server.go starts listenAndServe in an anonymous function (closes #162)

The server's OnStart hook started the HTTP listener with
`go s.listenAndServe()`; the Go style guide asks for a function called
with `go` to be wrapped in an anonymous function, so it is now
`go func() { s.listenAndServe() }()`. No other go statement in the Go
code called a function directly. Nothing the server does changed.

Model: opus-5-5

.dockerignore

@@ -73,9 +73,12 @@
 
 # This repository's own entries, after the shared content above: the
 # frontend build, the viewport test's output, the binary `make build`
-# writes and the reports `make run` stores, and log files at the root.
+# writes and the reports `make run` stores, and Go test binaries,
+# coverage output and log files at the root.
 /dist
 /tmp
 /netwatch-server
 /data
+/*.test
+/*.out
 /*.log

.editorconfig

@@ -11,7 +11,8 @@ insert_final_newline = true
 [Makefile]
 indent_style = tab
 
-# This repository's own entries, after the shared content above.
-
 [*.go]
 indent_style = tab
+
+# This repository's own sections, such as one for another language it
+# uses, go below this comment, and a re-vendor keeps them.

.gitea/workflows/check.yml

@@ -1,9 +1,20 @@
 name: check
 on: [push]
+# Free the shared runner: a new push cancels only the same branch's older run.
+concurrency:
+    group: ${{ github.workflow }}-${{ github.ref }}
+    cancel-in-progress: true
 jobs:
     check:
         runs-on: ubuntu-latest
+        # Free the shared runner from a hung build.
+        timeout-minutes: 20
         steps:
             # actions/checkout v4.2.2, 2026-02-22
             - uses: actions/checkout@11bd71901bbe5b1630ceea73d27597364c9af683
+              # script/cibuild needs no token, so none is left in .git/config.
+              with:
+                  persist-credentials: false
+                  # All history and tags, so git describe finds the version tag.
+                  fetch-depth: 0
             - run: script/cibuild

.gitignore

@@ -27,7 +27,7 @@ node_modules/
 # Environment files. `*.env` covers bare `.env` and the `prod.env`
 # convention. Only the templates `example.env` and `sample.env` are
 # re-included below. A repository that commits any other template adds
-# its own negation after these lines, for example `!.env.example`.
+# its own negation at the end of this file, for example `!.env.example`.
 *.[eE][nN][vV]
 .[eE][nN][vV].*
 .[eE][nN][vV][rR][cC]
@@ -46,7 +46,9 @@ node_modules/
 [iI][dD]_[eE][dD]25519
 [iI][dD]_[eE][dD]25519_[sS][kK]
 
-# This repository's own entries, after the shared content above.
+# This repository's own entries, such as its build outputs, go below
+# this comment, and a re-vendor keeps them. Anchor a binary built at the
+# root: `/myapp`, never `myapp`, which also ignores `cmd/myapp/`.
 
 # Build output
 dist/

.golangci.yml

@@ -25,6 +25,8 @@ linters:
     # silenced by disabling that name, not by enabling the successor.
     - wsl              # Deprecated, replaced by wsl_v5
     - gomodguard       # Deprecated, replaced by gomodguard_v2
+    # Misses findings at random in v2.14.0; back once a pinned release fixes it
+    - canonicalheader
   settings:
     lll:
       line-length: 88

Dockerfile

@@ -10,11 +10,22 @@
 # phase covers the frontend as well, through a copy from a node stage.
 # Inside them each tool is invoked directly, never through make or
 # script/, whose lint and test are themselves docker builds.
+#
+# In every stage ARG CACHEBUST comes right after the steps that install
+# system packages and download Go's modules, or right after FROM in a
+# stage with neither, and every step that lints, tests or compiles
+# comes after it, as does yarn install in the node stages. script/
+# passes a new value on every build, which reruns every RUN step after
+# it. Docker keeps the layers above it until the base image, their own
+# lines or go.mod and go.sum change. The steps that compile mount Go's
+# build cache, and golangci-lint run also mounts golangci-lint's; both
+# survive from build to build.
 
 # Frontend lint stage: eslint with the rules in eslint.config.js. The
 # lint phase below runs it.
 # node:22-alpine as of 2026-02-22
 FROM node@sha256:e4bf2a82ad0a4037d28035ae71529873c069b13eb0455466ae0bc13363826e34 AS frontend-lint
+ARG CACHEBUST
 WORKDIR /app
 COPY package.json yarn.lock ./
 RUN yarn install --frozen-lockfile
@@ -29,8 +40,11 @@ FROM golangci/golangci-lint@sha256:ad862ba6b3798cbe0fd9fd7408d498fd74fbd2623a924
 WORKDIR /src
 COPY go.mod go.sum ./
 RUN go mod download
+ARG CACHEBUST
 COPY . .
-RUN golangci-lint run --config .golangci.yml ./...
+RUN --mount=type=cache,target=/root/.cache/go-build \
+    --mount=type=cache,target=/root/.cache/golangci-lint \
+    golangci-lint run --config .golangci.yml ./...
 # Nothing is wanted from frontend-lint; the copy is what makes this
 # phase run it.
 COPY --from=frontend-lint /app/yarn.lock /dev/null
@@ -43,6 +57,7 @@ COPY --from=frontend-lint /app/yarn.lock /dev/null
 # node would take a reporter option for one more file.
 # node:22-alpine as of 2026-02-22
 FROM node@sha256:e4bf2a82ad0a4037d28035ae71529873c069b13eb0455466ae0bc13363826e34 AS frontend-test
+ARG CACHEBUST
 WORKDIR /app
 COPY package.json yarn.lock ./
 RUN yarn install --frozen-lockfile
@@ -58,19 +73,19 @@ RUN NODE_OPTIONS=--test-reporter=dot timeout 90 yarn --silent run test || \
 # ships and the alpine one does not. -timeout 90s is a backstop above
 # the 60-second cap on the suite. The rerun with -v only shows details:
 # the step fails however it ends, because the first run already failed.
-# Go's module and build caches are in memory (tmpfs) for the go test
+# -count=1 because the mounted cache also holds test results, which go
-# step alone, so the image make test tags does not carry them and the
+# test would otherwise replay.
-# build spends no time writing them into it. They start empty on every
-# build, so no stored test result can stand in for a run.
 # golang:1.25.7-trixie, 2026-10-07
 FROM golang@sha256:2b174ffcf56c7ad0c47d30d2630693265639ddf2a5141149c2da34db921791b4 AS test
 WORKDIR /src
+COPY go.mod go.sum ./
+RUN go mod download
+ARG CACHEBUST
 COPY . .
-RUN --mount=type=tmpfs,target=/go/pkg/mod \
+RUN --mount=type=cache,target=/root/.cache/go-build \
-    --mount=type=tmpfs,target=/root/.cache/go-build \
+    go test -count=1 -timeout 90s -race -cover ./... || \
-    go test -timeout 90s -race -cover ./... || \
     { echo "--- Rerunning with -v for details ---"; \
-      go test -timeout 90s -race -v ./...; exit 1; }
+      go test -count=1 -timeout 90s -race -v ./...; exit 1; }
 # Nothing is wanted from the frontend-test stage; the copy is what makes
 # this phase run its tests.
 COPY --from=frontend-test /app/yarn.lock /dev/null
@@ -83,23 +98,28 @@ COPY --from=frontend-test /app/yarn.lock /dev/null
 FROM frontend-test AS frontend-build
 COPY --from=lint /src/go.sum /dev/null
 COPY --from=test /src/go.sum /dev/null
-# vite.config.js reads the commit for the page's footer with git.
+# vite.config.js reads the commit for the page's footer with git. git is
+# installed again on every image build, although it comes above
+# ARG CACHEBUST, because this stage starts from frontend-test, whose
+# steps run again on every build.
 RUN apk add --no-cache git
+ARG CACHEBUST
 RUN yarn build
 
 # Backend build stage. Nothing is wanted from the two phases; the copies
-# are what make BuildKit build them first, so this stage cannot run
+# are what make BuildKit build them first, so nothing after them can
-# unless lint and test passed.
+# run unless lint and test passed.
 # golang:1.25-alpine (2026-02-27)
 FROM golang@sha256:f6751d823c26342f9506c03797d2527668d095b0a15f1862cddb4d927a7a4ced AS builder
-COPY --from=lint /src/go.sum /dev/null
-COPY --from=test /src/go.sum /dev/null
 RUN apk add --no-cache git
-# A tar-stream context keeps the sender's file owners, which git refuses.
-RUN git config --system --add safe.directory /src
 WORKDIR /src
 COPY go.mod go.sum ./
 RUN go mod download
+ARG CACHEBUST
+COPY --from=lint /src/go.sum /dev/null
+COPY --from=test /src/go.sum /dev/null
+# A tar-stream context keeps the sender's file owners, which git refuses.
+RUN git config --system --add safe.directory /src
 COPY . .
 
 # The same go build as in script/build, which builds the binary on the
@@ -112,7 +132,8 @@ COPY . .
 # still empty, dev or unknown fails the build: git is missing or could
 # not read the checkout.
 ARG VERSION
-RUN version="${VERSION:-$(git describe --tags --always)}"; \
+RUN --mount=type=cache,target=/root/.cache/go-build \
+    version="${VERSION:-$(git describe --tags --always)}"; \
     if [ -e .git ]; then \
         case "$version" in ""|dev|unknown) \
             echo "version is '$version' although .git is present" >&2; \
@@ -127,6 +148,7 @@ RUN version="${VERSION:-$(git describe --tags --always)}"; \
 # and the stages it copies from, the two phases included.
 # nginx:stable-alpine as of 2026-02-22
 FROM nginx@sha256:15e96e59aa3b0aada3a121296e3bce117721f42d88f5f64217ef4b18f458c6ab
+ARG CACHEBUST
 
 # netwatch-server runs as this user, which owns the report directory.
 # nginx keeps the image's own arrangement: its main process runs as

README.md

@@ -41,11 +41,13 @@ The scripts cover both the frontend and the Go backend, so `make check` fails if
 either one is broken. We provide:
 
 - `script/bootstrap` — install all dependencies (the pinned node via nvm unless
-  one new enough for the frontend's dependencies is installed, yarn via
+  the `node` on `PATH` has the pinned major version and is new enough for the
-  corepack, `yarn install --frozen-lockfile`, Go `GO_VERSION` unless the `go` on
+  frontend's dependencies, yarn `YARN_VERSION` via corepack, or npm where there
-  `PATH` is exactly that version, and the Go modules), linking what it installs
+  is no corepack, unless the `yarn` on `PATH` is exactly that version,
-  itself into `~/.local/bin`, which has to be on `PATH`. It installs no Go
+  `yarn install --frozen-lockfile`, Go `GO_VERSION` unless the `go` on `PATH` is
-  linter and not Docker: `make test` and `make lint` run in Docker
+  exactly that version, and the Go modules), linking what it installs itself
+  into `~/.local/bin`, which has to be on `PATH`. It installs no Go linter and
+  not Docker: `make test` and `make lint` run in Docker
 - `script/setup` — make a fresh clone ready for development: bootstrap plus the
   git pre-commit hook
 - `script/dev` — run the Vite dev server, which proxies `/api` to a locally
@@ -53,16 +55,23 @@ either one is broken. We provide:
 - `script/build` — build the frontend for production into `dist/`, then the
   static `netwatch-server` binary into the repo root with its version stamped
   in: `VERSION` from the environment, or when that is unset or empty,
-  `git describe` inside a git checkout, then `dev`
+  `git describe --tags --always --dirty` inside a git checkout, as
+  `script/docker` does, then `dev`
 - `script/run` — build with `script/build`, then run `netwatch-server` with its
   defaults (see [Settings](#settings))
 - `script/clean` — remove what `script/build` writes
 - `script/projectname` — print the project name (used for the Docker image tag)
-- `script/test` — build the `test` phase of `Dockerfile` without the cache: the
+- `script/yarn` — run yarn with the given arguments under the node
+  `script/bootstrap` would use: with `~/.local/bin` first on `PATH`, the first
+  `node` on that `PATH` when it has the pinned major version and is new enough
+  for the frontend's dependencies, and otherwise the pinned node in nvm. Every
+  script that runs yarn on the host runs it through this one, except
+  `script/bootstrap`, which picks its own node and yarn
+- `script/test` — build the `test` phase of `Dockerfile`, writing no image: the
   frontend's unit tests in its `frontend-test` stage, and the backend's Go tests
   with the race detector and coverage. The frontend's production build is left
   to the image build, in the `frontend-build` stage
-- `script/lint` — build the `lint` phase of `Dockerfile` without the cache:
+- `script/lint` — build the `lint` phase of `Dockerfile`, writing no image:
   eslint in its `frontend-lint` stage, and golangci-lint over the Go packages
 - `script/fmt` — format all files (writes): prettier over the JavaScript, CSS,
   HTML and Markdown, then gofmt over `cmd/` and `internal/`. It runs on the
@@ -90,13 +99,25 @@ either one is broken. We provide:
   starts, it checks that the page, and a path that falls back to it, are served
   with `Cache-Control: no-cache` and `Content-Security-Policy`. Not part of
   `script/check`: it takes minutes.
-- `script/docker` — build the image from `Dockerfile` without the build cache,
+- `script/docker` — build the image from `Dockerfile`, tagged `netwatch` via
-  tagged `netwatch` via `script/projectname`
+  `script/projectname`
 - `script/cibuild` — CI entrypoint: runs `script/bootstrap` and `script/check`,
-  then builds the image as `script/docker` does, without the build cache
+  then builds the image as `script/docker` does
 - `script/precommit` — run by the git pre-commit hook; runs `script/check`
 - `script/install-precommit` — install the git pre-commit hook
 
+`script/test`, `script/lint`, `script/docker` and `script/cibuild` each pass a
+new `CACHEBUST` value, so every step in `Dockerfile` that lints, tests or
+compiles runs again even when nothing has changed. Docker keeps the system
+packages and the Go modules from earlier builds until the base image, the step
+that installs them, or `go.mod` and `go.sum` change. The `frontend-build` stage
+is the exception: it installs git again on every image build, because it starts
+from the `frontend-test` stage, whose steps run again on every build. Go's build
+cache and golangci-lint's are kept as well, and each reuses a result only when
+everything it was made from is unchanged. Both `go test` runs carry `-count=1`,
+so no test result is reused. The yarn packages are installed again on every
+build.
+
 ## Responsive layout
 
 The narrow-viewport layout lives in the `max-width: 768px` media block in
@@ -155,8 +176,26 @@ report a delta, so only new samples are sent; the mark advances only on a
 delivered report, and while paused nothing is sent. Changing the interval clears
 each host's history on the page; the samples in it not yet reported still go in
 the next report. Delivery failure is quiet — one debug-log line per outage,
-retried at the next interval, never blocking probing. The report-building step
+never blocking probing. The report-building step is a pure function of host
-is a pure function of host state.
+state.
+
+A sample in the history that no report has delivered is sent only as long as the
+history holds it, so after a refused or failed report the next report carries
+the samples the history still has. Samples an interval change cleared from the
+history are kept apart, the last 100 per host, until they are delivered. The
+history keeps each host's last 100 rounds: 100 seconds at the 1-second interval,
+200 seconds at 2 seconds, 5 minutes at the default of 3 seconds and 100 minutes
+at 60 seconds. At 1 second, one refused or failed report loses the oldest 20
+seconds of each host's samples, as the next delivered report comes 2 minutes
+after the last. At 2 seconds and longer one loses nothing; samples are lost from
+the third refused report in a row at 2 seconds, from the fifth at 3 seconds, and
+later still at the longer intervals.
+
+No report is sent while the page is paused, but each paused round still adds an
+entry to the history, so each report due during a pause counts as a refused one.
+A pause longer than the history spans loses the samples in the history recorded
+between the last delivered report and the pause: up to a minute's worth for each
+host if the report before the pause was delivered.
 
