broad-except
             logger.exception("Plugin system initialization failed")
             self.plugin_manager = None
+            # A restored session has no plugin to resume on. It may have been
+            # read already (on_demand_active) or not yet, if initialization
+            # failed before the restore ran; either way, end it visibly.
+            try:
+                cached_session = self.cache_manager.get('display_on_demand_config',
+                                                        max_age=3600)
+            except Exception:  # pylint: disable=broad-except
+                cached_session = None
+            if self.on_demand_active or cached_session:
+                self.cache_manager.clear_cache('display_on_demand_config')
+                self._set_on_demand_error('restore-failed')
             # Its state machine no longer describes what runs; let the last
             # snapshot go stale (readers then say unknown) rather than keep
             # refreshing it.
@@ -1215,6 +1230,13 @@ class DisplayController:
                 note = getattr(self.plugin_manager, 'note_display_duration', None)
                 if note is not None and plugin_id:
                     note(plugin_id, time.monotonic() - started)
+                # A screen that drew once and holds makes no more
+                # update_display() calls, so a frame the preview throttle
+                # skipped would otherwise never reach the snapshot.
+                write_owed = getattr(getattr(self, 'display_manager', None),
+                                     'write_owed_snapshot', None)
+                if write_owed is not None:
+                    write_owed()
 
     def _health_tracker(self):
         """The plugin circuit breaker, or None when it is not enabled."""
@@ -1665,6 +1687,7 @@ class DisplayController:
         self.on_demand_expires_at = None
         self.on_demand_pinned = False
         self.on_demand_schedule_override = False
+        self._on_demand_named_mode = None
         # While the session ran, _evaluate_schedule may have forced
         # is_display_active on over a scheduled-off answer. Drop the minute
         # gate so the next _check_schedule recomputes it; otherwise the panel
@@ -1831,6 +1854,7 @@ class DisplayController:
         self.on_demand_pinned = on_demand_config.get('pinned', False)
         self.on_demand_requested_at = on_demand_config.get('requested_at')
         self.on_demand_expires_at = on_demand_config.get('expires_at')
+        self._on_demand_named_mode = on_demand_config.get('named_mode')
         self.on_demand_status = 'active'
         self.on_demand_schedule_override = True
         logger.info("On-demand mode detected during initialization: resuming on plugin '%s'; "
@@ -2317,13 +2341,25 @@ class DisplayController:
                 return modes[0]
         return plugin_id
 
-    def _on_demand_modes_for_plugin(self, plugin_id: str) -> List[str]:
+    def _on_demand_modes_for_plugin(self, plugin_id: str,
+                                    named_mode: Optional[str] = None) -> List[str]:
         """Every loaded display mode belonging to `plugin_id`, in rotation order.
 
         Live modes that actually have content lead, then the rest, then live
         modes with nothing to show -- so an on-demand request for a sports
         plugin opens on a game in progress rather than an empty live screen.
         Returns an empty list when the plugin has no loaded modes.
+
+        `named_mode` is a mode the request asked for by name. It is always
+        in the list, first when the checks below would have dropped it.
+        Those checks ask has_live_content(), which is the live-priority
+        question -- "should this plugin take the panel from the rotation?"
+        -- and the sports plugins answer it for favourite teams only. Asking
+        for ncaa_fb_live with fifteen games on and no favourite playing got
+        a 200 and nfl_recent on the panel. The plugin's display() is what
+        knows whether the mode has anything to draw; when it has not, the
+        session moves to the plugin's next mode like any empty on-demand
+        mode.
         """
         plugin_modes = self.plugin_display_modes.get(plugin_id, [])
         if not plugin_modes:
@@ -2364,6 +2400,15 @@ class DisplayController:
             # Only live modes available but no content - use them anyway
             ordered_modes = live_modes
 
+        if named_mode and named_mode in available_plugin_modes:
+            # The named mode leads whether or not the live check kept it: a
+            # second live mode with content is already in the list, but behind
+            # the first, so the session would rotate away before reaching it.
+            if named_mode not in ordered_modes:
+                logger.info("On-demand: showing %s as requested; plugin '%s' reports no "
+                            "live-priority content for it", named_mode, plugin_id)
+            ordered_modes = [named_mode] + [m for m in ordered_modes if m != named_mode]
+
         return ordered_modes
 
     def _apply_on_demand_pin(self, ordered_modes: List[str], resolved_mode: Optional[str],
@@ -2392,10 +2437,20 @@ class DisplayController:
 
         plugin_id = self.on_demand_plugin_id
 
-        ordered_modes = self._on_demand_modes_for_plugin(plugin_id)
+        ordered_modes = self._on_demand_modes_for_plugin(plugin_id, self._on_demand_named_mode)
         if not ordered_modes:
-            logger.warning("No valid display modes found for on-demand plugin '%s' after restoration", plugin_id)
-            self.on_demand_modes = []
+            # The plugin did not load this time (seen on a rig: its config
+            # failed validation after the crash that caused the restart), so
+            # there is nothing to resume. Leaving the session active with no
+            # modes published it as active for a plugin that was not running
+            # until the first pass ended it as an ordinary 'idle', and kept
+            # the cached request for the next restart to trip over. End it
+            # as a failure the status endpoint reports, and drop the cache.
+            logger.error("On-demand session for plugin '%s' cannot resume after the "
+                         "restart: the plugin has no loaded display modes (did it "
+                         "fail to load?); ending it", plugin_id)
+            self.cache_manager.clear_cache('display_on_demand_config')
+            self._set_on_demand_error('restore-failed')
             return
 
         # A restart must not silently un-pin: the pin is part of the request
@@ -2596,7 +2651,10 @@ class DisplayController:
         if resolved_mode in self.available_modes:
             self.current_mode_index = self.available_modes.index(resolved_mode)
 
-        ordered_modes = self._on_demand_modes_for_plugin(resolved_plugin_id)
+        # Named: the request gave this mode itself, rather than a plugin id
+        # (or a mode the plugin doesn't have) that resolved to a default.
+        named_mode = resolved_mode if mode == resolved_mode else None
+        ordered_modes = self._on_demand_modes_for_plugin(resolved_plugin_id, named_mode)
         if not ordered_modes:
             logger.error("No valid display modes found for plugin '%s'", resolved_plugin_id)
             self._set_on_demand_error("no-modes")
@@ -2613,6 +2671,7 @@ class DisplayController:
         self.on_demand_requested_at = now
         self.on_demand_expires_at = (now + duration) if duration else None
         self.on_demand_pinned = pinned
+        self._on_demand_named_mode = named_mode
         self.on_demand_status = 'active'
         self.on_demand_last_error = None
         self.on_demand_last_event = 'started'
@@ -2641,6 +2700,7 @@ class DisplayController:
             'mode': resolved_mode,
             'duration': duration,
             'pinned': pinned,
+            'named_mode': named_mode,
             'requested_at': now,
             'expires_at': self.on_demand_expires_at
         }

src/display_manager.py

@@ -317,6 +317,11 @@ class DisplayManager:
         # is handed to the writer; this only once it has been saved, so an
         # mtime touch never vouches for a frame still waiting to be written.
         self._saved_snapshot_digest: Optional[int] = None
+        # A changed frame reached _write_snapshot_if_due() inside the write
+        # interval and was skipped. Nothing writes it unless update_display()
+        # runs again, and a screen that draws once and holds never calls it
+        # again -- see write_owed_snapshot().
+        self._snapshot_owed = False
         self._snapshot_dir_prepared = False
         # Background writer used mid-scroll; see _write_snapshot_if_due.
         self._snapshot_cond = threading.Condition()
@@ -1788,9 +1793,10 @@ class DisplayManager:
 
             if frame_checksum is not None:
                 digest = frame_checksum
+                frame_changed = digest != self._last_snapshot_digest
                 action = snapshot_policy.decide(
                     now, self._last_snapshot_ts, self._last_snapshot_touch_ts,
-                    viewer_fresh, digest != self._last_snapshot_digest)
+                    viewer_fresh, frame_changed)
             else:
                 # Ask as if the frame had changed before paying to find out.
                 # decide() is monotone in frame_changed -- a SKIP for a
@@ -1802,22 +1808,36 @@ class DisplayManager:
                     now, self._last_snapshot_ts, self._last_snapshot_touch_ts,
                     viewer_fresh, True)
                 if action is snapshot_policy.SnapshotAction.SKIP:
+                    # Not hashed, so not known to be unchanged: owed until a
+                    # later look finds it written or unchanged.
+                    self._snapshot_owed = True
                     return
                 digest = zlib.adler32(self.image.tobytes())
-                if digest == self._last_snapshot_digest:
+                frame_changed = digest != self._last_snapshot_digest
+                if not frame_changed:
                     # Unchanged after all: the decision an unchanged frame gets.
                     action = snapshot_policy.decide(
                         now, self._last_snapshot_ts,
                         self._last_snapshot_touch_ts, viewer_fresh, False)
             if action is snapshot_policy.SnapshotAction.SKIP:
+                # A changed frame inside the write interval stays owed: the
+                # next update_display() would write it, but a static screen
+                # may not make one -- write_owed_snapshot() covers that.
+                self._snapshot_owed = frame_changed
                 return
             if (action is snapshot_policy.SnapshotAction.TOUCH
                     and self._saved_snapshot_digest == digest):
                 # mtime bump only: keeps the health check (snapshot age)
                 # green without paying for a PNG encode of an unchanged frame
+                # (this frame is already on disk, so nothing is owed).
+                self._snapshot_owed = False
                 os.utime(self._snapshot_path, None)
                 self._last_snapshot_touch_ts = now
                 return
+            # Owed until the write below succeeds: if it raises, the frame
+            # stays owed and write_owed_snapshot() retries it, rather than a
+            # held screen leaving the preview stale after one failed write.
+            self._snapshot_owed = True
             # (A TOUCH for a frame that isn't on disk yet -- still queued, or
             # its write failed -- is written instead: touching would make the
             # older file on disk look current.)
@@ -1842,9 +1862,39 @@ class DisplayManager:
             self._last_snapshot_ts = now
             self._last_snapshot_touch_ts = now
             self._last_snapshot_digest = digest
+            self._snapshot_owed = False
         except Exception as e:
             self._log_snapshot_failure(e)
 