 ### Backend
 
@@ -169,7 +208,9 @@ stops. Its routes:
   address may send a limited number a minute, and the report files are capped in
   size, the oldest deleted first
 - `GET /.well-known/healthcheck` — answers 200 with `"status":"ok"`, the
-  server's version and its uptime
+  server's version and its uptime; while reports cannot be stored, as the last
+  write of a report file failed, it answers 503 with `"status":"error"` and the
+  same fields (see [Report storage](#report-storage))
 - `GET /metrics` — Prometheus metrics behind basic auth, only when
   `METRICS_USERNAME` and `METRICS_PASSWORD` are set; each client address may
   make a limited number of requests to it a minute
@@ -177,6 +218,10 @@ stops. Its routes:
 It is an `fx`-wired Go application: `cmd/netwatch-server/` starts it, and its
 packages are in `internal/`:
 
+- **`cli`**: Runs `netwatch-server` from its command line: the `check-cidr` and
+  `prepare-data-dir` subcommands, which `bin/entrypoint.sh` runs before it
+  starts the server, and, given no arguments, the server, with the `fx` wiring.
+  It refuses any other arguments with a usage message and exit code 2.
 - **`config`**: Loads configuration from environment variables and config files
   via Viper.
 - **`handlers`**: HTTP request handlers for the API (health check, report
@@ -201,7 +246,7 @@ image, `make run` builds it and runs it on port 8080.
 | `PORT`                 | `8080`               | HTTP listen port                                                                                         |
 | `DATA_DIR`             | `./data/reports`     | Directory for compressed reports                                                                         |
 | `DATA_DIR_MAX_BYTES`   | `1073741824` (1 GiB) | Most bytes of report files kept in `DATA_DIR`, oldest deleted first; see [Report limits](#report-limits) |
-| `DEBUG`                | `false`              | Enable debug logging                                                                                     |
+| `DEBUG`                | empty                | Debug logging; any non-empty value, `false` and `0` included, turns it on                                |
 | `TRUSTED_PROXIES`      | RFC 1918 ranges      | Comma-separated CIDRs whose `X-Forwarded-For` / `X-Real-IP` headers are trusted for client IP resolution |
 | `REPORTS_PER_MINUTE`   | `60`                 | Reports each client address may send a minute; see [Report limits](#report-limits)                       |
 | `CORS_ALLOWED_ORIGINS` | empty                | Comma-separated origins whose pages may call the API; see [CORS](#cors)                                  |
@@ -224,10 +269,10 @@ an entry that is not an address at all, such as `unknown`, the direct peer is
 used. `X-Real-IP` is read only when `X-Forwarded-For` is missing or empty. The
 container image does not use this default; see [Deployment](#deployment).
 
-A variable set to a value the server cannot use, such as `PORT=abc`,
+A variable set to a value the server cannot use, such as `PORT=abc` or a
-`DEBUG=maybe` or a `BIND_ADDRESS` that is not an IP address, stops it from
+`BIND_ADDRESS` that is not an IP address, stops it from starting, with an error
-starting, with an error naming the variable. An empty variable counts as unset,
+naming the variable. An empty variable counts as unset, except
-except `TRUSTED_PROXIES`.
+`TRUSTED_PROXIES`. No value of `DEBUG` stops the start.
 
 #### Report storage
 
@@ -241,18 +286,23 @@ toward `DATA_DIR_MAX_BYTES` from the next start, only if removing it fails too.
 Files are created with `O_EXCL` to prevent overwrites.
 
 The reports of a failed write stay in memory, and the next flush, within a
-minute, tries again to write them. From the failed write until a write succeeds,
+minute, tries again to write them. The server writes one file at a time, so a
-`POST /api/v1/reports` refuses every report with 500 and keeps nothing of it, so
+file holds its reports in the order they arrived, those of a failed write
-the page keeps its samples and sends them again at its next report. Reports
+included. From the failed write until a write succeeds, `POST /api/v1/reports`
-already answered 200 are lost only if the server stops before a write succeeds;
+refuses every report with 500 and keeps nothing of it, and the health check
-the final flush at stop tries them too, and if it fails, the process exits
+answers 503 with `"status":"error"`. The page sends the refused samples again in
-non-zero.
+its next report, as long as it still holds them: at the 1-second interval one
+refused report already loses some of those in the history (see
+[Reporting](#reporting)). Reports already answered 200 are lost only if the
+server stops before a write succeeds; the final flush at stop tries them too,
+and if it fails, the process exits non-zero.
 
 #### Report limits
 
 `POST /api/v1/reports` takes reports from anyone who can reach it, without
 credentials, so it is bounded instead. Each refusal below answers with the same
-`{"status":"error"}` body as any other error.
+`{"status":"error"}` body as any other error, except nginx's refusal of a body
+over the limit in the image.
 
 - **JSON only.** A report must be sent with `Content-Type: application/json`,
   parameters such as `charset` allowed, as the page sends it; any other is
@@ -260,6 +310,16 @@ credentials, so it is bounded instead. Each refusal below answers with the same
   send a `text/plain` POST, or a form, without a CORS preflight; refusing them
   keeps any site from making its visitors' browsers store reports (see
   [CORS](#cors)).
+- **Body limit.** A report body may be at most 1 MiB, the backend's limit for
+  any request. Run on its own, the backend answers 413 before the route to a
+  request that declares a longer body. Without a declared length, the route
+  answers 413 only when the report it reads runs past 1 MiB; nothing after the
+  end of the report is read, so a longer body whose report ends sooner is
+  stored, and one that stops being valid JSON sooner is answered 400. In the
+  image, nginx refuses a body over 1 MiB first, with or without a declared
+  length, as its `client_max_body_size`, which `configs/nginx.conf` does not
+  set, is 1 MiB by default. nginx answers 413 with its own HTML error page, and
+  the backend never sees the request.
 - **Rate limit.** Each client address, resolved through `TRUSTED_PROXIES`, may
   send `REPORTS_PER_MINUTE` reports a minute; past that it gets 429 with
   `Retry-After: 60`. The minute slides: reports from the minute before still
@@ -283,6 +343,12 @@ credentials, so it is bounded instead. Each refusal below answers with the same
   how much of the newest reports is kept: the default of 1 GiB is small enough
   for any host; set it to the space you can give `DATA_DIR`.
 
+At the defaults, one address may send 60 reports of up to 1 MiB a minute, and a
+report made up to compress badly takes about as much room in the report files as
+its request body, so in about 17 minutes one address can fill the 1 GiB cap with
+its own reports, every report stored before them deleted to make room. Lowering
+`REPORTS_PER_MINUTE` or raising `DATA_DIR_MAX_BYTES` makes that take longer.
+
 #### CORS
 
 The page calls the API from the origin it is served from, so by default the
@@ -304,12 +370,16 @@ the metrics record its duration and response size, labelled with its path,
 method and status; they also count those requests in progress, and include Go's
 runtime and process metrics. No other request is recorded: not those to
 `/metrics` itself, and not those answered before they reach either route, such
-as a CORS preflight, or a request refused with 404 for a path no route has, 405
+as a CORS preflight, or a request refused with 404 for a path no route has or
-for a method its route does not take, or 413 for declaring a body length over
+405 for a method its route does not take. Run on its own, the backend also
-the 1 MiB limit. A report whose body goes over the limit without declaring its
+answers 413 before the route to a request declaring a body length over the 1 MiB
-length reaches the route, is answered 413 there, and is recorded with that
+limit, which is not recorded. A body over the limit sent without a declared
-status. Clients can make up any number of paths and methods, and each would add
+length reaches the route and is recorded with the status the route answers,
-labels to the metrics for as long as the server runs.
+which is 413 only when the report it reads runs past the limit: nothing after
+the end of the report is read. In the image neither reaches the backend: nginx
+answers both with 413 itself (see [Report limits](#report-limits)), and nothing
+of them is recorded. Clients can make up any number of paths and methods, and
+each would add labels to the metrics for as long as the server runs.
 
 `/metrics` is rate limited, so that its password cannot be guessed quickly: each
 client address, resolved through `TRUSTED_PROXIES`, may make 60 requests to it a
@@ -468,7 +538,8 @@ What the [upaas](https://git.eeqj.de/sneak/upaas) app for netwatch needs:
       report files may take; the oldest are deleted to stay under it
     - `CORS_ALLOWED_ORIGINS`, default empty: other origins whose pages may call
       the API
-    - `DEBUG`, default `false`: debug logging
+    - `DEBUG`, default empty: debug logging; any non-empty value, `false` and
+      `0` included, turns it on
     - `DATA_DIR`, default `/data/reports`: the directory the reports are kept
       in: `/data` or a path below it, with no `.` or `..` part and no extra `/`.
       The container also stops if the path goes through a symbolic link that
@@ -496,9 +567,16 @@ What the [upaas](https://git.eeqj.de/sneak/upaas) app for netwatch needs:
       stops the container. Empty, the backend sends nothing to Sentry
 - **Health check:** the image's `HEALTHCHECK` requests
   `/.well-known/healthcheck` through nginx every 30 seconds, so it fails unless
-  both nginx and the backend answer. upaas reads the container's health 60
+  both nginx and the backend answer. It also fails while reports cannot be
-  seconds after a deploy and fails the deploy unless it is `healthy`. The
+  stored, as when the volume is full or `DATA_DIR` cannot be written: the health
-  container also stops when either process exits.
+  check answers 503 from a failed write of a report file until a write succeeds,
+  and the backend tries the write again every minute. Three failed checks in a
+  row make the container `unhealthy`, and one that passes makes it `healthy`
+  again. upaas reads the container's health once, 60 seconds after a deploy, and
+  fails the deploy unless it is `healthy`. A container that turns `unhealthy`
+  after that is neither restarted nor reported by upaas, whose `unless-stopped`
+  restart policy ignores health; only `docker ps` and `docker inspect` show it.
+  The container also stops when either process exits.
 
 ## Browser Compatibility
 

REPO_POLICIES.md

@@ -1,6 +1,6 @@
 ---
 title: Repository Policies
-last_modified: 2026-10-04
+last_modified: 2026-10-18
 ---
 
 This document covers repository structure, tooling, and workflow standards. Code
@@ -54,34 +54,39 @@ style conventions are in separate documents:
   `cibuild`. `script/bootstrap` installs all dependencies idempotently and
   assumes nothing is present: base tools come from nix, apt, brew, or apk
   (detected in that order; apt runs noninteractive). For node it uses the
-  installed node if present; otherwise it installs a PINNED node version via
+  installed node only when its major version is the pinned one; otherwise (node
-  nvm, first installing nvm itself if missing — from a hash-verified GitHub
+  missing, or another major, such as the node 24 the CI runner image ships) it
-  release archive (never `curl | sh`), with bash installed as an explicit
+  installs the PINNED node version via nvm and installs yarn and the packages
-  prerequisite since nvm requires bash. yarn is then pinned via
+  under it. nvm itself is installed first if missing, from a hash-verified
-  `corepack prepare yarn@<version> --activate`. Never install "latest" or "lts";
+  GitHub release archive (never `curl | sh`), with bash installed as an explicit
-  always exact versions. `script/cibuild` runs the CI build: it changes to the
+  prerequisite since nvm requires bash. Only the major version is compared
-  repo root, runs `script/bootstrap`, runs `script/check`, and builds the image
+  because the `Dockerfile` stages start from a node image whose exact version is
-  with the version; the Gitea workflow calls it. **`script/cibuild` runs
+  not the pin, and nvm cannot install a prebuilt node on alpine. yarn is pinned
+  via `corepack prepare yarn@<version> --activate`. Never install "latest" or
+  "lts"; always exact versions. `script/cibuild` runs the CI build: it changes
+  to the repo root, runs `script/bootstrap`, runs `script/check`, and builds the
+  image with the version; the Gitea workflow calls it. **`script/cibuild` runs
   `script/bootstrap` first**, because the workflow checks out the repo and runs
   nothing else, while `script/fmt-check` runs the formatter on the host: on a
   pristine checkout with nothing installed the run dies there, after the
-  containerised gates have passed. **The bootstrap alone is not enough**:
+  containerised gates have passed. **The bootstrap alone is not enough**: when
-  `script/bootstrap` installs node and yarn under nvm and leaves neither on the
+  `script/bootstrap` installs node and yarn under nvm it leaves neither on the
-  `PATH` of the shell that called it, so a bare `yarn` still exits 127. The host
+  `PATH` of the shell that called it, so a bare `yarn` either exits 127 or runs
-  entrypoints that need yarn — `script/fmt` and `script/fmt-check` — therefore
+  under another node. The host entrypoints that need yarn — `script/fmt` and
-  source nvm for the pinned node version before invoking it, exactly as
+  `script/fmt-check` — therefore use the same test as `script/bootstrap`: when
-  `script/bootstrap`'s own install step does. A runner carrying nothing but
+  the node on `PATH` has the pinned major version they run the `yarn` on `PATH`,
-  docker and git then gets through `script/check`. Four further scripts are our
+  and otherwise they load nvm through `$HOME/.nvm/nvm.sh` and run yarn under the
-  own extensions to the standard: `script/check` runs `script/test`,
+  pinned node. A runner carrying nothing but docker and git then gets through
-  `script/lint` and `script/fmt-check`; `script/precommit` is what the git
+  `script/check`. Four further scripts are our own extensions to the standard:
-  pre-commit hook runs, and it calls `script/check`; `script/install-precommit`
+  `script/check` runs `script/test`, `script/lint` and `script/fmt-check`;
-  installs the git pre-commit hook (the `make hooks` target shims to it); and
+  `script/precommit` is what the git pre-commit hook runs, and it calls
-  `script/projectname` (literally that filename) simply outputs the project's
+  `script/check`; `script/install-precommit` installs the git pre-commit hook
-  name. Scripts that need the name call `script/projectname` — e.g.
+  (the `make hooks` target shims to it); and `script/projectname` (literally
-  `script/docker` assembles its image tag from it — so those scripts stay
+  that filename) simply outputs the project's name. Scripts that need the name
-  byte-identical across all repos. Repo-type-specific pre-commit extras (e.g.
+  call `script/projectname` — e.g. `script/docker` assembles its image tag from
-  `go mod tidy` verification in Go repos) belong in `script/precommit`, not in
+  it — so those scripts stay byte-identical across all repos. Repo-type-specific
-  the hook itself. Model scripts are at
+  pre-commit extras (e.g. `go mod tidy` verification in Go repos) belong in
+  `script/precommit`, not in the hook itself. Model scripts are at
   `https://git.eeqj.de/sneak/prompts/raw/branch/main/script/<name>`. The README
   must document the provided scripts in an **Entrypoints** section (see the
   README requirements below).
@@ -118,8 +123,10 @@ style conventions are in separate documents:
   and nothing else:
 
     ```sh
-    docker build --no-cache --target lint -t "$(script/projectname)-lint" .
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
-    docker build --no-cache --target test -t "$(script/projectname)-test" .
+        --target lint --output type=cacheonly .
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
+        --target test --output type=cacheonly .
     ```
 
     **A stage that is not the last one in the file is built only when the final
@@ -129,10 +136,15 @@ style conventions are in separate documents:
     plain `docker build .` builds the last stage alone and exits 0 having linted
     and tested nothing.
 