+    def write_owed_snapshot(self) -> None:
+        """Write a frame the snapshot throttle skipped, once it is due.
+
+        The preview snapshot is only ever written from update_display(), and
+        at most once per write interval (snapshot_policy). A frame pushed
+        inside that interval is skipped, and is written by the next
+        update_display() that comes after it -- but a screen that draws its
+        card once and then holds it makes no further call. Its frame was on
+        the panel and never in the preview: soccer's recent/upcoming cards
+        skip redundant redraws, and the first one after an on-demand start
+        (pushed a few milliseconds after the controller's clear) left
+        /api/v3/display/current and the web preview black for the whole
+        screen while the panel showed the card.
+
+        The render loop calls this after each frame. Cheap when nothing is
+        owed (one attribute read); otherwise the usual policy decides, so
+        the write still waits out the interval and an unchanged frame is
+        never re-encoded.
+        """
+        if not self._snapshot_owed:
+            return
+        try:
+            if self._writes_suppressed():
+                return
+            with self._update_lock:
+                self._write_snapshot_if_due()
+        except Exception as e:  # pylint: disable=broad-except
+            self._log_snapshot_failure(e)
+
     def _log_snapshot_failure(self, error: Exception) -> None:
         # Snapshot failures must never break display — but they must not
         # be silent either: the snapshot's mtime is the web UI's display

test/_run_loop_harness.py

@@ -724,11 +724,14 @@ class RunLoopHarness:
         self.clock.at(t, post)
 
     def restore_on_demand(self, plugin_id: str, mode: Optional[str] = None,
-                          duration: Optional[float] = None, pinned: bool = False):
+                          duration: Optional[float] = None, pinned: bool = False,
+                          named_mode: Optional[str] = None):
         """Start with an on-demand session resumed from the cache, as after
         a restart: the state _select_startup_plugins restores, then
-        _populate_on_demand_modes_from_plugin, as __init__ calls it."""
+        _populate_on_demand_modes_from_plugin, as __init__ calls it. A
+        session that cannot resume is logged as ``on-demand-error``."""
         dc = self.controller
+        dc._on_demand_named_mode = named_mode
         dc.on_demand_active = True
         dc.on_demand_plugin_id = plugin_id
         dc.on_demand_mode = mode
@@ -739,6 +742,8 @@ class RunLoopHarness:
         dc.on_demand_status = 'active'
         dc.on_demand_schedule_override = True
         dc._populate_on_demand_modes_from_plugin()
+        if dc.on_demand_status == 'error':
+            self.log("on-demand-error", dc.on_demand_last_error)
 
     def wifi_message(self, t: float, message: str, duration: float = 5):
         def write():

test/fixtures/run_loop_golden/on_demand_named_live.json

@@ -0,0 +1,29 @@
+{
+  "screens": [
+    [0.0, "clock", 5.0, "on-demand-start", 6, false],
+    [5.0, "sports_live", 15.0, "duration", 15, true],
+    [20.0, "sports_recent", 15.0, "duration", 15, true],
+    [35.0, "sports_upcoming", 5.0, "on-demand-requested-stop", 6, true],
+    [40.0, "clock", 20.0, "duration", 20, true],
+    [60.0, "sports_live", 15.0, "display-false", 11, true],
+    [75.0, "sports_recent", 15.0, "duration", 15, true],
+    [90.0, "sports_upcoming", 10.0, "on-demand-start", 11, true],
+    [100.0, "sports_live", 0.0, "empty", 1, true],
+    [100.0, "sports_recent", 15.0, "duration", 15, true],
+    [115.0, "sports_upcoming", 15.0, "duration", 15, true],
+    [130.0, "sports_live", 0.0, "empty", 1, true],
+    [130.0, "sports_recent", 10.0, "on-demand-requested-stop", 11, true],
+    [140.0, "sports_upcoming", 15.0, "duration", 15, true],
+    [155.0, "clock", 5.0, "horizon", 5, true]
+  ],
+  "events": [
+    [5.0, "request", "start:n1"],
+    [5.0, "on-demand-start", "sports"],
+    [40.0, "request", "stop:n2"],
+    [40.0, "on-demand-requested-stop"],
+    [100.0, "request", "start:n3"],
+    [100.0, "on-demand-start", "sports"],
+    [140.0, "request", "stop:n4"],
+    [140.0, "on-demand-requested-stop"]
+  ]
+}

test/fixtures/run_loop_golden/on_demand_restore_failed.json

@@ -0,0 +1,10 @@
+{
+  "screens": [
+    [0.0, "clock", 20.0, "duration", 20, false],
+    [20.0, "weather", 20.0, "duration", 20, true],
+    [40.0, "clock", 20.0, "horizon", 20, true]
+  ],
+  "events": [
+    [0.0, "on-demand-error", "restore-failed"]
+  ]
+}

test/test_api_v3_status_accuracy.py

@@ -0,0 +1,326 @@
+"""Four web answers that disagreed with the rig they describe (found on ledpi).
+
+1. POST /config/schedule refused the schedule GET returns on a fresh install
+   (config.template.json: per-day, every day off, schedule disabled) with
+   "At least one day must be enabled", as did /config/dim-schedule. A
+   disabled schedule needs no enabled day.
+2. A brightness-only POST /config/main answered ``restart_required: true``,
+   though the display applies brightness live (brightness.set over the
+   socket, and the config watcher). The flag now says whether anything
+   changed that the running display does not pick up by itself.
+3. /health stayed "healthy" with the display service stopped: only the
+   sub-checks changed. Service inactive, no socket and no live heartbeat
+   is now ``display_loop: stopped`` and "degraded".
+4. /display/current-status kept answering ``is_display_active: true`` from
+   the cache for up to 120 s after the display stopped. With no socket and
+   no live heartbeat it is now unknown.
+"""
+
+import copy
+import json
+import os
+import sys
+import time
+from pathlib import Path
+from unittest.mock import patch
+
+import pytest
+
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+from test._api_v3_test_helpers import api_v3_client, api_v3_module  # noqa: F401,E402
+
+from src import display_watchdog  # noqa: E402
+from src.ipc import client as control_client  # noqa: E402
+from web_interface import display_state  # noqa: E402
+
+REPO = Path(__file__).resolve().parent.parent
+TEMPLATE = json.loads((REPO / 'config' / 'config.template.json').read_text(encoding='utf-8'))
+
+
+@pytest.fixture
+def store(api_v3_module, monkeypatch):
+    state = {'config': {}, 'saves': 0}
+    api_v3_module.api_v3.config_manager.load_config.side_effect = \
+        lambda *a, **k: copy.deepcopy(state['config'])
+
+    def fake_save(_manager, config, **_kwargs):
+        state['config'] = copy.deepcopy(config)
+        state['saves'] += 1
+        return True, ''
+
+    monkeypatch.setattr(api_v3_module, '_save_config_atomic', fake_save)
+    return state
+
+
+# --- 1. schedules ---------------------------------------------------------------
+
+SCHEDULE_ROUTES = [('/api/v3/config/schedule', 'schedule'),
+                   ('/api/v3/config/dim-schedule', 'dim_schedule')]
+
+
+@pytest.mark.parametrize('route,section', SCHEDULE_ROUTES)
+def test_the_templates_disabled_per_day_schedule_saves_back(api_v3_client, store,
+                                                            route, section):
+    stored = copy.deepcopy(TEMPLATE[section])
+    stored['mode'] = 'per-day'
+    assert stored['enabled'] is False
+    assert not any(day['enabled'] for day in stored['days'].values())
+    store['config'] = {section: copy.deepcopy(stored)}
+
+    read = api_v3_client.get(route).get_json()['data']
+    resp = api_v3_client.post(route, json=read)
+
+    assert resp.status_code == 200, resp.get_json()
+    saved = store['config'][section]
+    assert saved['enabled'] is False and saved['mode'] == 'per-day'
+    # The disabled days keep their times: switching one on finds them.
+    assert saved['days'] == stored['days']
+
+
+@pytest.mark.parametrize('route,section', SCHEDULE_ROUTES)
+def test_an_enabled_per_day_schedule_still_needs_a_day(api_v3_client, store, route, section):
+    body = copy.deepcopy(TEMPLATE[section])
+    body.update(enabled=True, mode='per-day')
+    resp = api_v3_client.post(route, json=body)
+    assert resp.status_code == 400
+    assert 'At least one day must be enabled' in resp.get_json()['message']
+    assert store['saves'] == 0
+
+
+@pytest.mark.parametrize('route', [r for r, _ in SCHEDULE_ROUTES])
+def test_the_pickers_form_post_with_every_day_off_saves(api_v3_client, store, route):
+    """What schedule-picker.js posts: flat hidden inputs, booleans as strings,
+    times for every day."""
+    body = {'enabled': 'false', 'mode': 'per_day', 'start_time': '07:00', 'end_time': '23:00'}
+    for day in ('monday', 'tuesday', 'wednesday', 'thursday', 'friday', 'saturday', 'sunday'):
+        body.update({f'{day}_enabled': 'false', f'{day}_start': '06:30', f'{day}_end': '22:15'})
+    resp = api_v3_client.post(route, json=body)
+    assert resp.status_code == 200, resp.get_json()
+
+
+def test_an_invalid_time_on_a_disabled_day_is_dropped_not_refused(api_v3_client, store):
+    body = {'enabled': False, 'mode': 'per-day',
+            'days': {'monday': {'enabled': False, 'start_time': 'soon', 'end_time': '22:00'}}}
+    resp = api_v3_client.post('/api/v3/config/schedule', json=body)
+    assert resp.status_code == 200, resp.get_json()
+    assert store['config']['schedule']['days']['monday'] == {'enabled': False,
+                                                             'end_time': '22:00'}
+
+
+# --- 2. restart_required on /config/main ------------------------------------------
+
+STORED_MAIN = {
+    'timezone': 'America/Chicago',
+    'display': {
+        'hardware': {'rows': 32, 'cols': 64, 'chain_length': 2, 'brightness': 90,
+                     'disable_hardware_pulsing': False, 'inverse_colors': False,
+                     'show_refresh_rate': False},
+        'runtime': {'gpio_slowdown': 4},
+        'display_durations': {'clock': 15},
+        'use_short_date_format': False,
+    },
+}
+
+
+@pytest.fixture
+def main_store(store):
+    store['config'] = copy.deepcopy(STORED_MAIN)
+    return store
+
+
+def _save_main(client, body):
+    with patch('web_interface.blueprints.api_v3.control_client.brightness_set',
+               side_effect=control_client.ControlError('no_socket', 'x')):
+        resp = client.post('/api/v3/config/main', data=json.dumps(body),
+                           content_type='application/json')
+    assert resp.status_code == 200, resp.get_json()
+    return resp.get_json()
+
+
+def test_a_brightness_only_save_needs_no_restart(api_v3_client, main_store):
+    body = _save_main(api_v3_client, {'brightness': 40})
+    assert main_store['config']['display']['hardware']['brightness'] == 40
+    assert body['restart_required'] is False
+
+
+def test_a_brightness_save_on_a_config_without_a_display_section(api_v3_client, store):
+    """The route creates display.hardware and display.runtime on the way;
+    empty sections are not a change."""
+    store['config'] = {}
+    assert _save_main(api_v3_client, {'brightness': 40})['restart_required'] is False
+
+
+def test_the_display_form_with_only_brightness_changed_needs_no_restart(api_v3_client,
+                                                                       main_store):
+    hw = STORED_MAIN['display']['hardware']
+    body = {'__form_section': 'display', 'rows': 32, 'cols': 64, 'chain_length': 2,
+            'brightness': 55, 'gpio_slowdown': 4}
+    body.update({k: 'on' for k in ('disable_hardware_pulsing', 'inverse_colors',
+                                   'show_refresh_rate') if hw[k]})
+    assert _save_main(api_v3_client, body)['restart_required'] is False
+
+
+def test_a_mode_duration_needs_no_restart(api_v3_client, main_store):
+    body = _save_main(api_v3_client, {'duration__clock': 40})
+    assert main_store['config']['display']['display_durations']['clock'] == 40
+    assert body['restart_required'] is False
+
+
+@pytest.mark.parametrize('change', [{'rows': 64}, {'brightness': 40, 'chain_length': 3},
+                                    {'gpio_slowdown': 2}, {'timezone': 'UTC'}])
+def test_a_setting_the_display_reads_at_startup_still_needs_one(api_v3_client, main_store,
+                                                                change):
+    assert _save_main(api_v3_client, change)['restart_required'] is True
+
+
+def test_restart_needed_compares_leaves():
+    from web_interface.blueprints.api_v3.config import restart_needed
+    before = {'display': {'hardware': {'brightness': 90, 'rows': 32}}}
+    assert not restart_needed(before, copy.deepcopy(before))
+    assert not restart_needed(before, {'display': {'hardware': {'brightness': 10, 'rows': 32},
+                                                   'runtime': {}}})
+    assert restart_needed(before, {'display': {'hardware': {'brightness': 90}}})  # removed
+    assert not restart_needed({}, {'clock': {'enabled': True}}, live_paths=[('clock',)])
+    assert restart_needed({}, {'clockwork': {'enabled': True}}, live_paths=[('clock',)])
+
+
+# --- 3 and 4. a stopped display -----------------------------------------------------
+
+@pytest.fixture
+def no_display(monkeypatch, tmp_path):
+    """A Pi whose display service has stopped: the socket is expected here
+    but does not answer, and systemd took the heartbeat's directory away."""
+    monkeypatch.setattr(display_state, 'socket_supported', lambda: True)
+    monkeypatch.setattr(display_state, 'client_socket_paths', lambda: [str(tmp_path / 'gone')])
+    monkeypatch.setattr(display_state, 'read_state', lambda: None)
+    path = tmp_path / 'display-heartbeat.json'
+    monkeypatch.setattr(display_watchdog, 'HEARTBEAT_PATH', str(path))
+
+    def beat(age, pid=None):
+        path.write_text(json.dumps({'pid': os.getpid() if pid is None else pid,
+                                    'mono': time.monotonic() - age,
+                                    'wall': time.time() - age}))
+    return beat
+
+
+@pytest.fixture
+def service(monkeypatch):
+    status = {'active': False, 'returncode': 3, 'stdout': 'inactive', 'stderr': ''}
+    monkeypatch.setattr('web_interface.blueprints.api_v3.misc._get_display_service_status',
+                        lambda: dict(status))
+    return status
+
+
+@pytest.fixture
+def fresh_preview(tmp_path, monkeypatch):
+    """The preview frame the display left behind, under 60 s old: on its own
+    it kept the hardware check "connected"."""
+    from web_interface import display_preview
+    snapshot = tmp_path / 'preview.png'
+    snapshot.write_bytes(b'png')
+    monkeypatch.setattr(display_preview, 'SNAPSHOT_PATH', str(snapshot))
+
+
+def _health(client):
+    resp = client.get('/api/v3/health')
+    assert resp.status_code == 200, resp.get_json()
+    return resp.get_json()['data']
+
+
+class TestHealth:
+    def test_a_stopped_display_service_is_degraded(self, api_v3_client, no_display, service,
+                                                   fresh_preview):
+        data = _health(api_v3_client)
+        assert data['services']['display_service']['status'] == 'inactive'
+        assert data['checks']['display_loop']['status'] == 'stopped'
+        assert data['status'] == 'degraded'
+
+    def test_a_service_still_starting_is_not(self, api_v3_client, no_display, service,
+                                             fresh_preview):
+        """Active, before its socket and first heartbeat: not stopped."""
+        service.update(active=True, stdout='active', returncode=0)
+        data = _health(api_v3_client)
+        assert data['checks']['display_loop']['status'] == 'not_reported'
+        assert data['status'] == 'healthy'
+
+    def test_a_display_run_by_hand_is_not_stopped(self, api_v3_client, no_display, service,
+                                                  fresh_preview):
+        """The service is off but a display process beats (sudo python3 run.py)."""
+        no_display(age=2)
+        data = _health(api_v3_client)
+        assert data['checks']['display_loop']['status'] == 'running'
+        assert data['status'] == 'healthy'
+
+    @pytest.mark.parametrize('platform', ['no_unix_sockets', 'socket_off'])
+    def test_without_a_socket_to_expect_nothing_changes(self, api_v3_client, no_display,
+                                                        service, fresh_preview, monkeypatch,
+                                                        platform):
+        """Windows and the dev server (no systemd unit), or the socket
+        deliberately off: no heartbeat is no signal, as before."""
+        if platform == 'no_unix_sockets':
+            monkeypatch.setattr(display_state, 'socket_supported', lambda: False)
+        else:
+            monkeypatch.setattr(display_state, 'client_socket_paths', lambda: [])
+        service.update(returncode=-1, stdout='', stderr='systemctl not found')
+        data = _health(api_v3_client)
+        assert data['checks']['display_loop']['status'] == 'not_reported'
+        assert data['status'] == 'healthy'
+
+    def test_the_status_only_answer_says_degraded(self, api_v3_client, no_display, service,
+                                                  fresh_preview, monkeypatch):
+        monkeypatch.setattr('web_interface.blueprints.api_v3.misc.request_is_authenticated',
+                            lambda: False)
+        resp = api_v3_client.get('/api/v3/health')
+        assert resp.get_json()['data'] == {'status': 'degraded'}
+
+
+class TestCurrentStatus:
+    CACHED = {'mode': 'clock', 'plugin_id': 'clock', 'is_display_active': True,
+              'on_demand_active': False, 'last_updated': None}
+
+    @pytest.fixture
+    def cached(self, api_v3_module):
+        entry = dict(self.CACHED, last_updated=time.time() - 30)
+        cache = api_v3_module.api_v3.cache_manager
+        cache.get.side_effect = lambda key, *a, **kw: (
+            dict(entry) if key == 'display_current_state' else None)
+        return entry
+
+    def _status(self, client):
+        resp = client.get('/api/v3/display/current-status')
+        assert resp.status_code == 200
+        return resp.get_json()['data']
+
+    def test_a_stopped_display_is_not_reported_active(self, api_v3_client, no_display, cached):
+        data = self._status(api_v3_client)
+        assert not data.get('is_display_active')
+        assert data['mode'] is None and data['last_updated'] is None
+        assert data['source'] == 'cache'
+
+    def test_a_stale_heartbeat_is_not_active_either(self, api_v3_client, no_display, cached):
+        no_display(age=display_watchdog.HEARTBEAT_STALE_SECONDS + 5)
+        assert self._status(api_v3_client)['mode'] is None
+
+    @pytest.mark.skipif(os.name != 'posix', reason='process_exists answers only on POSIX')
+    def test_a_heartbeat_from_a_dead_process_is_not_active(self, api_v3_client, no_display,
+                                                           cached):
+        no_display(age=1, pid=2 ** 22 + 12345)
+        assert self._status(api_v3_client)['mode'] is None
+
+    def test_a_live_heartbeat_without_a_socket_reads_the_cache(self, api_v3_client,
+                                                               no_display, cached):
+        """An older display with no socket, still running."""
+        no_display(age=2)
+        data = self._status(api_v3_client)
+        assert data['mode'] == 'clock' and data['is_display_active'] is True
+
+    @pytest.mark.parametrize('platform', ['no_unix_sockets', 'socket_off'])
+    def test_without_a_socket_to_expect_the_cache_answers(self, api_v3_client, no_display,
+                                                          cached, monkeypatch, platform):
+        if platform == 'no_unix_sockets':
+            monkeypatch.setattr(display_state, 'socket_supported', lambda: False)
+        else:
+            monkeypatch.setattr(display_state, 'client_socket_paths', lambda: [])
+        data = self._status(api_v3_client)
+        assert data['mode'] == 'clock' and data['is_display_active'] is True