-    **Every `docker build` in `script/` is tagged**, here and in
+    **The gate builds write no image.** With `--output type=cacheonly` the phase
-    `script/cibuild` and `script/docker`. An untagged build leaves a dangling
+    runs and a failing step fails the build, but the result is not exported.
-    image behind on every invocation, on every developer host and every CI
+    Nothing uses those images, and writing one out is slow: a Go test phase's
-    runner; a tagged one replaces the previous image.
+    image holds the toolchain. A build given neither `--output` nor `-t` writes
+    an untagged image and leaves it dangling, on every developer host and every
+    CI runner. `script/cibuild` and `script/docker` build the image that ships
+    and tag it, so each build replaces the previous image; each assigns the tag
+    on its own line before the build, so `set -e` stops it where
+    `script/projectname` fails.
 
     Inside a phase the tool is invoked directly — `golangci-lint`, `go test`,
     `eslint`, `prettier` — never through `make lint` or `script/test`, which are
@@ -147,22 +159,42 @@ style conventions are in separate documents:
     one's findings, and a host-global lock in `$TMPDIR` makes concurrent runs
     exit non-zero with `parallel golangci-lint is running` — a status a caller
     cannot tell from real findings. Both have produced wrong verdicts in this
-    org, in both directions. A container has its own cache, its own `TMPDIR` and
+    org, in both directions. A container has its own `TMPDIR`, so the lock is
-    a digest-pinned binary, so neither is reachable.
+    out of reach. The lint phase does keep golangci-lint's cache between check
+    builds (below), but it lints every checkout at `/src`, and a finding is
+    taken from that cache only for the same files, settings and golangci-lint
+    version.
 
-- **Any build that runs checks is built with `--no-cache`.** Docker invalidates
+- **Any build that runs checks is given a new `CACHEBUST` value.** Docker
-  a `COPY` layer only when the copied content changes, so on an unchanged tree
+  invalidates a `COPY` layer only when the copied content changes, so on an
-  the check `RUN` is served from cache, nothing executes, and the build still
+  unchanged tree the check `RUN` is served from cache, nothing executes, and the
-  exits 0. Every `docker build` in `script/` therefore passes `--no-cache`:
+  build still exits 0. Every `docker build` in `script/` therefore passes
-  `script/lint`, `script/test`, `script/cibuild` and `script/docker` are the
+  `--build-arg CACHEBUST="$(date +%s)-$$"`, the time and the script's process
-  four, and there is no fifth — `script/check` runs the two gate phases and
+  id, so no two builds share a value, not even the three in one
-  `script/fmt-check`, and builds no image of its own. A bare `docker build .` is
+  `script/cibuild`. `script/lint`, `script/test`, `script/cibuild` and
-  not evidence that anything ran: a sub-second build reporting success is a
+  `script/docker` are the four, and there is no fifth — `script/check` runs the
-  cache hit, not a result. Never invalidate by pruning — `docker builder prune`
+  two gate phases and `script/fmt-check`, and builds no image of its own. In
-  and friends destroy a build cache shared with every other build on the host.
+  every stage of the `Dockerfile`, `ARG CACHEBUST` comes right after the steps
-  When a check is added or changed, prove it works by planting a defect it must
+  that install system packages and download Go's modules, or right after `FROM`
-  catch and watching the run fail on it, then revert the defect. A green run
+  in a stage with neither, and every step that lints, tests or compiles comes
-  alone shows neither that the check ran nor that it covers what it should.
+  after it. A step that installs anything else as well, such as a
+  `script/bootstrap` that also installs node packages, comes after it too. A new
+  value reruns every `RUN` step after that line; `COPY` steps are compared by
+  content and need nothing. Docker keeps the layers above the line until the
+  base image, their own lines or the files they copy change. A stage without the
+  `ARG` line is served from cache on an unchanged tree, so a repository
+  re-vendors the scripts and adds the `ARG` lines in one commit. The scripts do
+  not pass `--no-cache`: it reruns every step too, but it also builds the layers
+  above the line again and empties every cache mount, so Go's build cache and
+  golangci-lint's, which the canonical Go `Dockerfile` below mounts so that they
+  survive between check builds, would start empty every time. A bare
+  `docker build .` is not evidence that anything ran: a sub-second build
+  reporting success is a cache hit, not a result. Never invalidate by pruning —
+  `docker builder prune` and friends destroy a build cache shared with every
+  other build on the host. When a check is added or changed, prove it works by
+  planting a defect it must catch and watching the run fail on it, then revert
+  the defect. A green run alone shows neither that the check ran nor that it
+  covers what it should.
 
 - **The gate phases are separate stages, and the build stage depends on both.**
   The lint phase is based on the `golangci/golangci-lint` image (pinned by
@@ -171,40 +203,56 @@ style conventions are in separate documents:
   `Dockerfile`:
 
     ```dockerfile
+    # In every stage ARG CACHEBUST comes right after the steps that install
+    # system packages and download Go's modules, or right after FROM in a
+    # stage with neither, and every step that lints, tests or compiles
+    # comes after it. script/ passes a new value on every build, which
+    # reruns every RUN step after it. Docker keeps the layers above it
+    # until the base image, their own lines or go.mod and go.sum change.
+    # The steps that compile mount Go's build cache, and golangci-lint run
+    # also mounts golangci-lint's; both survive from build to build.
+
     # Lint phase
     # golangci/golangci-lint:v2.x.x, YYYY-MM-DD
     FROM golangci/golangci-lint@sha256:... AS lint
     WORKDIR /src
     COPY go.mod go.sum ./
     RUN go mod download
+    ARG CACHEBUST
     COPY . .
-    RUN golangci-lint run --config .golangci.yml ./...
+    RUN --mount=type=cache,target=/root/.cache/go-build \
+        --mount=type=cache,target=/root/.cache/golangci-lint \
+        golangci-lint run --config .golangci.yml ./...
 
     # Test phase. -race needs cgo and so a C compiler, which the Debian Go
-    # image ships and the alpine one does not.
+    # image ships and the alpine one does not. -count=1 because the mounted
+    # cache also holds test results, which go test would otherwise replay.
     # golang:1.x, YYYY-MM-DD
     FROM golang@sha256:... AS test
     WORKDIR /src
     COPY go.mod go.sum ./
     RUN go mod download
+    ARG CACHEBUST
     COPY . .
-    RUN go test -timeout 90s -race -cover ./... || \
+    RUN --mount=type=cache,target=/root/.cache/go-build \
+        go test -count=1 -timeout 90s -race -cover ./... || \
         { echo "--- Rerunning with -v for details ---"; \
-          go test -timeout 90s -race -v ./...; exit 1; }
+          go test -count=1 -timeout 90s -race -v ./...; exit 1; }
 
     # Build stage. Nothing is wanted from either phase above; the copies
-    # are what make BuildKit build them first, so this stage cannot run
+    # are what make BuildKit build them first, so nothing after them can
-    # unless lint and test passed.
+    # run unless lint and test passed.
     # golang:1.x-alpine, YYYY-MM-DD
     FROM golang@sha256:... AS builder
-    COPY --from=lint /src/go.sum /dev/null
-    COPY --from=test /src/go.sum /dev/null
     RUN apk add --no-cache git
-    # A tar-stream context keeps the sender's file owners, which git refuses.
-    RUN git config --system --add safe.directory /src
     WORKDIR /src
     COPY go.mod go.sum ./
     RUN go mod download
+    ARG CACHEBUST
+    COPY --from=lint /src/go.sum /dev/null
+    COPY --from=test /src/go.sum /dev/null
+    # A tar-stream context keeps the sender's file owners, which git refuses.
+    RUN git config --system --add safe.directory /src
     COPY . .
 
     # The VERSION build arg when one is given, otherwise
@@ -212,7 +260,8 @@ style conventions are in separate documents:
     # .git present, a version that is still empty, dev or unknown fails the
     # build: git is missing or could not read the checkout.
     ARG VERSION
-    RUN VERSION="${VERSION:-$(git describe --tags --always)}"; \
+    RUN --mount=type=cache,target=/root/.cache/go-build \
+        VERSION="${VERSION:-$(git describe --tags --always)}"; \
         if [ -e .git ]; then \
             case "$VERSION" in ""|dev|unknown) \
                 echo "version is '$VERSION' although .git is present" >&2; \
@@ -225,6 +274,7 @@ style conventions are in separate documents:
 
     # Runtime stage, and the last one
     FROM alpine@sha256:...
+    ARG CACHEBUST
     COPY --from=builder /app /usr/local/bin/app
     ENTRYPOINT ["app"]
     ```
@@ -239,6 +289,26 @@ style conventions are in separate documents:
     - Keep the runtime stage last, and if you add a stage after it, give it the
       same two copies. A plain `docker build .` builds the last stage's chain
       and nothing else.
+    - Check builds keep these, and none of them can hand back a stale result:
+        - The layers above `ARG CACHEBUST`: the system packages and Go's
+          modules. Docker installs the packages again when the base image or the
+          install line changes, and downloads the modules again when `go.mod` or
+          `go.sum` changes. Go checks every module against `go.sum` as it
+          downloads it, and no later build can change a kept layer. The modules
+          are kept in a layer rather than a cache mount because a cache mount is
+          shared by every build on the host, and Go builds from whatever files
+          it holds without checking them again (only `go mod verify` does), so a
+          file changed there would go unnoticed.
+        - Go's build cache (`/root/.cache/go-build`, where Go keeps it for
+          root), mounted on the steps that compile: `golangci-lint run`, both
+          `go test` runs and `go build`. Go reuses a compiled package only when
+          all its inputs are unchanged. The cache also holds test results, which
+          is why both `go test` runs carry `-count=1`.
+        - golangci-lint's cache (`/root/.cache/golangci-lint`, where the pinned
+          image keeps it), mounted on `golangci-lint run`. golangci-lint reuses
+          a package's findings only when the package's files and those of every
+          package it imports, `go.mod`, the linters and their settings in
+          `.golangci.yml`, and the golangci-lint version are all unchanged.
     - If the project uses `//go:embed` directives that reference build artifacts
       (e.g. a web frontend compiled in a separate stage), the lint phase must
       create placeholder files so the embed directives resolve. Example:
@@ -247,12 +317,20 @@ style conventions are in separate documents:
       in the lint phase. The `golangci/golangci-lint` image is Debian-based and
       has no `apk`, so install with `apt-get` under the Debian package name
       (`libvips-dev`, where alpine says `vips-dev`), and delete the package
-      lists in the same `RUN`, so the layer does not keep them:
+      lists in the same `RUN`, so the layer does not keep them. The test phase,
+      on the Debian Go image, installs the libraries its tests need the same
+      way. Either install goes right after `FROM`, above `ARG CACHEBUST`, so
+      Docker keeps it until the base image or the install line changes:
 
         ```dockerfile
+        FROM golangci/golangci-lint@sha256:... AS lint
         RUN apt-get update \
             && apt-get install -y --no-install-recommends libvips-dev \
             && rm -rf /var/lib/apt/lists/*
+        WORKDIR /src
+        COPY go.mod go.sum ./
+        RUN go mod download
+        ARG CACHEBUST
         ```
 
     - `.dockerignore` lets `.git` into the build context. It keeps out every git
@@ -278,20 +356,39 @@ style conventions are in separate documents:
       `.git` and the version still comes out empty, `dev` or `unknown`. A plain
       `docker build .` with no build arguments must succeed; a Dockerfile that
       refuses an empty build argument drops that refusal and keeps the argument.
+      A checkout whose `.git` is a file (a linked worktree, or a repository
+      checked out as a submodule) is the exception: that file points to a git
+      directory outside the build context, so the build cannot read the version
+      and a plain `docker build .` fails; pass the version with
+      `--build-arg VERSION=...`, as `script/docker` and `script/cibuild` already
+      do.
 
 - Every repo should have a Gitea Actions workflow (`.gitea/workflows/`) that
-  runs `script/cibuild` on push, and checks out the repo as its only other step.
+  runs `script/cibuild` on push, and checks out the repo as its only other step,
-  That script bootstraps, runs the gate phases, and then builds the image, so a
+  with `persist-credentials: false`: `script/cibuild` needs no token, and
-  successful run means every check passed; a bare `docker build .` does not
+  without it the checkout leaves the job's token in `.git/config` for every
+  later step. The checkout step also sets `fetch-depth: 0`, which fetches the
+  tags `git describe` needs: by default it clones shallow with no tags, and a
+  tagged repository's CI build would stamp a bare short commit id. The
+  workflow's `concurrency` block groups runs by workflow and branch
+  (`${{ github.workflow }}-${{ github.ref }}`) with `cancel-in-progress: true`,
+  so a new push cancels the older run on the same branch, queued or running, and
+  no other: runs for replaced commits do not hold up the shared runner.
+  `script/cibuild` bootstraps, runs the gate phases, and then builds the image,
+  so a successful run means every check passed; a bare `docker build .` does not
   carry the same guarantee, because its gate phases may come from the cache. The
-  image build is uncached and so runs the gate phases a second time. That is the
+  image build gets a `CACHEBUST` value of its own and so runs the gate phases a
-  price of the rule above, and it is worth paying: the image that ships is built
+  second time. That is the price of the rule above, and it is worth paying: the
-  from a run of its own gates rather than from a cache entry. A separate
+  image that ships is built from a run of its own gates rather than from a cache
-  workflow limited to `main` by a `branches` list under `on: push` cannot be
+  entry. The `check` job sets `timeout-minutes: 20`, so a hung build frees the
-  checked by review: to try a change to it, add the feature branch to that list
+  shared runner after 20 minutes. That allows for the three Docker builds
-  and push, then remove the branch from the list again before merging. Keep any
+  described above (the test phase, the lint phase, then the image), each held to
-  job in it that publishes behind `if: github.ref_name == 'main'`, so the run
+  the 5-minute Docker build limit below, plus the bootstrap. A separate workflow
-  from the feature branch publishes nothing.
+  limited to `main` by a `branches` list under `on: push` cannot be checked by
+  review: to try a change to it, add the feature branch to that list and push,
+  then remove the branch from the list again before merging. Keep any job in it
+  that publishes behind `if: github.ref_name == 'main'`, so the run from the
+  feature branch publishes nothing.
 