test/test_espn_dates.py

@@ -489,3 +489,171 @@ class TestConcurrency:
 
         assert live["peak"] <= espn_dates.ESPN_CHUNK_WORKERS
         assert live["peak"] > 1, "chunks should actually overlap"
+
+
+class TestEdgeMonths:
+    """A window's partial edge months are asked whole and trimmed.
+
+    The default scoreboard window -- a fortnight either side of today -- spans
+    two partial months, so it used to cost 29 day requests per league. ESPN's
+    ``dates=YYYYMMDD`` means a US Eastern day (verified against the live API
+    on 2026-10-03, 417 of 417 soccer events), so a month answer trimmed to
+    the window's Eastern days is what the day requests returned.
+    """
+
+    def test_a_fortnight_either_side_is_two_requests(self):
+        planned = espn_dates.espn_request_chunks(date(2026, 9, 20), date(2026, 10, 18))
+        assert planned == [
+            ("202609", (date(2026, 9, 20), date(2026, 9, 30))),
+            ("202610", (date(2026, 10, 1), date(2026, 10, 18))),
+        ]
+
+    def test_a_live_polls_two_days_stay_two_days(self):
+        planned = espn_dates.espn_request_chunks(date(2026, 10, 2), date(2026, 10, 3))
+        assert planned == [("20261002", None), ("20261003", None)]
+
+    def test_the_threshold_is_inclusive(self):
+        n = espn_dates.ESPN_MONTH_COVER_MIN_DAYS
+        short = espn_dates.espn_request_chunks(date(2026, 10, 1), date(2026, 10, n - 1))
+        assert [chunk for chunk, _ in short] == [
+            "202610%02d" % day for day in range(1, n)]
+        enough = espn_dates.espn_request_chunks(date(2026, 10, 1), date(2026, 10, n))
+        assert enough == [("202610", (date(2026, 10, 1), date(2026, 10, n)))]
+
+    def test_whole_months_and_short_edges_are_unchanged(self):
+        planned = espn_dates.espn_request_chunks(date(2026, 8, 30), date(2026, 10, 2))
+        assert planned == [
+            ("20260830", None), ("20260831", None), ("202609", None),
+            ("20261001", None), ("20261002", None),
+        ]
+
+    def test_without_time_zone_data_edges_stay_days(self, monkeypatch):
+        monkeypatch.setattr(espn_dates, "_EASTERN", None)
+        planned = espn_dates.espn_request_chunks(date(2026, 9, 20), date(2026, 10, 18))
+        assert len(planned) == 29
+        assert all(trim is None for _, trim in planned)
+
+    @pytest.mark.parametrize("start,end", [
+        (date(2026, 9, 20), date(2026, 10, 18)),
+        (date(2026, 1, 25), date(2026, 3, 3)),
+        (date(2026, 12, 20), date(2027, 1, 9)),
+        (date(2026, 10, 5), date(2026, 10, 9)),
+    ])
+    def test_the_planned_requests_still_cover_every_day_exactly_once(self, start, end):
+        covered = []
+        for chunk, trim in espn_dates.espn_request_chunks(start, end):
+            if trim is None:
+                covered.extend(days_covered_by([chunk]))
+            else:
+                assert chunk == trim[0].strftime("%Y%m") == trim[1].strftime("%Y%m")
+                covered.extend(trim[0] + timedelta(days=offset)
+                               for offset in range((trim[1] - trim[0]).days + 1))
+        expected = [start + timedelta(days=offset) for offset in range((end - start).days + 1)]
+        assert covered == expected
+
+    def test_a_trimmed_month_keeps_only_the_windows_eastern_days(self):
+        september = [
+            # 03:30Z on the 20th is still the 19th in New York: outside.
+            {"id": "before", "date": "2026-09-20T03:30Z"},
+            {"id": "first", "date": "2026-09-20T14:00Z"},
+            {"id": "late", "date": "2026-09-30T23:30Z"},
+        ]
+        october = [
+            {"id": "oct1", "date": "2026-10-01T19:00Z"},
+            # 03:30Z on the 19th is the evening of the 18th in New York: inside.
+            {"id": "last", "date": "2026-10-19T03:30Z"},
+            {"id": "after", "date": "2026-10-19T14:00Z"},
+            {"id": "undated"},
+        ]
+        session = FakeSession({"202609": september, "202610": october})
+
+        data = fetch_espn_date_chunks(session, URL, params={"dates": "20260920-20261018"})
+
+        assert sorted(call["dates"] for call in session.calls) == ["202609", "202610"]
+        # An event with no readable date is kept, never dropped on a guess.
+        assert [e["id"] for e in data["events"]] == ["first", "late", "oct1", "last", "undated"]
+
+    def test_eastern_standard_time_is_honoured_after_the_clocks_change(self):
+        # 2026-11-01 ends daylight saving: Eastern is UTC-5 from then on.
+        november = [
+            {"id": "out", "date": "2026-11-15T04:30Z"},  # Nov 14, 23:30 EST
+            {"id": "in", "date": "2026-11-15T05:30Z"},   # Nov 15, 00:30 EST
+        ]
+        session = FakeSession({"202611": november})
+        data = fetch_espn_date_chunks(session, URL, params={"dates": "20261115-20261121"})
+        assert [e["id"] for e in data["events"]] == ["in"]
+
+    def test_a_capped_edge_month_re_asks_only_the_windows_days(self):
+        full = [{"id": "cap%d" % i, "date": "2026-10-05T18:00Z"} for i in range(ESPN_MAX_LIMIT)]
+        by_chunk = {"202610": full}
+        by_chunk.update({"202610%02d" % day: [{"id": "o%02d" % day}] for day in range(1, 32)})
+        session = FakeSession(by_chunk)
+
+        data = fetch_espn_date_chunks(session, URL, params={"dates": "20261001-20261010"})
+
+        sent = [call["dates"] for call in session.calls]
+        assert sent[0] == "202610"
+        assert sorted(sent[1:]) == ["202610%02d" % day for day in range(1, 11)]
+        assert [e["id"] for e in data["events"]] == ["o%02d" % day for day in range(1, 11)]
+
+
+class TestProcessWideChunkCap:
+    """The chunk cap holds across windows, not per window.
+
+    A soccer board starting eight leagues fetches sixteen windows at once.
+    With a pool of ``ESPN_CHUNK_WORKERS`` each, ~40 requests were in flight
+    and every one past a session's pool opened a connection -- and a DNS
+    lookup. On ledpi that was ~90 NameResolutionErrors per start.
+    """
+
+    def test_concurrent_windows_share_one_budget(self):
+        live = {"now": 0, "peak": 0}
+        guard = threading.Lock()
+
+        class CountingSession(FakeSession):
+            def get(self, url, params=None, headers=None, timeout=None):
+                with guard:
+                    live["now"] += 1
+                    live["peak"] = max(live["peak"], live["now"])
+                try:
+                    time.sleep(0.01)
+                    return super().get(url, params=params, headers=headers, timeout=timeout)
+                finally:
+                    with guard:
+                        live["now"] -= 1
+
+        sessions = [CountingSession() for _ in range(6)]
+        # Six leagues, so the fetch service cannot merge them into one, on a
+        # host with no token bucket: earlier tests may have spent ESPN's
+        # burst, and a bucket paced at 20/s would serialise these by itself.
+        threads = [
+            threading.Thread(target=fetch_espn_date_chunks,
+                             args=(session, "https://scores.example.test/league%d" % index),
+                             kwargs={"params": {"dates": "20260101-20261231"}})
+            for index, session in enumerate(sessions)
+        ]
+        for thread in threads:
+            thread.start()
+        for thread in threads:
+            thread.join(timeout=30)
+
+        assert all(len(session.calls) == 12 for session in sessions)
+        assert live["peak"] <= espn_dates.ESPN_CHUNK_WORKERS
+        assert live["peak"] > 1, "chunks should still overlap"
+
+
+def test_a_fresh_process_skips_the_doomed_range_request():
+    """Every start used to spend one 400 per window learning that ranges are
+    still rejected -- eleven at once from a soccer board. A new process now
+    starts inside the retry period instead."""
+    import subprocess
+    import sys
+    from pathlib import Path
+
+    out = subprocess.run(
+        [sys.executable, "-c",
+         "import src.common.espn_dates as e; print(e._ranges_known_rejected())"],
+        cwd=str(Path(__file__).resolve().parents[1]),
+        capture_output=True, text=True, timeout=60,
+    )
+    assert out.stdout.strip() == "True", out.stderr