 - Use platform-standard formatters: `black` for Python, `prettier` for
   JS/CSS/Markdown/HTML, `go fmt` for Go. Always use default configuration with
@@ -351,12 +448,11 @@ style conventions are in separate documents:
     That cache is Go's own, separate from Docker's layer cache. Go stores a
     passing result in its cache directory (`GOCACHE`), and when the same tests
     run again on unchanged code it prints that result, marked `(cached)`,
-    without running them. That matters on a developer's machine, where this
+    without running them. The directory lasts from one run to the next on a
-    target runs and the directory lasts from one run to the next. The `test`
+    developer's machine, where this target runs, and in the `test` phase of the
-    phase of the `Dockerfile` needs no `-count=1`: its base image holds no
+    `Dockerfile`, which mounts it to keep compiled packages between check
-    result for this repo's tests and nothing before its `go test` step runs a
+    builds, so the phase's `go test` runs carry `-count=1` too. A new
-    test, so there is nothing to replay. `--no-cache` (above) is what makes that
+    `CACHEBUST` value (above) is what makes that step run on an unchanged tree.
-    step run on an unchanged tree.
 
     Python example:
 
@@ -384,9 +480,12 @@ style conventions are in separate documents:
 
 - `.gitignore` should be comprehensive from the start: OS files (`.DS_Store`),
   editor files (`.swp`, `*~`), in-repo agent scratch directories (`.claude/`),
-  language build artifacts, and `node_modules/`. Fetch the standard `.gitignore`
+  `node_modules/`, and the repo's own build outputs. Fetch the standard
-  from `https://git.eeqj.de/sneak/prompts/raw/branch/main/.gitignore` when
+  `.gitignore` from
-  setting up a new repo. These patterns are written to `.gitignore`'s own
+  `https://git.eeqj.de/sneak/prompts/raw/branch/main/.gitignore` when setting up
+  a new repo. A repo's `.gitignore` is the standard file followed by the repo's
+  own entries, such as its binaries; a re-vendor replaces the standard part and
+  keeps those entries. These patterns are written to `.gitignore`'s own
   semantics, in which an unanchored pattern already matches at every depth; they
   are not a `.dockerignore` and must not be transplanted into one unmodified.
 
@@ -434,13 +533,15 @@ style conventions are in separate documents:
   byte-identically across repos:
 
     ```sh
-    # Own line: a failing command substitution inside an argument does not
+    # The version and the tag each get their own line: a failing command
-    # trip `set -e`, so the inline form degrades to an empty constant.
+    # substitution inside an argument does not trip `set -e`, so the inline
+    # form degrades to an empty constant.
     version="$(git describe --tags --always --dirty 2>/dev/null || true)"
     [ -n "$version" ] || version="unknown"
-    docker build --no-cache \
+    tag="$(script/projectname)"
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
         --build-arg VERSION="$version" \
-        -t "$(script/projectname)" .
+        -t "$tag" .
     ```
 
     `--always` makes an untagged repo yield an abbreviated commit hash rather
@@ -452,8 +553,8 @@ style conventions are in separate documents:
     there because `ARG` is stage-scoped; passing `VERSION` to a repo whose
     Dockerfile declares no such `ARG` is ignored and costs nothing, which is why
     the scripts stay byte-identical. One consequence for CI: the standard
-    checkout action clones shallow and fetches no tags, so a repo that embeds a
+    checkout action clones shallow and fetches no tags, so the canonical
-    tag-derived version must set `fetch-depth: 0` on its checkout step.
+    `.gitea/workflows/check.yml` sets `fetch-depth: 0` on its checkout step.
 
 - **Verify `.dockerignore` by enumerating the image, not by reading the
   patterns.** Plant files at the root _and_ at least two directories deep, build
@@ -520,6 +621,11 @@ style conventions are in separate documents:
       function defined and never invoked has the same exit status and the same
       empty output as one that worked.
 
+    Node is the exception to the whole-token comparison: the canonical
+    `script/bootstrap` compares only its major version, because the node 22
+    alpine image the `Dockerfile` stages start from is not the exact pin, and
+    nvm cannot install a prebuilt node on alpine.
+
     Keep it POSIX sh: no arrays, no `[[`, no `grep -P`.
 
     A Go tool a repo needs on the host is installed with `go install` pinned to
@@ -636,7 +742,11 @@ style conventions are in separate documents:
       Never edit existing migrations after release.
 
 - All repos should have an `.editorconfig` enforcing the project's indentation
-  settings.
+  settings: the standard file from
+  `https://git.eeqj.de/sneak/prompts/raw/branch/main/.editorconfig`, which sets
+  tabs for `Makefile` and Go files, followed by the repo's own sections, such as
+  one for another language it uses. A re-vendor replaces the standard part and
+  keeps those sections.
 
 - Avoid putting files in the repo root unless necessary. Root should contain
   only project-level config files (`README.md`, `AGENTS.md`, `Makefile`,

TODO.md

@@ -9,10 +9,12 @@
 
 # Status
 
-pre-1.0. No git tags. `main` is the stable branch and `next` the development
+pre-1.0. Release candidates are tagged on `next`, the first being `v1.0.0-rc1`;
-branch, which every PR targets. The frontend and the Go backend ship as one
+`git tag --merged origin/next` lists them. `main` is the stable branch and
-Docker image, and the Gitea workflow `.gitea/workflows/check.yml` runs
+`next` the development branch, which every PR targets. The frontend and the Go
-`script/cibuild` on every push. Working toward 1.0.0.
+backend ship as one Docker image, and the Gitea workflow
+`.gitea/workflows/check.yml` runs `script/cibuild` on every push, and cancels
+the older run on the same branch. Working toward 1.0.0.
 
 # Next Step
 
@@ -21,18 +23,181 @@ Run `make frontend-viewport-test` in CI as its own step; it is not part of
 
 # Completed Steps
 
+- 2026-10-19: the server's `OnStart` hook in `internal/server/server.go` starts
+  `listenAndServe` with `go func() { s.listenAndServe() }()`, as the Go style
+  guide asks ([#162](https://git.eeqj.de/sneak/netwatch/issues/162)); it was
+  `go s.listenAndServe()`. No other `go` statement in the Go code called a
+  function directly. Nothing the server does changed
+- 2026-10-19: `.gitignore`, `.editorconfig` and `.golangci.yml` are the
+  sneak/prompts copies at `0653579`
+  ([#174](https://git.eeqj.de/sneak/netwatch/issues/174)). `.gitignore` is the
+  standard file followed by this repository's own entries; `.editorconfig` is
+  the standard file alone, which now carries the `[*.go]` section.
+  `.golangci.yml` disables `canonicalheader`, as the canonical copy does,
+  because golangci-lint v2.14.0 misses its findings at random
+- 2026-10-19: `package.json` has no `version` field and is marked
+  `"private": true` ([#158](https://git.eeqj.de/sneak/netwatch/issues/158)). It
+  said `1.0.0`, a release that has not been made, and nothing read it; the git
+  tag is the one version, which the image and `make build` take from
+  `git describe --tags`. The package is never published
+- 2026-10-19: Status names `v1.0.0-rc1`, the first release candidate, and says
+  release candidates are tagged on `next`, where it said there were no git tags
+  ([#157](https://git.eeqj.de/sneak/netwatch/issues/157)). The rest of Status
+  matches the tree
+- 2026-10-19: `netwatch-server` starts the server only when given no arguments
+  ([#179](https://git.eeqj.de/sneak/netwatch/issues/179)). `check-cidr CIDR` and
+  `prepare-data-dir DATA_DIR` behave as before; any other arguments, such as a
+  subcommand with no value or two, or a misspelt one, print a one-line usage
+  message naming both forms to stderr and exit 2. They used to start the server.
+  Unit tests call `cli.Main` with no arguments and check that it starts the
+  server, and with each subcommand given no value and two values and with a
+  misspelt subcommand, and check the exit code, the message and that the server
+  does not start
+- 2026-10-19: `main()` in `cmd/netwatch-server/main.go` is one call into a new
+  package, `internal/cli`, as the Go style guide asks
+  ([#160](https://git.eeqj.de/sneak/netwatch/issues/160)). `internal/cli` runs
+  the `check-cidr` and `prepare-data-dir` subcommands, with the same arguments,
+  messages and exit codes, and otherwise the server, with the `fx` wiring.
+  `Appname` and `Version` stay in `main`, where the build sets `Version`, and
+  `main` gives `cli.Main` stderr and the function that runs the server, which
+  the tests replace with a buffer and a function that fails them. Unit tests
+  call `cli.Main` with `check-cidr` and a CIDR the server takes and one it
+  refuses, and check the exit code and what it writes to stderr, and with
+  `prepare-data-dir` and a path outside `/data`, and check that it exits 1 with
+  the error from looking up the `netwatch` user, which the test image does not
+  have. Others give the functions behind the subcommands input they take and
+  input they refuse, and check the exit code and what they write to stderr
+- 2026-10-19: when `VERSION` is unset or empty, `script/build` stamps the
+  version from `git describe --tags --always --dirty`, as `script/docker` does
+  ([#156](https://git.eeqj.de/sneak/netwatch/issues/156)). Without `--tags` it
+  skipped lightweight tags, so `make run` could report a commit where the image
+  built from the same tree reported the tag. Checked once with a lightweight tag
+  on a scratch clone: `make build` stamped it
+- 2026-10-19: `.dockerignore` keeps Go test binaries and coverage output at the
+  root out of the build context, with `/*.test` and `/*.out` in the repository's
+  own section ([#155](https://git.eeqj.de/sneak/netwatch/issues/155)). Such
+  files went into the build context of every `docker build` in `script/` and
+  were copied into the lint, test and build stages. Checked with a probe image
+  that copies the whole context: `x.test` and `x.out` planted at the root no
+  longer land in it, and nothing else changed. The entries match only at the
+  root, so the same files in a subdirectory still reach the image
+- 2026-10-19: `.gitea/workflows/check.yml` is the sneak/prompts copy at
+  `0653579` ([#153](https://git.eeqj.de/sneak/netwatch/issues/153)). Its
+  checkout step sets `fetch-depth: 0`, so CI clones all history and tags, and
+  `git describe` in `script/cibuild` gives the image built from a tagged commit
+  the tag as its version; it gave the short commit. The checkout step also sets
+  `persist-credentials: false`, so the job's token is not left in `.git/config`.
+  A new push cancels the older run on the same branch, and a run stops after 20
+  minutes
+- 2026-10-19: `script/bootstrap` keeps the `yarn` on `PATH` only when
+  `yarn --version` is exactly `YARN_VERSION`, and otherwise installs that
+  version as it did when no yarn was installed
+  ([#154](https://git.eeqj.de/sneak/netwatch/issues/154)); any yarn used to be
+  kept. It uses the `node` on `PATH` only when that node has the major version
+  of `NODE_VERSION`, as `REPO_POLICIES.md` asks, and is at least
+  `NODE_MIN_VERSION`; a node 24 was kept before. After installing node or yarn
+  it looks the tool up through `PATH` again and stops if that one fails the same
+  check, as it does for Go, and it prints the node and yarn versions it uses.
+  Every script that runs yarn on the host, except `script/bootstrap`, which
+  picks its own node and yarn, runs it through the new `script/yarn`. That
+  script puts `~/.local/bin` first on `PATH`, uses the first `node` on that
+  `PATH` when it passes the same test, and otherwise runs yarn under the pinned
+  node in nvm; `make build` ran yarn under whatever node was first on `PATH`
+- 2026-10-19: `bin/entrypoint.sh` has each of its steps in a function, and a
+  `main` at the end calls them in order, as the shell section of the code style
+  guide asks ([#163](https://git.eeqj.de/sneak/netwatch/issues/163)). It is
+  still POSIX sh with no `set -e`, and does what it did, with the same messages
+  and exit codes. Checked by running the image with each script: a normal start,
+  bad `PORT`, `TRUSTED_PROXIES` and `DATA_DIR` values, `docker stop`, INT, and
+  nginx exiting on its own
+- 2026-10-19: while report files cannot be written, the health check answers 503
+  with `"status":"error"` and the same fields, so the image's `HEALTHCHECK`
+  reports the container `unhealthy`
+  ([#142](https://git.eeqj.de/sneak/netwatch/issues/142)). It answered 200
+  `"status":"ok"` whatever happened, so the platform saw a healthy container
+  while every report was refused with 500. It answers 200 at start, before any
+  write has failed, and again from the first write that succeeds. A unit test
+  removes `DATA_DIR` to make a write fail and checks the 503, then puts it back
+  and checks the 200 after the next write
+- 2026-10-19: report file writes no longer overlap
+  ([#150](https://git.eeqj.de/sneak/netwatch/issues/150)), so the reports of
+  failed writes are written in the order they arrived. Two writes that failed at
+  the same time put their reports back in the wrong order when the older failed
+  first. One write runs at a time, and each takes the waiting reports only when
+  it starts, after a failed write has put its own back. The write `Append`
+  starts when 10 MiB are waiting waits for any write under way, and `Append`
+  starts no other while it waits. Files written since the start are deleted to
+  make room in the order they were written, after those found at start. A unit
+  test holds a write open, starts a second at 10 MiB, fails both and checks the
+  order in the next file. Others check that `Append` starts no second write at
+  10 MiB while one waits, that it starts one again the next time 10 MiB are
+  waiting, and that the first file written since the start is the first deleted
+- 2026-10-19: `DEBUG` unset or empty leaves debug logging off, and any other
+  value, `false` and `0` included, turns it on, as the code style guide says
+  ([#159](https://git.eeqj.de/sneak/netwatch/issues/159)). No value of it stops
+  the start any more; before, only `1`, `t`, `T`, `TRUE`, `true`, `True`, `0`,
+  `f`, `F`, `FALSE`, `false` and `False` were accepted. `README.md` no longer
+  gives `false` as its default
+- 2026-10-19: check builds keep Go's modules and compiled packages between runs
+  ([#118](https://git.eeqj.de/sneak/netwatch/issues/118)), on sneak's rulings in
+  [prompts#113](https://git.eeqj.de/sneak/prompts/issues/113) and
+  [prompts#130](https://git.eeqj.de/sneak/prompts/issues/130).
+  `REPO_POLICIES.md`, `script/test`, `script/lint`, `script/docker` and
+  `script/cibuild` are the sneak/prompts copies at `0653579`: the scripts pass a
+  new `CACHEBUST` value where they passed `--no-cache`, and the lint and test
+  builds write no image. In each stage of `Dockerfile`, `ARG CACHEBUST` comes
+  after the system packages and the Go module download, and before every step
+  that lints, tests or compiles and the node stages' `yarn install`. Go's build
+  cache and golangci-lint's are cache mounts, and both `go test` runs carry
+  `-count=1`. The test phase's Go caches used to be in memory, empty on every
+  build
+- 2026-10-19: the three unit tests of a check whose result is discarded (its
+  round given up, the user paused, or the first round) also check that the
+  host's history is still empty, its status still `pending`, and nothing was
+  written to the console
+  ([#151](https://git.eeqj.de/sneak/netwatch/issues/151)). They checked only the
+  row, so a change that recorded the result and skipped only the row would have
+  kept them passing
+- 2026-10-18: when the backend runs on its own, a `DATA_DIR` that cannot be
+  created or read, or where no file can be created, stops the start with an
+  error naming `DATA_DIR` and its value, as the other settings' errors do
+  ([#148](https://git.eeqj.de/sneak/netwatch/issues/148)); the first two named
+  only the path, and the third was found only at the first write of a report
+  file. At start the server creates a file in `DATA_DIR` and removes it. Unit
+  tests start the buffer with `DATA_DIR` below a regular file and with
+  `DATA_DIR` set to `/proc`, and check the error. The report file number is an
+  `atomic.Int64`, as the Go style guide asks for int types
+  ([#161](https://git.eeqj.de/sneak/netwatch/issues/161)); file names are
+  unchanged
 - 2026-10-18: the image's nginx sends the page with `Cache-Control: no-cache`
   ([#141](https://git.eeqj.de/sneak/netwatch/issues/141)), so a browser that
   keeps a copy checks with the server that it is current before using it again.
   The page went out with no `Cache-Control`, so a browser could keep it after a
   deploy and ask for files under `/assets/` that the new build no longer has,
-  and show an empty page. A path with no file of its own falls back to the page
+  and show an empty page. A path that `location /` answers with the page through
-  and gets the same header. `location /` now includes the security headers
+  its `try_files` fallback gets the same header. `location /` now includes the
-  again, as `/assets/` does, since its own `add_header` drops the server's.
+  security headers again, as `/assets/` does, since its own `add_header` drops
-  `/assets/` keeps its immutable caching. `make frontend-viewport-test` fails
+  the server's. `/assets/` keeps its immutable caching.
-  unless `/` and a path that falls back to the page carry
+  `make frontend-viewport-test` fails unless `/` and a path that falls back to
-  `Cache-Control: no-cache` and `Content-Security-Policy`, which comes from the
+  the page carry `Cache-Control: no-cache` and `Content-Security-Policy`, which
-  same file as the other security headers
+  comes from the same file as the other security headers
+- 2026-10-18: `README.md` says what the page loses while reports are refused or
+  the page is paused ([#145](https://git.eeqj.de/sneak/netwatch/issues/145)),
+  how fast one address can replace the stored reports
+  ([#146](https://git.eeqj.de/sneak/netwatch/issues/146)), and that in the image
+  nginx answers a body over 1 MiB itself
+  ([#147](https://git.eeqj.de/sneak/netwatch/issues/147)). A sample in the
+  history that no report has delivered is sent only as long as the history holds
+  it: at the 1-second interval one refused or failed report loses the oldest 20
+  seconds of each host's samples, and a pause longer than the history loses
+  every sample between the last delivered report and the pause. Samples an
+  interval change cleared are kept apart, the last 100 per host, until they are
+  delivered. At the defaults, one address sending 1 MiB reports made up to
+  compress badly can fill the 1 GiB cap in about 17 minutes. nginx's
+  `client_max_body_size` is also 1 MiB by default, so in the image it answers a
+  larger body with 413 and its own HTML page, and the backend never sees or
+  records it; the backend's own 413 sentences now say they apply to it run on
+  its own. The reporter, the body limit and nginx are unchanged
 - 2026-10-18: a browser that cannot format the header's clocks still probes
   ([#149](https://git.eeqj.de/sneak/netwatch/issues/149)). The clocks name the
   local time zone through `Intl.DateTimeFormat`'s `shortGeneric` option, which
@@ -586,9 +751,3 @@ Run `make frontend-viewport-test` in CI as its own step; it is not part of
 - A configurable host list (an environment variable or a config file)
 - Export of the latency history (CSV or JSON)
 - A notification when the health status changes to DEGRADED
-- `make test` within the 20-second target
-  ([#118](https://git.eeqj.de/sneak/netwatch/issues/118)): every run compiles
-  the Go standard library and every dependency with the race detector from
-  nothing. Whether test builds may keep Go's compiled packages between runs
-  waits on sneak's ruling in
-  [prompts#113](https://git.eeqj.de/sneak/prompts/issues/113)

bin/entrypoint.sh

@@ -13,24 +13,27 @@ set -u
 # address and start anyway, and reports a bad port without naming
 # PORT, so a value that is not a usable port stops the container here,
 # before either process starts.
-export PORT="${PORT:-8080}"
+check_port() {
-case "$PORT" in
+    export PORT="${PORT:-8080}"
+    case "$PORT" in
         *[!0-9]*)
             echo "entrypoint: PORT must be a port number, not '$PORT'" >&2
             exit 1
             ;;
-esac
+    esac
-# The length is checked first because, for a number too big for it,
+    # The length is checked first because, for a number too big for it,
-# the shell's test prints an error and is false, so the range checks
+    # the shell's test prints an error and is false, so the range checks
-# alone would let it through.
+    # alone would let it through.
-if [ "${#PORT}" -gt 5 ] || [ "$PORT" -lt 1 ] || [ "$PORT" -gt 65535 ]; then
+    if [ "${#PORT}" -gt 5 ] || [ "$PORT" -lt 1 ] || [ "$PORT" -gt 65535 ]; then
         echo "entrypoint: PORT must be from 1 to 65535, not '$PORT'" >&2
         exit 1
-fi
+    fi
-if [ "$PORT" -eq 8081 ]; then
+    if [ "$PORT" -eq 8081 ]; then
-    echo "entrypoint: PORT cannot be 8081, netwatch-server listens there" >&2
+        echo "entrypoint: PORT cannot be 8081, netwatch-server" \
+            "listens there" >&2
         exit 1
-fi
+    fi
+}
 
 # TRUSTED_PROXIES names the reverse proxies in front of the container,
 # as IP addresses or CIDRs separated by commas. nginx takes the client
@@ -47,9 +50,10 @@ fi
 # TRUSTED_PROXIES. Its error, naming the CIDR, is dropped for the one
 # below, naming the entry as written. set -f keeps a * in an entry from
 # becoming a list of file names.
-TRUSTED_PROXIES="${TRUSTED_PROXIES-10.0.0.0/8,172.16.0.0/12,192.168.0.0/16}"
+write_trusted_proxies() {
-set -f
+    TRUSTED_PROXIES="${TRUSTED_PROXIES-10.0.0.0/8,172.16.0.0/12,192.168.0.0/16}"
-for proxy in $(printf '%s' "$TRUSTED_PROXIES" | tr ',' ' '); do
+    set -f
+    for proxy in $(printf '%s' "$TRUSTED_PROXIES" | tr ',' ' '); do
         case "$proxy" in
             */*) cidr="$proxy" ;;
             *:*) cidr="$proxy/128" ;;
@@ -61,7 +65,8 @@ for proxy in $(printf '%s' "$TRUSTED_PROXIES" | tr ',' ' '); do
             exit 1
         fi
         echo "set_real_ip_from $cidr;"
-done > /etc/nginx/trusted-proxies.conf
+    done > /etc/nginx/trusted-proxies.conf
+}
 
 # netwatch-server keeps its report files in DATA_DIR, on the /data
 # volume, which may be a host directory owned by root or by another
@@ -73,12 +78,10 @@ done > /etc/nginx/trusted-proxies.conf
 # naming DATA_DIR, unless DATA_DIR is /data or a path below it, and it
 # acts on nothing outside /data, whatever symbolic or hard links it
 # meets there.
-export DATA_DIR="${DATA_DIR:-/data/reports}"
+prepare_data_dir() {
-netwatch-server prepare-data-dir "$DATA_DIR" || exit 1
+    export DATA_DIR="${DATA_DIR:-/data/reports}"
-
+    netwatch-server prepare-data-dir "$DATA_DIR" || exit 1
-# A stop signal is only noted here; the loop below acts on it.
+}
-stop_requested=""
-trap 'stop_requested=yes' TERM INT
 
 # netwatch-server runs as the netwatch user and listens on loopback
 # only, on a port other than the public one; configs/nginx.conf proxies
@@ -87,9 +90,11 @@ trap 'stop_requested=yes' TERM INT
 # the container has. The netwatch user has no login shell, hence -s
 # /bin/sh. busybox su replaces itself with the command instead of
 # staying on as its parent, so $! is the server's own PID.
-BIND_ADDRESS=127.0.0.1 PORT=8081 TRUSTED_PROXIES=127.0.0.1/32 \
+start_backend() {
+    BIND_ADDRESS=127.0.0.1 PORT=8081 TRUSTED_PROXIES=127.0.0.1/32 \
         su -s /bin/sh netwatch -c 'exec netwatch-server' &
-backend=$!
+    backend=$!
+}
 
 # nginx starts through the nginx image's own entrypoint, which applies
 # the image's start-up configuration and then replaces itself with
@@ -98,9 +103,11 @@ backend=$!
 # that rendering to PORT: a variable nginx itself uses, such as $uri,
 # would otherwise be replaced by an environment variable of the same
 # name.
-NGINX_ENVSUBST_FILTER='^PORT$' \
+start_nginx() {
+    NGINX_ENVSUBST_FILTER='^PORT$' \
         /docker-entrypoint.sh nginx -g 'daemon off;' &
-nginx=$!
+    nginx=$!
+}
 
 running() {
     kill -0 "$1" 2>/dev/null
@@ -109,26 +116,47 @@ running() {
 # POSIX sh cannot wait for whichever of two children exits first, so
 # look once a second. The shell collects a child that has exited while
 # it runs sleep, and running() is false for that child from then on.
-while [ -z "$stop_requested" ] && running "$backend" && running "$nginx"; do
+wait_for_stop_or_exit() {
+    while [ -z "$stop_requested" ] && running "$backend" && running "$nginx"; do
         sleep 1
-done
+    done
+}
 
 # Stop both, then wait until neither is left.
-kill -TERM "$backend" "$nginx" 2>/dev/null
+stop_both() {
-while running "$backend" || running "$nginx"; do
+    kill -TERM "$backend" "$nginx" 2>/dev/null
+    while running "$backend" || running "$nginx"; do
         sleep 1
-done
+    done
-
+}
-wait "$backend"
-backend_status=$?
-wait "$nginx"
-nginx_status=$?
-echo "entrypoint: netwatch-server exited $backend_status," \
-    "nginx exited $nginx_status"
 
 # Success is a requested stop that both processes exited cleanly from.
-if [ -n "$stop_requested" ] && [ "$backend_status" -eq 0 ] &&
+exit_with_status() {
+    wait "$backend"
+    backend_status=$?
+    wait "$nginx"
+    nginx_status=$?
+    echo "entrypoint: netwatch-server exited $backend_status," \
+        "nginx exited $nginx_status"
+    if [ -n "$stop_requested" ] && [ "$backend_status" -eq 0 ] &&
         [ "$nginx_status" -eq 0 ]; then
         exit 0
-fi
+    fi
-exit 1
+    exit 1
+}
+
+main() {
+    check_port
+    write_trusted_proxies
+    prepare_data_dir
+    # A stop signal is only noted here; wait_for_stop_or_exit acts on it.
+    stop_requested=""
+    trap 'stop_requested=yes' TERM INT
+    start_backend
+    start_nginx
+    wait_for_stop_or_exit
+    stop_both
+    exit_with_status
+}
+
+main "$@"

cmd/netwatch-server/main.go

@@ -1,22 +1,11 @@
-// Package main is the entry point for netwatch-server.
+// Package main is the entry point for netwatch-server; package cli
+// does the work.
 package main
 
 import (
-	"fmt"
 	"os"
-	"os/user"
 
-	"sneak.berlin/go/netwatch/internal/config"
+	"sneak.berlin/go/netwatch/internal/cli"
-	"sneak.berlin/go/netwatch/internal/globals"
-	"sneak.berlin/go/netwatch/internal/handlers"
-	"sneak.berlin/go/netwatch/internal/healthcheck"
-	"sneak.berlin/go/netwatch/internal/logger"
-	"sneak.berlin/go/netwatch/internal/middleware"
-	"sneak.berlin/go/netwatch/internal/reportbuf"
-	"sneak.berlin/go/netwatch/internal/server"
-
-	"go.uber.org/fx"
-	"go.uber.org/fx/fxevent"
 )
 