test/test_fetch_service.py

@@ -576,6 +576,42 @@ class TestCounters:
         assert snap["hosts"]["site.api.espn.com"]["requests"] == 1
         assert snap["totals"]["bytes"] == 3 * len(b'{"ok": 1}')
 
+    def test_wire_bytes_are_the_compressed_size(self, service):
+        # Built the way requests builds a real response: a urllib3
+        # HTTPResponse carrying a gzip body, decoded when .content is read.
+        import gzip
+        import io
+
+        from requests.adapters import HTTPAdapter
+        from urllib3.response import HTTPResponse
+
+        decoded = json.dumps({"events": [{"id": str(i), "name": "x" * 200}
+                                         for i in range(50)]}).encode()
+        wire = gzip.compress(decoded)
+
+        def handler(url, kwargs):
+            raw = HTTPResponse(body=io.BytesIO(wire), status=200,
+                               headers={"Content-Encoding": "gzip",
+                                        "Content-Type": "application/json"},
+                               preload_content=False, decode_content=True)
+            request = requests.Request("GET", url).prepare()
+            response = HTTPAdapter().build_response(request, raw)
+            response.content  # what Session.get does for a non-streamed call
+            return response
+
+        response = service.get(FakeSession(handler), "https://site.api.espn.com/x")
+        assert response.content == decoded
+        totals = _counters(service)
+        assert totals["bytes"] == len(decoded)
+        assert totals["wire_bytes"] == len(wire) < len(decoded)
+
+    def test_wire_bytes_fall_back_to_the_decoded_size(self, service):
+        # No urllib3 response behind it (a test double, another adapter):
+        # count what is known rather than nothing.
+        service.get(FakeSession(), "https://api.test/x")
+        totals = _counters(service)
+        assert totals["wire_bytes"] == totals["bytes"] == len(b'{"ok": 1}')
+
     def test_errors_and_http_errors(self, service):
         def handler(url, kwargs):
             if url.endswith("/down"):
@@ -670,14 +706,14 @@ class TestCallerIdentity:
         assert _counters(global_service, plugin="football-scoreboard")["requests"] == 1
 
     def test_espn_chunks_on_worker_threads_count_against_the_caller(self, global_service):
-        from src.common.espn_dates import espn_date_chunks, fetch_espn_date_chunks, parse_espn_date_range
+        from src.common.espn_dates import espn_request_chunks, fetch_espn_date_chunks, parse_espn_date_range
 
         session = FakeSession(lambda url, kw: make_response(body=b'{"events": []}', url=url))
         dates = "20260801-20261015"
         with plugin_scope("baseball-scoreboard"):
             fetch_espn_date_chunks(session, "https://site.api.espn.com/s/scoreboard",
                                    params={"dates": dates})
-        chunks = len(espn_date_chunks(*parse_espn_date_range(dates)))
+        chunks = len(espn_request_chunks(*parse_espn_date_range(dates)))
         assert chunks > 1
         assert len(session.calls) == chunks
         assert _counters(global_service, plugin="baseball-scoreboard")["requests"] == chunks