 //nolint:gochecknoglobals // set via ldflags at build time
@@ -26,64 +15,5 @@ var (
 )
 
 func main() {
-	// "netwatch-server check-cidr CIDR" exits 1, with the error, if
+	os.Exit(cli.Main(os.Stderr, Appname, Version, os.Args[1:], cli.RunServer))
-	// this server would refuse CIDR in its TRUSTED_PROXIES.
-	// bin/entrypoint.sh runs it on each entry it gives nginx.
-	if len(os.Args) == 3 && os.Args[1] == "check-cidr" {
-		_, err := middleware.ParseTrustedProxies(os.Args[2:])
-		if err != nil {
-			fmt.Fprintln(os.Stderr, err)
-			os.Exit(1)
-		}
-
-		return
-	}
-
-	// "netwatch-server prepare-data-dir DATA_DIR" gets DATA_DIR ready
-	// for the netwatch user, or exits 1 with the error; see
-	// reportbuf.PrepareDataDir. bin/entrypoint.sh runs it as root
-	// before it starts this server as that user.
-	if len(os.Args) == 3 && os.Args[1] == "prepare-data-dir" {
-		netwatch, err := user.Lookup("netwatch")
-		if err != nil {
-			fmt.Fprintln(os.Stderr, err)
-			os.Exit(1)
-		}
-
-		err = reportbuf.PrepareDataDir("/data", os.Args[2], netwatch)
-		if err != nil {
-			fmt.Fprintf(os.Stderr, "DATA_DIR '%s': %v\n", os.Args[2], err)
-			os.Exit(1)
-		}
-
-		return
-	}
-
-	globals.Appname = Appname
-	globals.Version = Version
-
-	fx.New(
-		// fx logs each step of starting and stopping through the
-		// server's own logger, so off a terminal those lines are
-		// JSON like every other line.
-		fx.WithLogger(func(log *logger.Logger) fxevent.Logger {
-			return &fxevent.SlogLogger{Logger: log.Get()}
-		}),
-		fx.Provide(
-			config.New,
-			globals.New,
-			handlers.New,
-			healthcheck.New,
-			logger.New,
-			middleware.New,
-			reportbuf.New,
-			server.New,
-		),
-		fx.Invoke(
-			// First, so the name and version are logged even when
-			// a setting stops the start.
-			func(log *logger.Logger) { log.Identify() },
-			func(*server.Server) {},
-		),
-	).Run()
 }

internal/cli/cli.go

@@ -0,0 +1,143 @@
+// Package cli runs netwatch-server from its command line: the
+// check-cidr and prepare-data-dir subcommands, which bin/entrypoint.sh
+// runs before it starts the server, and, given no arguments, the
+// server, with the fx wiring.
+package cli
+
+import (
+	"fmt"
+	"io"
+	"os/user"
+
+	"sneak.berlin/go/netwatch/internal/config"
+	"sneak.berlin/go/netwatch/internal/globals"
+	"sneak.berlin/go/netwatch/internal/handlers"
+	"sneak.berlin/go/netwatch/internal/healthcheck"
+	"sneak.berlin/go/netwatch/internal/logger"
+	"sneak.berlin/go/netwatch/internal/middleware"
+	"sneak.berlin/go/netwatch/internal/reportbuf"
+	"sneak.berlin/go/netwatch/internal/server"
+
+	"go.uber.org/fx"
+	"go.uber.org/fx/fxevent"
+)
+
+// usageExitCode is what Main returns for a command line it does not
+// understand, the usual exit code for a usage error.
+const usageExitCode = 2
+
+// Main runs netwatch-server with args, its command line after the
+// program's name, and returns the exit code. With no args, Main runs
+// the server by calling serve with appname and version, the name and
+// version the server logs and reports. With "check-cidr CIDR" or
+// "prepare-data-dir DATA_DIR" it runs that subcommand, which writes its
+// errors to stderr. Any other args get a one-line usage message on
+// stderr and exit code 2. main passes os.Stderr and RunServer; the
+// tests pass a buffer and a function of their own, so that no test
+// starts the server. When serve is RunServer and the server fails to
+// start, to listen or to stop, fx's Run exits the process with status 1
+// and Main never returns.
+func Main(
+	stderr io.Writer,
+	appname, version string,
+	args []string,
+	serve func(appname, version string),
+) int {
+	// "netwatch-server check-cidr CIDR" exits 1, with the error, if
+	// this server would refuse CIDR in its TRUSTED_PROXIES.
+	// bin/entrypoint.sh runs it on each entry it gives nginx.
+	if len(args) == 2 && args[0] == "check-cidr" {
+		return checkCIDR(stderr, args[1])
+	}
+
+	// "netwatch-server prepare-data-dir DATA_DIR" gets DATA_DIR ready
+	// for the netwatch user, or exits 1 with the error; see
+	// reportbuf.PrepareDataDir. bin/entrypoint.sh runs it as root
+	// before it starts this server as that user.
+	if len(args) == 2 && args[0] == "prepare-data-dir" {
+		return prepareDataDir(stderr, "/data", args[1], "netwatch")
+	}
+
+	// Any other command line is refused, so that a mistyped one does
+	// not start the server.
+	if len(args) > 0 {
+		_, _ = fmt.Fprintln(stderr,
+			"usage: netwatch-server [check-cidr CIDR | prepare-data-dir DATA_DIR]")
+
+		return usageExitCode
+	}
+
+	serve(appname, version)
+
+	return 0
+}
+
+// checkCIDR returns 0 if the server would take cidr in its
+// TRUSTED_PROXIES, and otherwise writes the error to stderr and
+// returns 1.
+func checkCIDR(stderr io.Writer, cidr string) int {
+	_, err := middleware.ParseTrustedProxies([]string{cidr})
+	if err != nil {
+		_, _ = fmt.Fprintln(stderr, err)
+
+		return 1
+	}
+
+	return 0
+}
+
+// prepareDataDir gets dataDir ready for the user named username with
+// reportbuf.PrepareDataDir, volume being where the data volume is
+// mounted, and returns 0. Otherwise it writes the error to stderr and
+// returns 1.
+func prepareDataDir(stderr io.Writer, volume, dataDir, username string) int {
+	owner, err := user.Lookup(username)
+	if err != nil {
+		_, _ = fmt.Fprintln(stderr, err)
+
+		return 1
+	}
+
+	err = reportbuf.PrepareDataDir(volume, dataDir, owner)
+	if err != nil {
+		_, _ = fmt.Fprintf(stderr, "DATA_DIR '%s': %v\n", dataDir, err)
+
+		return 1
+	}
+
+	return 0
+}
+
+// RunServer runs the server until it is told to stop, with appname and
+// version as the name and version it logs and reports. If the server
+// fails to start, to listen or to stop, fx's Run exits the process
+// with status 1 itself.
+func RunServer(appname, version string) {
+	globals.Appname = appname
+	globals.Version = version
+
+	fx.New(
+		// fx logs each step of starting and stopping through the
+		// server's own logger, so off a terminal those lines are
+		// JSON like every other line.
+		fx.WithLogger(func(log *logger.Logger) fxevent.Logger {
+			return &fxevent.SlogLogger{Logger: log.Get()}
+		}),
+		fx.Provide(
+			config.New,
+			globals.New,
+			handlers.New,
+			healthcheck.New,
+			logger.New,
+			middleware.New,
+			reportbuf.New,
+			server.New,
+		),
+		fx.Invoke(
+			// First, so the name and version are logged even when
+			// a setting stops the start.
+			func(log *logger.Logger) { log.Identify() },
+			func(*server.Server) {},
+		),
+	).Run()
+}

internal/cli/cli_test.go

@@ -0,0 +1,219 @@
+package cli_test
+
+import (
+	"bytes"
+	"os"
+	"os/user"
+	"path/filepath"
+	"strconv"
+	"strings"
+	"testing"
+
+	"sneak.berlin/go/netwatch/internal/cli"
+	"sneak.berlin/go/netwatch/internal/reportbuf"
+)
+
+// TestMainRunsCheckCIDR: Main given check-cidr returns 0, writing
+// nothing, for a CIDR the server takes in its TRUSTED_PROXIES, and 1,
+// with an error naming TRUSTED_PROXIES and the CIDR, for one it
+// refuses, and does not start the server.
+func TestMainRunsCheckCIDR(t *testing.T) {
+	t.Parallel()
+
+	var stderr bytes.Buffer
+
+	code := cli.Main(&stderr, "netwatch", "test",
+		[]string{"check-cidr", "10.0.0.0/8"}, failIfServerStarts(t))
+	if code != 0 || stderr.Len() != 0 {
+		t.Errorf("check-cidr 10.0.0.0/8: exit %d, stderr %q; want exit 0 and no output",
+			code, stderr.String())
+	}
+
+	stderr.Reset()
+
+	code = cli.Main(&stderr, "netwatch", "test",
+		[]string{"check-cidr", "1.2.3/32"}, failIfServerStarts(t))
+
+	want := `TRUSTED_PROXIES "1.2.3/32": `
+	if code != 1 || !strings.HasPrefix(stderr.String(), want) {
+		t.Errorf("check-cidr 1.2.3/32: exit %d, stderr %q; want exit 1 and %q...",
+			code, stderr.String(), want)
+	}
+}
+
+// TestMainRunsPrepareDataDir: Main given prepare-data-dir runs the
+// prepare-data-dir subcommand for the netwatch user, and does not start
+// the server. The test image has no netwatch user, so that subcommand
+// exits 1 with the error from looking the user up;
+// TestPrepareDataDirRefusesPathOutsideVolume checks the path.
+func TestMainRunsPrepareDataDir(t *testing.T) {
+	t.Parallel()
+
+	var stderr bytes.Buffer
+
+	code := cli.Main(&stderr, "netwatch", "test",
+		[]string{"prepare-data-dir", "/etc"}, failIfServerStarts(t))
+
+	want := user.UnknownUserError("netwatch").Error() + "\n"
+	if code != 1 || stderr.String() != want {
+		t.Errorf("prepare-data-dir /etc: exit %d, stderr %q; want exit 1 and %q",
+			code, stderr.String(), want)
+	}
+}
+
+// TestMainRunsServerWithNoArguments: Main given no arguments starts the
+// server with the name and version it was given, writes nothing and
+// returns 0.
+func TestMainRunsServerWithNoArguments(t *testing.T) {
+	t.Parallel()
+
+	var (
+		stderr                 bytes.Buffer
+		gotAppname, gotVersion string
+	)
+
+	code := cli.Main(&stderr, "netwatch", "test", []string{},
+		func(appname, version string) {
+			gotAppname, gotVersion = appname, version
+		})
+	if code != 0 || stderr.Len() != 0 {
+		t.Errorf("exit %d, stderr %q; want exit 0 and no output",
+			code, stderr.String())
+	}
+
+	if gotAppname != "netwatch" || gotVersion != "test" {
+		t.Errorf("server started with %q %q; want %q %q",
+			gotAppname, gotVersion, "netwatch", "test")
+	}
+}
+
+// TestMainRefusesOtherCommandLines: Main given a subcommand with too
+// few or too many values, or a misspelt subcommand, writes the usage
+// message to stderr, returns 2 and does not start the server.
+func TestMainRefusesOtherCommandLines(t *testing.T) {
+	t.Parallel()
+
+	want := "usage: netwatch-server " +
+		"[check-cidr CIDR | prepare-data-dir DATA_DIR]\n"
+
+	for _, commandLine := range []string{
+		"netwatch-server check-cidr",
+		"netwatch-server check-cidr 192.0.2.0/24 198.51.100.0/24",
+		"netwatch-server prepare-data-dir",
+		"netwatch-server prepare-data-dir /data/reports /data/other",
+		"netwatch-server check-cdir 192.0.2.0/24",
+	} {
+		var stderr bytes.Buffer
+
+		// Main takes the command line after the program's name.
+		args := strings.Fields(commandLine)[1:]
+
+		code := cli.Main(&stderr, "netwatch", "test", args, failIfServerStarts(t))
+		if code != 2 || stderr.String() != want {
+			t.Errorf("%q: exit %d, stderr %q; want exit 2 and %q",
+				commandLine, code, stderr.String(), want)
+		}
+	}
+}
+
+// TestCheckCIDRTakesCIDR: check-cidr exits 0, writing nothing, for a
+// CIDR the server takes in its TRUSTED_PROXIES.
+func TestCheckCIDRTakesCIDR(t *testing.T) {
+	t.Parallel()
+
+	for _, cidr := range []string{"10.0.0.0/8", "192.0.2.1/32", "2001:db8::/32"} {
+		var stderr bytes.Buffer
+
+		code := cli.CheckCIDR(&stderr, cidr)
+		if code != 0 || stderr.Len() != 0 {
+			t.Errorf("%q: exit %d, stderr %q; want exit 0 and no output",
+				cidr, code, stderr.String())
+		}
+	}
+}
+
+// TestCheckCIDRRefusesOtherInput: check-cidr exits 1, with an error
+// naming TRUSTED_PROXIES and the input, for anything the server would
+// refuse in its TRUSTED_PROXIES, such as 1.2.3/32, which nginx would
+// look up as a hostname.
+func TestCheckCIDRRefusesOtherInput(t *testing.T) {
+	t.Parallel()
+
+	for _, cidr := range []string{
+		"1.2.3/32", "example.com/32", "10.0.0.0/33", "10.0.0.1", "",
+	} {
+		var stderr bytes.Buffer
+
+		code := cli.CheckCIDR(&stderr, cidr)
+
+		want := "TRUSTED_PROXIES " + strconv.Quote(cidr) + ": "
+		if code != 1 || !strings.HasPrefix(stderr.String(), want) {
+			t.Errorf("%q: exit %d, stderr %q; want exit 1 and %q...",
+				cidr, code, stderr.String(), want)
+		}
+	}
+}
+
+// TestPrepareDataDirTakesPathInVolume: prepare-data-dir exits 0,
+// writing nothing, for a DATA_DIR below the volume, and creates it.
+func TestPrepareDataDirTakesPathInVolume(t *testing.T) {
+	t.Parallel()
+
+	volume := t.TempDir()
+	dataDir := filepath.Join(volume, "reports")
+
+	var stderr bytes.Buffer
+
+	code := cli.PrepareDataDir(&stderr, volume, dataDir, currentUsername(t))
+	if code != 0 || stderr.Len() != 0 {
+		t.Errorf("exit %d, stderr %q; want exit 0 and no output",
+			code, stderr.String())
+	}
+
+	info, err := os.Stat(dataDir)
+	if err != nil || !info.IsDir() {
+		t.Errorf("DATA_DIR not created: %v", err)
+	}
+}
+
+// TestPrepareDataDirRefusesPathOutsideVolume: prepare-data-dir exits
+// 1, with an error naming DATA_DIR and its value, for a DATA_DIR
+// outside the volume.
+func TestPrepareDataDirRefusesPathOutsideVolume(t *testing.T) {
+	t.Parallel()
+
+	volume := t.TempDir()
+
+	var stderr bytes.Buffer
+
+	code := cli.PrepareDataDir(&stderr, volume, "/etc", currentUsername(t))
+
+	want := "DATA_DIR '/etc': " + reportbuf.ErrDataDirOutsideVolume.Error() + "\n"
+	if code != 1 || stderr.String() != want {
+		t.Errorf("exit %d, stderr %q; want exit 1 and %q",
+			code, stderr.String(), want)
+	}
+}
+
+// currentUsername is the name of the user the test runs as, the only
+// owner a test not run as root can give files to.
+func currentUsername(t *testing.T) string {
+	t.Helper()
+
+	current, err := user.Current()
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	return current.Username
+}
+
+// failIfServerStarts is what the tests give Main in place of
+// cli.RunServer: it fails the test instead of starting the server.
+func failIfServerStarts(t *testing.T) func(appname, version string) {
+	t.Helper()
+
+	return func(string, string) {
+		t.Error("Main started the server")
+	}
+}

internal/cli/export_test.go

@@ -0,0 +1,16 @@
+package cli
+
+import "io"
+
+// CheckCIDR exposes the check-cidr subcommand to the external tests,
+// which read what it writes to stderr.
+func CheckCIDR(stderr io.Writer, cidr string) int {
+	return checkCIDR(stderr, cidr)
+}
+
+// PrepareDataDir exposes the prepare-data-dir subcommand to the
+// external tests, which give it a volume and a user of their own in
+// place of /data and netwatch.
+func PrepareDataDir(stderr io.Writer, volume, dataDir, username string) int {
+	return prepareDataDir(stderr, volume, dataDir, username)
+}

internal/config/config.go

@@ -41,7 +41,6 @@ var (
 		"must be an origin, scheme://host with an optional port",
 	)
 	errNotPort = errors.New("must be a port number, 1 to 65535")
-	errNotBool = errors.New("must be true or false")
 	errNotIP   = errors.New("must be an IP address, or empty")
 
 	errMetricsCredentials = errors.New(
@@ -101,7 +100,7 @@ func New(
 	viper.SetDefault("CORS_ALLOWED_ORIGINS", "")
 	viper.SetDefault("DATA_DIR", "./data/reports")
 	viper.SetDefault("DATA_DIR_MAX_BYTES", defaultDataDirMaxBytes)
-	viper.SetDefault("DEBUG", "false")
+	viper.SetDefault("DEBUG", "")
 	// An empty BIND_ADDRESS listens on every interface.
 	viper.SetDefault("BIND_ADDRESS", "")
 	viper.SetDefault("PORT", "8080")
@@ -120,19 +119,17 @@ func New(
 		}
 	}
 