test/test_on_demand_live_and_restore.py

@@ -0,0 +1,162 @@
+"""Two on-demand edges seen on a rig.
+
+  * A request naming a ``*_live`` mode got HTTP 200 and a different mode on
+    the panel. The session's mode list kept live modes only when the plugin's
+    has_live_content() said so, and that is the live-priority question,
+    which the sports plugins answer for favourite teams only: fifteen college
+    games on, no favourite playing, and ``ncaa_fb_live`` became
+    ``nfl_recent``.
+  * A restart during a session whose plugin then failed to load (its config
+    no longer validated) logged "No valid display modes found ... after
+    restoration" and left the session active with no modes: published as
+    active for a plugin that was not running, with its cached request kept
+    for the next restart.
+"""
+
+from unittest.mock import MagicMock
+
+import pytest
+
+SPORTS_MODES = ['nfl_live', 'nfl_recent', 'nfl_upcoming',
+                'ncaa_fb_live', 'ncaa_fb_recent', 'ncaa_fb_upcoming']
+
+
+def _sports_plugin(has_live_content=False):
+    plugin = MagicMock(spec=['display', 'has_live_content', 'has_live_priority',
+                             'get_live_modes'])
+    plugin.has_live_content.return_value = has_live_content
+    plugin.has_live_priority.return_value = True
+    plugin.get_live_modes.return_value = []
+    return plugin
+
+
+def _register(controller, plugin_id, modes, plugin):
+    controller.plugin_display_modes[plugin_id] = list(modes)
+    for mode in modes:
+        controller.plugin_modes[mode] = plugin
+        controller.mode_to_plugin_id[mode] = plugin_id
+        if mode not in controller.available_modes:
+            controller.available_modes.append(mode)
+
+
+@pytest.fixture
+def football(test_display_controller):
+    c = test_display_controller
+    _register(c, 'football-scoreboard', SPORTS_MODES, _sports_plugin())
+    return c
+
+
+class TestANamedLiveModeIsShown:
+
+    def test_it_is_the_first_screen(self, football):
+        football._activate_on_demand({'plugin_id': 'football-scoreboard',
+                                      'mode': 'ncaa_fb_live'})
+        assert football.on_demand_active
+        assert football.current_display_mode == 'ncaa_fb_live'
+        assert football.on_demand_mode == 'ncaa_fb_live'
+
+    def test_the_plugins_other_modes_follow_it(self, football):
+        football._activate_on_demand({'plugin_id': 'football-scoreboard',
+                                      'mode': 'ncaa_fb_live'})
+        assert football.on_demand_modes[0] == 'ncaa_fb_live'
+        assert set(football.on_demand_modes[1:]) == {
+            'nfl_recent', 'nfl_upcoming', 'ncaa_fb_recent', 'ncaa_fb_upcoming'}
+
+    def test_pinned_holds_it(self, football):
+        football._activate_on_demand({'plugin_id': 'football-scoreboard',
+                                      'mode': 'ncaa_fb_live', 'pinned': True})
+        assert football.on_demand_modes == ['ncaa_fb_live']
+
+    def test_a_bare_plugin_request_still_skips_quiet_live_modes(self, football):
+        """Only a mode asked for by name is kept: a plugin-only request
+        resolves to the plugin's first mode (nfl_live), and opening on an
+        empty live screen there is what the ordering exists to avoid."""
+        football._activate_on_demand({'plugin_id': 'football-scoreboard'})
+        assert not any(m.endswith('_live') for m in football.on_demand_modes)
+
+    def test_a_named_second_live_mode_with_content_leads(self, test_display_controller):
+        """With live content both live modes are kept, nfl_live first; a
+        request naming ncaa_fb_live must still open on it, not rotate away."""
+        c = test_display_controller
+        _register(c, 'football-scoreboard', SPORTS_MODES, _sports_plugin(has_live_content=True))
+        c._activate_on_demand({'plugin_id': 'football-scoreboard', 'mode': 'ncaa_fb_live'})
+        assert c.on_demand_modes[0] == 'ncaa_fb_live'
+        assert c.on_demand_modes.count('ncaa_fb_live') == 1
+        assert 'nfl_live' in c.on_demand_modes[1:]
+
+    def test_the_named_mode_survives_a_restart(self, football):
+        football._activate_on_demand({'plugin_id': 'football-scoreboard',
+                                      'mode': 'ncaa_fb_live'})
+        saved = football.cache_manager.set.call_args_list[-1]
+        assert saved.args[0] == 'display_on_demand_config'
+        config = saved.args[1]
+        assert config['named_mode'] == 'ncaa_fb_live'
+
+        football._reset_on_demand_fields()
+        football._select_startup_plugins(['football-scoreboard'], config)
+        football._populate_on_demand_modes_from_plugin()
+        assert football.on_demand_modes[football.on_demand_mode_index] == 'ncaa_fb_live'
+
+
+class TestARestoreWithNothingToResume:
+
+    @pytest.fixture
+    def restored(self, test_display_controller):
+        c = test_display_controller
+        c.config['clock-simple'] = {'enabled': True}
+        c._select_startup_plugins(['clock-simple'],
+                                  {'plugin_id': 'clock-simple', 'mode': 'clock-simple'})
+        assert c.on_demand_active
+        # The plugin's load then fails: nothing is registered for it.
+        c.cache_manager.clear_cache.reset_mock()
+        c._populate_on_demand_modes_from_plugin()
+        return c
+
+    def test_the_session_ends(self, restored):
+        assert not restored.on_demand_active
+        assert restored.on_demand_plugin_id is None
+        assert not restored.on_demand_schedule_override
+
+    def test_it_is_reported_as_an_error(self, restored):
+        assert restored.on_demand_status == 'error'
+        assert restored.on_demand_last_error == 'restore-failed'
+        # The last on-demand state write, not the last write of any key: the
+        # font-usage publisher thread writes its own key at its own pace.
+        published = [c for c in restored.cache_manager.set.call_args_list
+                     if c.args and c.args[0] == 'display_on_demand_state'][-1]
+        assert published.args[1]['status'] == 'error'
+        assert published.args[1]['error'] == 'restore-failed'
+
+    def test_the_cached_request_is_dropped(self, restored):
+        restored.cache_manager.clear_cache.assert_any_call('display_on_demand_config')
+
+
+def test_a_plugin_system_failure_ends_a_cached_session_not_yet_restored(
+        mock_config_manager, mock_display_manager, mock_cache_manager,
+        test_config_with_plugins, emulator_mode):
+    """Initialization can fail before the cached session is read, with
+    on_demand_active still False: the session must still end, visibly."""
+    from unittest.mock import patch
+    from src.display_controller import DisplayController
+
+    mock_config_manager.get_config.return_value = test_config_with_plugins
+    mock_config_manager.load_config.return_value = test_config_with_plugins
+    mock_cache_manager._memory_cache['display_on_demand_config'] = {
+        'plugin_id': 'clock-simple', 'mode': 'clock-simple'}
+    with patch('src.display_controller.ConfigManager', return_value=mock_config_manager), \
+         patch('src.display_controller.DisplayManager', return_value=mock_display_manager), \
+         patch('src.display_controller.CacheManager', return_value=mock_cache_manager), \
+         patch('src.display_controller.FontManager'), \
+         patch('src.plugin_system.PluginManager', side_effect=RuntimeError("boom")):
+        controller = DisplayController()
+        try:
+            assert controller.plugin_manager is None
+            assert not controller.on_demand_active
+            assert controller.on_demand_status == 'error'
+            assert controller.on_demand_last_error == 'restore-failed'
+            mock_cache_manager.clear_cache.assert_any_call('display_on_demand_config')
+        finally:
+            try:
+                controller.cleanup()
+            except Exception:
+                pass

test/test_run_loop_golden.py

@@ -133,6 +133,32 @@ def scenario_on_demand_restored(h: RunLoopHarness):
     h.restore_on_demand("sports", mode="sports_upcoming", duration=40)
 
 
+def scenario_on_demand_named_live(h: RunLoopHarness):
+    # Games are on until t=70, but none involves a favourite, so
+    # has_live_content() (the live-priority answer) stays False throughout.
+    # A request naming sports_live still opens on it (it opened on
+    # sports_recent); asked for again after the games end, it has nothing to
+    # draw and the session moves on to the plugin's next mode.
+    h.add_plugin(FakePlugin("clock", ["clock"], duration=20))
+    h.add_plugin(FakePlugin(
+        "sports", ["sports_live", "sports_recent", "sports_upcoming"], duration=15,
+        live_priority=True,
+        content=lambda t, mode: mode != "sports_live" or t < 70))
+    h.on_demand_request(5, "n1", plugin_id="sports", mode="sports_live")
+    h.on_demand_request(40, "n2", action="stop")
+    h.on_demand_request(100, "n3", plugin_id="sports", mode="sports_live")
+    h.on_demand_request(140, "n4", action="stop")
+
+
+def scenario_on_demand_restore_failed(h: RunLoopHarness):
+    # A restart during a session whose plugin then fails to load: the
+    # session ends as an error before the first screen, and the rotation
+    # runs normally from the top.
+    h.add_plugin(FakePlugin("clock", ["clock"], duration=20))
+    h.add_plugin(FakePlugin("weather", ["weather"], duration=20))
+    h.restore_on_demand("gone", mode="gone", duration=40)
+
+
 def scenario_schedule(h: RunLoopHarness):
     # The clock starts at 22:59:30. Off from 23:01 until 23:05 (the window
     # spans midnight); dimmed from 23:00 until 23:01.
@@ -194,6 +220,8 @@ SCENARIOS = {
     "on_demand": (scenario_on_demand, 240),
     "on_demand_pinned": (scenario_on_demand_pinned, 160),
     "on_demand_restored": (scenario_on_demand_restored, 100),
+    "on_demand_named_live": (scenario_on_demand_named_live, 160),
+    "on_demand_restore_failed": (scenario_on_demand_restore_failed, 60),
     "schedule": (scenario_schedule, 400),
     "wifi_notice": (scenario_wifi_notice, 150),
     "follower": (scenario_follower, 80),

test/test_snapshot_owed_frame.py

@@ -0,0 +1,183 @@
+"""A frame the preview throttle skipped still reaches the snapshot.
+
+The preview snapshot (/api/v3/display/current, the web UI's live preview) is
+only written from update_display(), at most once per write interval. A screen
+that draws its card once and then holds it -- soccer's recent/upcoming cards
+skip redundant redraws -- pushes exactly one frame. When that push lands inside
+the interval, e.g. a few milliseconds after the on-demand start's clear wrote a
+black frame, the throttle skips it and nothing ever writes it: on ledpi the
+preview stayed black for soccer's whole 15 s screen while the panel showed the
+card, and the next screen "rendered immediately".
+
+Runs the real DisplayManager on the emulator, like test_display_dirty_tracking.
+"""
+
+import os
+import sys
+import types
+
+os.environ["EMULATOR"] = "true"
+
+import pytest
+from PIL import Image
+
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), ".."))
+
+
+@pytest.fixture(scope="module")
+def dm(tmp_path_factory):
+    from src.display_manager import DisplayManager
+    DisplayManager._instance = None
+    manager = DisplayManager({
+        "display": {
+            "hardware": {"rows": 32, "cols": 64, "chain_length": 2,
+                         "parallel": 1, "brightness": 90},
+            "runtime": {"gpio_slowdown": 0},
+        },
+    }, suppress_test_pattern=True)
+    manager._snapshot_path = str(
+        tmp_path_factory.mktemp("owed_snapshot") / "led_matrix_preview.png")
+    yield manager
+    DisplayManager._instance = None
+
+
+@pytest.fixture
+def viewer(dm, monkeypatch, tmp_path):
+    """A preview is open (1 s write interval); fresh snapshot bookkeeping."""
+    monkeypatch.setattr(dm, "_viewer_is_fresh", lambda now: True)
+    dm._viewer_was_fresh = True
+    dm._snapshot_path = str(tmp_path / "snap.png")
+    dm._last_snapshot_ts = 0.0
+    dm._last_snapshot_touch_ts = 0.0
+    dm._last_snapshot_digest = None
+    dm._saved_snapshot_digest = None
+    dm._snapshot_owed = False
+    dm.set_scrolling_state(False)
+    return dm
+
+
+def _lit(path):
+    with Image.open(path) as img:
+        return sum(1 for p in img.convert("RGB").getdata() if max(p) > 20)
+
+
+def _age_last_write(dm, seconds=2.0):
+    """As if `seconds` had passed since the last snapshot write."""
+    dm._last_snapshot_ts -= seconds
+    dm._last_snapshot_touch_ts -= seconds
+
+
+def _clear_then_draw_card(dm):
+    """The on-demand start's clear, then the card a few ms later."""
+    dm.clear()
+    dm.update_display()                       # black frame: written
+    assert _lit(dm._snapshot_path) == 0
+    dm.draw.rectangle([4, 4, 40, 20], fill=(255, 255, 0))
+    dm.update_display()                       # the card: inside the interval
+
+
+def _controller(dm):
+    from src import display_controller as dc_module
+    controller = dc_module.DisplayController.__new__(dc_module.DisplayController)
+    controller.plugin_manager = None
+    controller.display_manager = dm
+    return controller
+
+
+class _HoldingPlugin:
+    """Already showing its card: display() returns True and draws nothing."""
+
+    plugin_id = "holding"
+
+    def __init__(self):
+        self.calls = 0
+
+    def display(self, display_mode=None, force_clear=False):
+        self.calls += 1
+        return True
+
+
+def test_a_held_card_reaches_the_preview_on_the_next_frame(viewer):
+    dm = viewer
+    _clear_then_draw_card(dm)
+    assert _lit(dm._snapshot_path) == 0       # the throttle skipped the card
+
+    controller = _controller(dm)
+    plugin = _HoldingPlugin()
+    _age_last_write(dm)
+    # The render loop's next frame: the plugin draws nothing and makes no
+    # update_display() call, as soccer's switch cards do.
+    assert controller._display_once(plugin, "soccer_eng.1_recent", True) is True
+    assert plugin.calls == 1
+    assert _lit(dm._snapshot_path) > 0
+
+
+def test_the_owed_write_still_waits_out_the_interval(viewer, monkeypatch):
+    dm = viewer
+    _clear_then_draw_card(dm)
+    saves = []
+    monkeypatch.setattr(dm, "_save_snapshot", lambda image: saves.append(image))
+    dm.write_owed_snapshot()                  # still inside the interval
+    assert saves == []
+    _age_last_write(dm)
+    dm.write_owed_snapshot()
+    assert len(saves) == 1
+    # Written: nothing is owed, so later frames do no work and the unchanged
+    # frame is not encoded again.
+    assert dm._snapshot_owed is False
+    _age_last_write(dm)
+    dm.write_owed_snapshot()
+    assert len(saves) == 1
+
+
+def test_a_failed_owed_write_stays_owed_and_is_retried(viewer, monkeypatch):
+    dm = viewer
+    _clear_then_draw_card(dm)
+    _age_last_write(dm)
+    attempts = []
+
+    def failing_save(image):
+        attempts.append(image)
+        raise OSError("disk full")
+
+    monkeypatch.setattr(dm, "_save_snapshot", failing_save)
+    dm.write_owed_snapshot()                  # the write fails
+    assert len(attempts) == 1
+    assert dm._snapshot_owed is True          # still owed: a held screen
+    saves = []                                # makes no update_display()
+    monkeypatch.setattr(dm, "_save_snapshot", lambda image: saves.append(image))
+    dm.write_owed_snapshot()                  # retried on the next frame
+    assert len(saves) == 1
+    assert dm._snapshot_owed is False
+
+
+def test_nothing_owed_after_a_frame_that_was_written(viewer, monkeypatch):
+    dm = viewer
+    dm.draw.rectangle([0, 0, 8, 8], fill=(0, 255, 0))
+    dm.update_display()                       # due: written at once
+    assert dm._snapshot_owed is False
+    calls = []
+    monkeypatch.setattr(dm, "_write_snapshot_if_due",
+                        lambda *a, **k: calls.append(a))
+    dm.write_owed_snapshot()
+    assert calls == []
+
+
+def test_an_unchanged_frame_inside_the_interval_is_not_owed(viewer):
+    dm = viewer
+    dm.draw.rectangle([0, 0, 8, 8], fill=(0, 0, 255))
+    dm.update_display()
+    dm.update_display()                       # same frame, inside the interval
+    assert dm._snapshot_owed is False
+
+
+def test_a_controller_without_the_hook_still_draws():
+    """Controllers built without a display manager (tests) are unaffected."""
+    from src import display_controller as dc_module
+    controller = dc_module.DisplayController.__new__(dc_module.DisplayController)
+    controller.plugin_manager = None
+    plugin = _HoldingPlugin()
+    assert controller._display_once(plugin, "x", True) is True
+    controller.display_manager = types.SimpleNamespace()
+    assert controller._display_once(plugin, "x", True) is True
+    assert plugin.calls == 2

test/test_state_stream_readers.py

@@ -612,6 +612,12 @@ class TestEndToEnd:
         cached['display_current_state'] = {'mode': 'from-cache', 'last_updated': 1}
         path = str(tmp_path / 'control.sock')
         monkeypatch.setenv(c.SOCKET_PATH_ENV, path)
+        # The display is this process, and its render loop is beating: the
+        # cache is then still its answer once the socket goes.
+        heartbeat = tmp_path / 'display-heartbeat.json'
+        heartbeat.write_text(json.dumps({'pid': os.getpid(), 'mono': time.monotonic(),
+                                         'wall': time.time()}))
+        monkeypatch.setattr(display_watchdog, 'HEARTBEAT_PATH', str(heartbeat))
         hub = _hub_with_everything()
         server = ControlServer(path, state_hub=hub, keepalive=0.2)
         assert server.start()
@@ -638,3 +644,8 @@ class TestEndToEnd:
                 break
             time.sleep(0.05)
         assert (data['mode'], data['source']) == ('from-cache', 'cache')
+        # Stopped: systemd takes the heartbeat's directory with it, and the
+        # cache's last answer is no longer anyone's.
+        heartbeat.unlink()
+        data = _data(client, '/api/v3/display/current-status')
+        assert (data['mode'], data['source']) == (None, 'cache')

test/web_interface/test_web_process_runs_no_plugin_code.py

@@ -221,7 +221,9 @@ class TestTheWebProcessNeverRunsAPlugin:
     def test_saving_its_section_through_the_main_config(self, web):
         body = web.post("/api/v3/config/main", {PLUGIN_ID: {"message": "via main"}})
         assert web.stored()["message"] == "via main"
-        assert body["restart_required"] is True
+        # The display's config watcher hands the section to the running
+        # plugin (on_config_change), as for /plugins/config: no restart.
+        assert body["restart_required"] is False
         assert web.ran() == []
 
     def test_resetting_its_config(self, web):

web_interface/blueprints/api_v3/config.py

@@ -17,6 +17,8 @@ from src.pi5_matrix_support import is_raspberry_pi_5
 from web_interface.cache import invalidate_cache
 from web_interface.auth import SECTION as _WEB_AUTH_SECTION, strip_auth_section
 import web_interface.blueprints.api_v3 as _pkg
+import copy
+from typing import Any, Dict, Iterable, Tuple
 
 # Read through the module rather than bound by value: tests patch these
 # as module attributes, and a value binding would not see the patch.
@@ -33,6 +35,50 @@ FORM_SECTION_FIELD = '__form_section'
 GENERAL_FIELDS = ('timezone', 'city', 'state', 'country', 'web_display_autostart',
                   'plugins_directory', 'auto_update_enabled', 'auto_update_channel')
 
+#: Settings in config.json the running display applies without a restart,
+#: as key paths (a path covers everything under it). Brightness goes over the
+#: control socket (brightness.set) and the config watcher's refresh
+#: (DisplayController._refresh_config_cache); the per-mode durations are read
+#: from the live config each time a mode starts (_get_display_duration).
+#: Plugin sections are live as well (each plugin's on_config_change), and
+#: save_main_config adds the ones a request saves.
+LIVE_CONFIG_PATHS: Tuple[Tuple[str, ...], ...] = (
+    ('display', 'hardware', 'brightness'),
+    ('display', 'display_durations'),
+)
+
+
+_MISSING = object()
+
+
+def _config_leaves(config: Any, prefix: Tuple[str, ...] = ()) -> Dict[Tuple[str, ...], Any]:
+    """Every non-dict value in ``config``, by key path. An empty dict has none,
+    so a section created empty on the way to a field is not a change."""
+    if not isinstance(config, dict):
+        return {prefix: config}
+    leaves: Dict[Tuple[str, ...], Any] = {}
+    for key, value in config.items():
+        leaves.update(_config_leaves(value, prefix + (str(key),)))
+    return leaves
+
+
+def restart_needed(before: Dict[str, Any], after: Dict[str, Any],
+                   live_paths: Iterable[Tuple[str, ...]] = LIVE_CONFIG_PATHS) -> bool:
+    """Does going from config ``before`` to ``after`` need a display restart?
+
+    True when anything changed outside ``live_paths``. A save that changes
+    only live settings, or nothing at all, does not.
+    """
+    live = tuple(live_paths)
+    old, new = _config_leaves(before), _config_leaves(after)
+    for path in set(old) | set(new):
+        if old.get(path, _MISSING) == new.get(path, _MISSING):
+            continue
+        if not any(path[:len(prefix)] == prefix for prefix in live):
+            return True
+    return False
+
+
 #: Top-level fields save_main_config stores somewhere of its own (location,
 #: plugin_system, ...), never as a config key of the same name.
 _MAPPED_TOP_LEVEL_FIELDS = GENERAL_FIELDS + (
@@ -72,6 +118,22 @@ def _day_setting(data, day, flat_key, nested_key):
     return False, None
 