-	// Read with strconv: viper's GetInt and GetBool would read a value
+	// Read with strconv: viper's GetInt would read a value it cannot
-	// they cannot parse as 0 or false instead of failing.
+	// parse as 0 instead of failing.
 	port, err := strconv.Atoi(viper.GetString("PORT"))
 	if err != nil || port < 1 || port > math.MaxUint16 {
 		return nil, fmt.Errorf("PORT %q: %w",
 			viper.GetString("PORT"), errNotPort)
 	}
 
-	debug, err := strconv.ParseBool(viper.GetString("DEBUG"))
+	// Any non-empty value turns debug logging on, "false" and "0"
-	if err != nil {
+	// included, as the code style guide says.
-		return nil, fmt.Errorf("DEBUG %q: %w",
+	debug := viper.GetString("DEBUG") != ""
-			viper.GetString("DEBUG"), errNotBool)
-	}
 
 	// An empty TRUSTED_PROXIES trusts no proxy, but viper reads an
 	// empty environment variable as unset and would give the default,
@@ -172,8 +169,8 @@ func New(
 }
 
 // check fails with an error naming the first setting here whose value
-// the server cannot use. New checks PORT and DEBUG as it reads them,
+// the server cannot use. New checks PORT as it reads it, and the
-// and the middleware checks TRUSTED_PROXIES as it parses it.
+// middleware checks TRUSTED_PROXIES as it parses it.
 func (s *Config) check() error {
 	// viper reads a value that is not a number as 0, so this also
 	// catches a mistyped setting.

internal/config/config_test.go

@@ -13,8 +13,8 @@ import (
 	"go.uber.org/fx"
 )
 
-// requireConfigError builds the config as main does and fails the
+// requireConfigError builds the config as cli.RunServer does and fails
-// test unless that fails with an error naming setting. It uses
+// the test unless that fails with an error naming setting. It uses
 // fx.New, because fxtest.New fails the test itself on an error.
 func requireConfigError(t *testing.T, setting string) {
 	t.Helper()
@@ -31,8 +31,8 @@ func requireConfigError(t *testing.T, setting string) {
 	}
 }
 
-// requireTrustedProxies builds the config as main does and fails the
+// requireTrustedProxies builds the config as cli.RunServer does and
-// test unless its TrustedProxies are want.
+// fails the test unless its TrustedProxies are want.
 func requireTrustedProxies(t *testing.T, want []string) {
 	t.Helper()
 
@@ -127,12 +127,52 @@ func TestPortMustBeAPortNumber(t *testing.T) {
 	}
 }
 
-// TestDebugMustBeTrueOrFalse: viper reads any other value, such as
+// TestDebugIsOnForAnyNonEmptyValue: unset or empty, debug logging is
-// "yes", as false.
+// off; any other value turns it on, "false" and "0" included, and none
-func TestDebugMustBeTrueOrFalse(t *testing.T) {
+// stops the start.
-	t.Setenv("DEBUG", "yes")
+func TestDebugIsOnForAnyNonEmptyValue(t *testing.T) {
+	for _, tc := range []struct {
+		name  string
+		value string
+		unset bool
+		want  bool
+	}{
+		{name: "unset", unset: true, want: false},
+		{name: "empty", value: "", want: false},
+		{name: "false", value: "false", want: true},
+		{name: "0", value: "0", want: true},
+		{name: "yes", value: "yes", want: true},
+	} {
+		t.Run(tc.name, func(t *testing.T) {
+			// t.Setenv first, so the variable is put back as it was
+			// after the test.
+			t.Setenv("DEBUG", tc.value)
 
-	requireConfigError(t, "DEBUG")
+			if tc.unset {
+				err := os.Unsetenv("DEBUG")
+				if err != nil {
+					t.Fatal(err)
+				}
+			}
+
+			var cfg *config.Config
+
+			app := fx.New(
+				fx.NopLogger,
+				fx.Provide(globals.New, logger.New, config.New),
+				fx.Populate(&cfg),
+			)
+
+			err := app.Err()
+			if err != nil {
+				t.Fatalf("config error = %v", err)
+			}
+
+			if cfg.Debug != tc.want {
+				t.Fatalf("Debug = %t, want %t", cfg.Debug, tc.want)
+			}
+		})
+	}
 }
 
 // TestBindAddressMustBeAnIPAddress: a host name would be looked up

internal/globals/globals.go

@@ -4,7 +4,7 @@ package globals
 
 import "go.uber.org/fx"
 
-//nolint:gochecknoglobals // set from main before fx starts
+//nolint:gochecknoglobals // set by cli.RunServer before fx starts
 var (
 	// Appname is the application name.
 	Appname string

internal/handlers/healthcheck.go

@@ -3,9 +3,17 @@ package handlers
 import "net/http"
 
 // HandleHealthCheck returns a handler for the health check
-// endpoint.
+// endpoint. It answers 200 while the status is "ok" and 503 otherwise,
+// so that the image's HEALTHCHECK fails.
 func (s *Handlers) HandleHealthCheck() http.HandlerFunc {
 	return func(w http.ResponseWriter, r *http.Request) {
-		s.respondJSON(w, r, s.hc.Healthcheck(), http.StatusOK)
+		resp := s.hc.Healthcheck()
+
+		code := http.StatusOK
+		if resp.Status != "ok" {
+			code = http.StatusServiceUnavailable
+		}
+
+		s.respondJSON(w, r, resp, code)
 	}
 }

internal/handlers/healthcheck_test.go

@@ -9,17 +9,19 @@ import (
 	"testing"
 	"time"
 
+	"sneak.berlin/go/netwatch/internal/config"
 	"sneak.berlin/go/netwatch/internal/globals"
 	"sneak.berlin/go/netwatch/internal/handlers"
 	"sneak.berlin/go/netwatch/internal/healthcheck"
 	"sneak.berlin/go/netwatch/internal/logger"
+	"sneak.berlin/go/netwatch/internal/reportbuf"
 
 	"go.uber.org/fx/fxtest"
 )
 
 // newStartedHandlers builds Handlers with a real health check for the
-// server named in g, and starts them, which records the time the
+// server named in g, and a report buffer in a temporary directory, and
-// uptime counts from.
+// starts them, which records the time the uptime counts from.
 func newStartedHandlers(t *testing.T, g *globals.Globals) *handlers.Handlers {
 	t.Helper()
 
@@ -30,14 +32,23 @@ func newStartedHandlers(t *testing.T, g *globals.Globals) *handlers.Handlers {
 		t.Fatalf("logger: %v", err)
 	}
 
+	buf, err := reportbuf.New(lc, reportbuf.Params{
+		Config: &config.Config{DataDir: t.TempDir(), DataDirMaxBytes: 1 << 20},
+		Logger: log,
+	})
+	if err != nil {
+		t.Fatalf("report buffer: %v", err)
+	}
+
 	hc, err := healthcheck.New(lc,
-		healthcheck.Params{Globals: g, Logger: log})
+		healthcheck.Params{Buffer: buf, Globals: g, Logger: log})
 	if err != nil {
 		t.Fatalf("health check: %v", err)
 	}
 
-	h, err := handlers.New(lc,
+	h, err := handlers.New(lc, handlers.Params{
-		handlers.Params{Globals: g, Healthcheck: hc, Logger: log})
+		Buffer: buf, Globals: g, Healthcheck: hc, Logger: log,
+	})
 	if err != nil {
 		t.Fatalf("handlers: %v", err)
 	}
@@ -48,10 +59,12 @@ func newStartedHandlers(t *testing.T, g *globals.Globals) *handlers.Handlers {
 	return h
 }
 
-// TestHandleHealthCheck checks the health check's answer: 200, a JSON
+// TestHandleHealthCheck checks the health check's answer at start, when
-// content type, and a JSON object with exactly the fields of
+// no write of a report file has failed: 200, a JSON content type, and a
-// healthcheck.HealthcheckResponse, carrying this server's name and
+// JSON object with exactly the fields of healthcheck.HealthcheckResponse,
-// version and an uptime counted from its start.
+// carrying "status":"ok", this server's name and version and an uptime
+// counted from its start. The answer while writes fail is checked in
+// internal/reportbuf, whose tests can make a write fail.
 func TestHandleHealthCheck(t *testing.T) {
 	t.Parallel()
 

internal/healthcheck/healthcheck.go

@@ -10,6 +10,7 @@ import (
 	"sneak.berlin/go/netwatch/internal/config"
 	"sneak.berlin/go/netwatch/internal/globals"
 	"sneak.berlin/go/netwatch/internal/logger"
+	"sneak.berlin/go/netwatch/internal/reportbuf"
 
 	"go.uber.org/fx"
 )
@@ -18,6 +19,7 @@ import (
 type Params struct {
 	fx.In
 
+	Buffer  *reportbuf.Buffer
 	Config  *config.Config
 	Globals *globals.Globals
 	Logger  *logger.Logger
@@ -68,12 +70,18 @@ func New(
 	return s, nil
 }
 
-// Healthcheck returns the current health status of the application.
+// Healthcheck returns the current health status of the application:
+// "ok", or "error" while report files cannot be written.
 func (s *Healthcheck) Healthcheck() *HealthcheckResponse {
+	status := "ok"
+	if s.params.Buffer.LastWriteFailed() {
+		status = "error"
+	}
+
 	return &HealthcheckResponse{
 		Appname:       s.params.Globals.Appname,
 		Now:           time.Now().UTC().Format(time.RFC3339Nano),
-		Status:        "ok",
+		Status:        status,
 		UptimeHuman:   s.uptime().String(),
 		UptimeSeconds: int64(s.uptime().Seconds()),
 		Version:       s.params.Globals.Version,

internal/reportbuf/reportbuf.go

@@ -16,7 +16,6 @@ import (
 	"log/slog"
 	"os"
 	"path/filepath"
-	"slices"
 	"strings"
 	"sync"
 	"sync/atomic"
@@ -73,11 +72,9 @@ type Buffer struct {
 	// except in tests that hold the write open or make it fail.
 	fileCreated func(f *os.File)
 	// files are the report files that may be deleted to make room,
-	// in name order, which is oldest first: those in dataDir at
+	// oldest first: those in dataDir at start, in name order, then
-	// start, and each one this buffer writes, put in at its place by
+	// each one this buffer writes, once it is complete. A file still
-	// name once it is complete, even when an older file's write
+	// being written is not among them. filesBytes is their total size.
-	// completes after a newer one's. A file still being written is
-	// not among them. filesBytes is their total size.
 	files      []reportFile
 	filesBytes int64
 	log        *slog.Logger
@@ -88,8 +85,12 @@ type Buffer struct {
 	now func() time.Time
 	// seq numbers the report files, so that two named in the same
 	// millisecond still get different names.
-	seq      atomic.Uint64
+	seq      atomic.Int64
 	stopOnce sync.Once
+	// thresholdWriteWaiting is set while a write Append started at
+	// flushSizeThreshold waits for writeMu; Append starts no other
+	// while it is set.
+	thresholdWriteWaiting atomic.Bool
 	// usedBytes is what Append checks against maxBytes: the size
 	// of the report files in dataDir, plus the reports waiting to
 	// be written to one at their uncompressed size.
@@ -99,6 +100,11 @@ type Buffer struct {
 	// refuses reports, and the reports of that write wait in buf for
 	// the next flush.
 	writeErr error
+	// writeMu is held across each write of a report file, so writes
+	// never overlap: a failed write puts its reports back in buf before
+	// the next write takes them, which keeps them in the order they
+	// were appended.
+	writeMu sync.Mutex
 	// writers are the goroutines that write report files: flushLoop,
 	// and each one Append starts at flushSizeThreshold. OnStop waits
 	// for them before the final flush.
@@ -127,16 +133,33 @@ func New(
 
 	lc.Append(fx.Hook{
 		OnStart: func(_ context.Context) error {
+			// The errors name DATA_DIR, as config.New's do for the
+			// settings it checks.
 			err := os.MkdirAll(b.dataDir, dirPerms)
 			if err != nil {
-				return fmt.Errorf("create data dir: %w", err)
+				return fmt.Errorf("DATA_DIR %q: %w", b.dataDir, err)
+			}
+
+			// Creating and removing a file stops the start when no
+			// file can be created in DATA_DIR, instead of the first
+			// write of a report file failing after reports were
+			// answered 200. A full disk is not caught: the empty
+			// file is still created.
+			f, err := os.CreateTemp(b.dataDir, "")
+			if err != nil {
+				return fmt.Errorf("DATA_DIR %q: %w", b.dataDir, err)
+			}
+
+			err = errors.Join(f.Close(), os.Remove(f.Name()))
+			if err != nil {
+				return fmt.Errorf("DATA_DIR %q: %w", b.dataDir, err)
 			}
 
 			// Report files left by earlier runs count too, and are
 			// the first to be deleted to make room.
 			files, err := reportFiles(b.dataDir)
 			if err != nil {
-				return err
+				return fmt.Errorf("DATA_DIR %q: %w", b.dataDir, err)
 			}
 
 			b.mu.Lock()
@@ -190,8 +213,9 @@ func New(
 // write of a report file failed, until a write succeeds again. If the
 // line would take usedBytes past maxBytes, the oldest report files are
 // deleted to make room; it stores nothing and returns ErrFull if that
-// cannot make room. If the buffer reaches the size threshold, it is
+// cannot make room. If the buffer reaches the size threshold, it starts
-// drained and written to disk asynchronously.
+// a write of the reports in the background (see flushAtThreshold),
+// unless one it started is still waiting.
 func (b *Buffer) Append(v any) error {
 	line, err := json.Marshal(v)
 	if err != nil {
@@ -221,18 +245,9 @@ func (b *Buffer) Append(v any) error {
 	b.buf.Write(line)
 	b.buf.WriteByte('\n')
 