 
+def _disabled_day_times(data, day, start_key, end_key):
+    """The times posted for a day that is off, the valid ones.
+
+    Nothing reads them while the day is off, but the schedule picker posts
+    them and GET returns them, so keeping them means turning the day back on
+    finds what was there. An invalid one is dropped rather than refused, for
+    the same reason.
+    """
+    times = {}
+    for field, key in (('start_time', start_key), ('end_time', end_key)):
+        value = _day_setting(data, day, key, field)[1]
+        if value and _validate_time_format(value)[0]:
+            times[field] = value
+    return times
+
+
 @api_v3.route('/config/main', methods=['GET'])
 def get_main_config():
     """Get main configuration, with credentials redacted."""
@@ -257,11 +319,16 @@ def save_schedule_config():
 
                     day_config['start_time'] = start_time
                     day_config['end_time'] = end_time
+                else:
+                    day_config.update(_disabled_day_times(data, day, start_key, end_key))
 
                 schedule_config['days'][day] = day_config
 
-            # Validate that at least one day is enabled in per-day mode
-            if enabled_days_count == 0:
+            # An enabled per-day schedule needs a day to be on. A disabled
+            # one does not: every day off with the schedule off is what
+            # config.template.json ships, so refusing it meant a fresh
+            # install could not post back the schedule GET returned.
+            if enabled_days_count == 0 and enabled_value:
                 return error_response(
                     ErrorCode.VALIDATION_ERROR,
                     "At least one day must be enabled in per-day schedule mode",
@@ -465,11 +532,13 @@ def save_dim_schedule_config():
 
                     day_config['start_time'] = start_time
                     day_config['end_time'] = end_time
+                else:
+                    day_config.update(_disabled_day_times(data, day, start_key, end_key))
 
                 dim_schedule_config['days'][day] = day_config
 
-            # Validate that at least one day is enabled in per-day mode
-            if enabled_days_count == 0:
+            # As for the on/off schedule: only an enabled one needs a day on.
+            if enabled_days_count == 0 and enabled_value:
                 return error_response(
                     ErrorCode.VALIDATION_ERROR,
                     "At least one day must be enabled in per-day dim schedule mode",
@@ -557,6 +626,8 @@ def save_main_config():
 
         # Merge with existing config (similar to original implementation)
         current_config = api_v3.config_manager.load_config()
+        # What was stored, to tell which settings this save changed.
+        stored_config = copy.deepcopy(current_config)
         was_auto_update_enabled = bool((current_config.get('auto_update') or {}).get('enabled'))
 
         is_general_update = any(k in data for k in GENERAL_FIELDS)
@@ -1194,13 +1265,15 @@ def save_main_config():
                 message = f'{message}. {note}'
         except Exception:
             logger.warning("Automatic update setup could not be started", exc_info=True)
-        # Display hardware, rotation/durations and general settings take
-        # effect after a display restart; the UI shows its restart banner on
-        # this flag.
-        extra = {'restart_required': True}
-        # Brightness is the exception: the display applies a saved one
-        # without a restart. Over the control socket it lands at once,
-        # instead of when the config watcher next looks (up to ~2 s).
+        # Display hardware, rotation order and general settings take effect
+        # after a display restart; the UI shows its restart banner on this
+        # flag. Brightness, mode durations and plugin settings are applied by
+        # the running display (LIVE_CONFIG_PATHS), so a save that changed
+        # only those -- or nothing -- does not ask for one.
+        live_paths = LIVE_CONFIG_PATHS + tuple((plugin_id,) for plugin_id in plugin_keys_to_remove)
+        extra = {'restart_required': restart_needed(stored_config, current_config, live_paths)}
+        # Over the control socket a saved brightness lands at once, instead
+        # of when the config watcher next looks (up to ~2 s).
         if 'brightness' in data:
             saved = (current_config.get('display', {}).get('hardware', {}) or {}).get('brightness')
             if isinstance(saved, int) and not isinstance(saved, bool):

web_interface/blueprints/api_v3/display.py

@@ -386,15 +386,22 @@ def get_current_display_status():
     Read from the display's state stream over the control socket when it is
     available (``source: "socket"``). Otherwise from what the display
     publishes to the shared cache (display_controller._publish_current_mode_state)
-    when the active mode changes (``source: "cache"``).
+    when the active mode changes (``source: "cache"``). Unknown (every field
+    None) when the socket and the heartbeat both say the display is gone
+    (display_state.display_gone).
     """
-    state = display_state.current_status(display_state.read_state())
+    snapshot = display_state.read_state()
+    state = display_state.current_status(snapshot)
     source = 'socket'
     if state is None:
         source = 'cache'
-        cache = _cache_manager()
-        # memory_ttl=0: written by the display service; see get_on_demand_status.
-        state = cache.get('display_current_state', max_age=120, memory_ttl=0)
+        # A stopped display leaves its last answer in the cache, where it
+        # read as on (is_display_active: true) for the 120 s max_age. With
+        # no socket and no live heartbeat there is no display behind it.
+        if not display_state.display_gone(snapshot):
+            cache = _cache_manager()
+            # memory_ttl=0: written by the display service; see get_on_demand_status.
+            state = cache.get('display_current_state', max_age=120, memory_ttl=0)
     if state is None:
         state = {
             'mode': None,

web_interface/blueprints/api_v3/misc.py

@@ -102,6 +102,7 @@ def get_health():
         # the only signal, as it always was.
         # The display reports the same beat's age over the control socket's
         # state stream, measured in memory; the file is the fallback.
+        snapshot = None
         try:
             snapshot = display_state.read_state()
             if snapshot is not None:
@@ -132,6 +133,26 @@ def get_health():
                 'error': 'see logs for details'
             }
 
+        # A stopped display service. The heartbeat's absence alone says
+        # nothing (the dev server, the emulator and Windows write none), so
+        # the overall status stayed "healthy" with the display down until the
+        # last preview frame it left aged past 60 s (hardware: stale). Together
+        # the three signals are definite: systemd says the service is not
+        # active, the control socket does not answer, and there is no live
+        # heartbeat (display_state.display_gone, which is never true where
+        # the platform has no socket or it is switched off).
+        try:
+            if (not display_service_status.get('active')
+                    and display_state.display_gone(snapshot)):
+                health_status['checks']['display_loop'] = {
+                    'status': 'stopped',
+                    'note': 'The display service is not running',
+                    'source': 'service',
+                }
+        except Exception:
+            logger.warning("Health check could not tell whether the display is stopped",
+                           exc_info=True)
+
         # Check hardware connectivity (if display manager available)
         try:
             snapshot_path = display_preview.SNAPSHOT_PATH

web_interface/display_state.py

@@ -133,6 +133,39 @@ def on_demand_state(snapshot: Optional[Dict[str, Any]],
     return state
 
 
+def display_gone(snapshot: Optional[Dict[str, Any]]) -> bool:
+    """Is there positively no display behind a fallback to the cache?
+
+    True only when the socket should be there (this platform has one and it
+    is not switched off) but gave no ``snapshot``, and the render loop's
+    heartbeat file says nothing is running either: it is absent (systemd
+    removes its directory when the service stops), stale, or written by a
+    process that no longer exists -- #726's rules for the runtime snapshot.
+    Then what the display last left in the cache is a dead process's answer.
+
+    False whenever the answer is in doubt: a snapshot came in, the socket is
+    off or unsupported (Windows, the test suite, a deliberate ``off``), or a
+    live heartbeat says the display is running without a socket (an older
+    display). Those read the cache exactly as before.
+    """
+    if snapshot is not None:
+        return False
+    if not socket_supported() or not client_socket_paths():
+        return False
+    from src import display_watchdog
+    from src.plugin_system.plugin_runtime import process_exists
+    heartbeat = display_watchdog.read_heartbeat(display_watchdog.HEARTBEAT_PATH)
+    if heartbeat is None:
+        return True
+    age = display_watchdog.heartbeat_age(heartbeat)
+    if age is None or age >= display_watchdog.HEARTBEAT_STALE_SECONDS:
+        return True
+    pid = heartbeat.get('pid')
+    if isinstance(pid, int) and not isinstance(pid, bool) and process_exists(pid) is False:
+        return True
+    return False
+
+
 def loop_heartbeat_age(snapshot: Optional[Dict[str, Any]]) -> Optional[float]:
     """The render loop's heartbeat age now; None when the display has no
     beat to report yet (or there is no snapshot)."""
@@ -141,5 +174,5 @@ def loop_heartbeat_age(snapshot: Optional[Dict[str, Any]]) -> Optional[float]:
     return control_client.snapshot_loop_age(snapshot)
 
 
-__all__ = ['current_status', 'loop_heartbeat_age', 'on_demand_state', 'read_state',
-           'stop_subscription']
+__all__ = ['current_status', 'display_gone', 'loop_heartbeat_age', 'on_demand_state',
+           'read_state', 'stop_subscription']