-	if b.buf.Len() >= flushSizeThreshold {
+	if b.buf.Len() >= flushSizeThreshold &&
-		data := b.drainBuf()
+		b.thresholdWriteWaiting.CompareAndSwap(false, true) {
-		b.mu.Unlock()
+		b.writers.Go(b.flushAtThreshold)
-
-		b.writers.Go(func() {
-			writeErr := b.writeFile(data)
-			if writeErr != nil {
-				b.log.Error("flush reports failed", "error", writeErr)
-			}
-		})
-
-		return nil
 	}
 
 	b.mu.Unlock()
@@ -240,6 +255,16 @@ func (b *Buffer) Append(v any) error {
 	return nil
 }
 
+// LastWriteFailed reports whether the last write of a report file to
+// finish failed: false at start until a write fails, and false again
+// once a write succeeds. While it is true, Append refuses reports.
+func (b *Buffer) LastWriteFailed() bool {
+	b.mu.Lock()
+	defer b.mu.Unlock()
+
+	return b.writeErr != nil
+}
+
 // deleteOldestFiles deletes report files, oldest first, until n more
 // bytes fit under maxBytes. It deletes none when the reports waiting
 // to be written leave no room for n even with every file gone, since
@@ -290,21 +315,28 @@ func (b *Buffer) flushLoop() {
 	}
 }
 
-// flushLocked acquires the lock, drains the buffer, and
+// flushLocked waits for any write of a report file under way, then
-// writes the data to a compressed file.
+// writes the reports waiting in the buffer to a new one.
 func (b *Buffer) flushLocked() error {
-	b.mu.Lock()
+	b.writeMu.Lock()
+	defer b.writeMu.Unlock()
 
-	if b.buf.Len() == 0 {
+	return b.writeFile()
-		b.mu.Unlock()
+}
 
-		return nil
+// flushAtThreshold is the write Append starts when the buffer reaches
+// flushSizeThreshold. Once any write under way is done, it lets Append
+// start another and writes the reports waiting in the buffer.
+func (b *Buffer) flushAtThreshold() {
+	b.writeMu.Lock()
+	defer b.writeMu.Unlock()
+
+	b.thresholdWriteWaiting.Store(false)
+
+	err := b.writeFile()
+	if err != nil {
+		b.log.Error("flush reports failed", "error", err)
 	}
-
-	data := b.drainBuf()
-	b.mu.Unlock()
-
-	return b.writeFile(data)
 }
 
 // drainBuf copies the buffer contents and resets it.
@@ -317,10 +349,22 @@ func (b *Buffer) drainBuf() []byte {
 	return data
 }
 
-// writeFile writes data, reports drained from the buffer, to a new
+// writeFile takes the reports waiting in the buffer, if any, and writes
-// timestamped zstd-compressed JSONL file in the data directory. If the
+// them to a new timestamped zstd-compressed JSONL file in the data
-// write fails, it puts the reports back in the buffer.
+// directory. If the write fails, it puts the reports back in the
-func (b *Buffer) writeFile(data []byte) error {
+// buffer. The caller must hold b.writeMu.
+func (b *Buffer) writeFile() error {
+	b.mu.Lock()
+
+	if b.buf.Len() == 0 {
+		b.mu.Unlock()
+
+		return nil
+	}
+
+	data := b.drainBuf()
+	b.mu.Unlock()
+
 	// The timestamp comes first, so the names sort by time; the number
 	// after it tells apart files named in the same millisecond.
 	ts := b.now().UTC().Format("2006-01-02T15-04-05.000Z")
@@ -351,14 +395,7 @@ func (b *Buffer) writeFile(data []byte) error {
 	b.usedBytes -= int64(len(data))
 	b.usedBytes += size
 	b.filesBytes += size
-
+	b.files = append(b.files, reportFile{name: name, size: size})
-	// At its place by name, not at the end: another write, of a newer
-	// file, may have completed while this one was being written.
-	i, _ := slices.BinarySearchFunc(b.files, name,
-		func(f reportFile, target string) int {
-			return strings.Compare(f.name, target)
-		})
-	b.files = slices.Insert(b.files, i, reportFile{name: name, size: size})
 
 	return nil
 }

internal/reportbuf/reportbuf_test.go

@@ -108,6 +108,68 @@ func TestFailedFinalFlushFailsStop(t *testing.T) {
 	}
 }
 
+// TestUncreatableDataDirStopsStart sets DATA_DIR below a regular file,
+// where no directory can be created, even by root, as the tests run in
+// the backend image. The start must fail with an error naming DATA_DIR
+// and its value, as the errors for the other settings do.
+func TestUncreatableDataDirStopsStart(t *testing.T) {
+	file := filepath.Join(t.TempDir(), "file")
+	writeBytes(t, file, 1)
+
+	dir := filepath.Join(file, "reports")
+	t.Setenv("DATA_DIR", dir)
+
+	app := fxtest.New(t,
+		fx.Provide(
+			globals.New,
+			logger.New,
+			config.New,
+			reportbuf.New,
+		),
+		fx.Invoke(func(*reportbuf.Buffer) {}),
+	)
+
+	err := app.Start(t.Context())
+	if err == nil {
+		t.Fatal("start succeeded with DATA_DIR below a regular file")
+	}
+
+	want := fmt.Sprintf("DATA_DIR %q: ", dir)
+	if !strings.Contains(err.Error(), want) {
+		t.Fatalf("start error = %q, want it to contain %q", err, want)
+	}
+}
+
+// TestUnwritableDataDirStopsStart sets DATA_DIR to /proc, which exists
+// and can be read, but where no file can be created, even by root, as
+// the tests run in the backend image: root ignores a directory the test
+// makes read-only. The start must fail with an error naming DATA_DIR and
+// its value, rather than the first write of a report file.
+func TestUnwritableDataDirStopsStart(t *testing.T) {
+	dir := "/proc"
+	t.Setenv("DATA_DIR", dir)
+
+	app := fxtest.New(t,
+		fx.Provide(
+			globals.New,
+			logger.New,
+			config.New,
+			reportbuf.New,
+		),
+		fx.Invoke(func(*reportbuf.Buffer) {}),
+	)
+
+	err := app.Start(t.Context())
+	if err == nil {
+		t.Fatal("start succeeded with DATA_DIR /proc")
+	}
+
+	want := fmt.Sprintf("DATA_DIR %q: ", dir)
+	if !strings.Contains(err.Error(), want) {
+		t.Fatalf("start error = %q, want it to contain %q", err, want)
+	}
+}
+
 // TestStopWaitsForPeriodicWrite holds open the write the periodic flush
 // starts, then stops the app. The stop must wait until the write is
 // released, and the report must then be in its file. The test runs in
@@ -272,9 +334,9 @@ func stopWhileHeld(
 	return <-stopped
 }
 
-// startApp starts a Buffer through fx, as main does, with the DATA_DIR
+// startApp starts a Buffer through fx, as cli.RunServer does, with the
-// and DATA_DIR_MAX_BYTES the calling test has set, and returns it with
+// DATA_DIR and DATA_DIR_MAX_BYTES the calling test has set, and returns
-// the app, for tests that stop the app themselves.
+// it with the app, for tests that stop the app themselves.
 func startApp(t *testing.T) (*reportbuf.Buffer, *fxtest.App) {
 	t.Helper()
 
@@ -296,8 +358,8 @@ func startApp(t *testing.T) (*reportbuf.Buffer, *fxtest.App) {
 	return buf, app
 }
 
-// startBuffer starts a Buffer through fx, as main does, with the
+// startBuffer starts a Buffer through fx, as cli.RunServer does, with
-// DATA_DIR and DATA_DIR_MAX_BYTES the calling test has set.
+// the DATA_DIR and DATA_DIR_MAX_BYTES the calling test has set.
 func startBuffer(t *testing.T) *reportbuf.Buffer {
 	t.Helper()
 
@@ -519,6 +581,62 @@ func TestWrittenFilesDeletedToMakeRoom(t *testing.T) {
 	}
 }
 
+// TestWrittenFilesDeletedOldestFirst writes three report files, then
+// appends a report that fits only once one of them is deleted. The
+// first one written must be the one deleted.
+func TestWrittenFilesDeletedOldestFirst(t *testing.T) {
+	const maxBytes = 1000
+
+	report := map[string]string{"id": "written"}
+
+	t.Setenv("DATA_DIR", t.TempDir())
+	t.Setenv("DATA_DIR_MAX_BYTES", strconv.Itoa(maxBytes))
+
+	buf := startBuffer(t)
+
+	var written []string
+
+	buf.OnFileCreated(func(f *os.File) { written = append(written, f.Name()) })
+
+	for range 3 {
+		err := buf.Append(report)
+		if err != nil {
+			t.Fatalf("append report: %v", err)
+		}
+
+		err = buf.Flush()
+		if err != nil {
+			t.Fatalf("flush: %v", err)
+		}
+	}
+
+	// The three files hold the same report, so they are the same size.
+	info, err := os.Stat(written[0])
+	if err != nil {
+		t.Fatalf("stat %s: %v", written[0], err)
+	}
+
+	// A report that fits beside two of the files, so deleting one makes
+	// exactly the room it needs.
+	pad := maxBytes - 2*int(info.Size()) -
+		lineBytes(t, map[string]string{"id": ""})
+
+	err = buf.Append(map[string]string{"id": strings.Repeat("a", pad)})
+	if err != nil {
+		t.Fatalf("report that needs room: %v", err)
+	}
+
+	if exists(t, written[0]) {
+		t.Fatal("first report file written kept when room was needed")
+	}
+
+	for _, path := range written[1:] {
+		if !exists(t, path) {
+			t.Fatalf("%s deleted before the first report file written", path)
+		}
+	}
+}
+
 // TestFailedDeletionStillCounts makes deleting the oldest report file
 // fail. It is still there, so it still takes room, and the next oldest
 // is deleted in its place.
@@ -604,17 +722,16 @@ func TestFileDeletedByHandFreesRoom(t *testing.T) {
 	}
 }
 
-// TestFileBeingWrittenIsNeverDeleted holds the write of one report file
+// TestFileBeingWrittenIsNeverDeleted holds the write of a report file
-// open while a second write completes, then sends a report that needs
+// open under a cap with room for two reports, then appends a second
-// room. Deleting either file would make it, and the one being written is
+// report and a third that needs room. The file being written is the
-// the older, but only the complete one may be deleted. Once the first
+// only report file, and it is not deleted to make room, so the third
-// write is complete, its file is deleted when room is needed.
+// report is refused. Once the write is complete, its file is deleted
+// when room is needed.
 func TestFileBeingWrittenIsNeverDeleted(t *testing.T) {
 	report := map[string]string{"id": "writing"}
 
 	t.Setenv("DATA_DIR", t.TempDir())
-	// Room for two reports waiting to be written, but not for two
-	// beside a report file.
 	t.Setenv("DATA_DIR_MAX_BYTES", strconv.Itoa(2*lineBytes(t, report)))
 
 	buf := startBuffer(t)
@@ -639,36 +756,25 @@ func TestFileBeingWrittenIsNeverDeleted(t *testing.T) {
 
 	writing := <-created
 
-	// Only the first write is held; the second goes through, and
+	// Errorf, not Fatalf, until the write is released, so that a
-	// so do the writes after it, the final one at stop included.
-	var complete string
-
-	buf.OnFileCreated(func(f *os.File) { complete = f.Name() })
-
-	// Errorf, not Fatalf, until the first write is released, so that a
 	// failure here does not leave it held.
 	err = buf.Append(report)
 	if err != nil {
 		t.Errorf("second report: %v", err)
 	}
 
-	err = buf.Flush()
-	if err != nil {
-		t.Errorf("flush of the second report: %v", err)
-	}
-
 	err = buf.Append(report)
-	if err != nil {
+	if !errors.Is(err, reportbuf.ErrFull) {
-		t.Errorf("report that needs room: %v", err)
+		t.Errorf("report that needs room: error = %v, want ErrFull", err)
 	}
 
 	if !exists(t, writing) {
 		t.Error("report file deleted while it was being written")
 	}
 
-	if exists(t, complete) {
+	// Only this write is held; the writes after it, the final one at
-		t.Error("complete report file kept, though room was needed")
+	// stop included, go through.
-	}
+	buf.OnFileCreated(func(*os.File) {})
 
 	close(release)
 
@@ -677,9 +783,11 @@ func TestFileBeingWrittenIsNeverDeleted(t *testing.T) {
 		t.Fatalf("flush of the first report: %v", err)
 	}
 
+	// The second report still waits to be written, so this one fits
+	// only once the file just written is deleted.
 	err = buf.Append(report)
 	if err != nil {
-		t.Fatalf("report after the first write was complete: %v", err)
+		t.Fatalf("report after the write was complete: %v", err)
 	}
 
 	if exists(t, writing) {
@@ -821,6 +929,82 @@ func TestFailedWriteKeepsReportsAppendedDuringIt(t *testing.T) {
 	}
 }
 
+// TestFailedWritesKeepReportsInOrder holds a write open, then appends a
+// report that takes the buffer to FlushSizeThreshold, so that Append
+// starts a second write. The held write fails first, then the second,
+// which waited for it and took both reports. The next flush writes them
+// in the order they were appended, the held write's first.
+func TestFailedWritesKeepReportsInOrder(t *testing.T) {
+	dir := t.TempDir()
+	t.Setenv("DATA_DIR", dir)
+
+	buf := startBuffer(t)
+
+	created := make(chan struct{})
+	release := make(chan struct{})
+
+	// Each write is held until release is closed, then fails, as its
+	// file is closed under it.
+	buf.OnFileCreated(func(f *os.File) {
+		created <- struct{}{}
+
+		<-release
+
+		_ = f.Close()
+	})
+
+	err := buf.Append(map[string]string{"id": "held"})
+	if err != nil {
+		t.Fatalf("report of the held write: %v", err)
+	}
+
+	flushed := make(chan error)
+
+	go func() { flushed <- buf.Flush() }()
+
+	<-created
+
+	big := strings.Repeat("a", reportbuf.FlushSizeThreshold)
+
+	// Errorf, not Fatalf, until the held write is released, so that a
+	// failure here does not leave it held.
+	err = buf.Append(map[string]string{"id": big})
+	if err != nil {
+		t.Errorf("report that starts the second write: %v", err)
+	}
+
+	close(release)
+
+	err = <-flushed
+	if err == nil {
+		t.Error("held write succeeded, though its file was closed under it")
+	}
+
+	// The second write has created its file, so it has taken the
+	// reports.
+	<-created
+
+	// Writes from here on, the final one at stop included, succeed.
+	buf.OnFileCreated(func(*os.File) {})
+
+	// Flush waits for the second write to fail before it writes.
+	err = buf.Flush()
+	if err != nil {
+		t.Fatalf("flush after the failed writes: %v", err)
+	}
+
+	files, err := readReportFiles(dir)
+	if err != nil {
+		t.Fatalf("read report files: %v", err)
+	}
+
+	want := `{"id":"held"}` + "\n" + `{"id":"` + big + `"}` + "\n"
+	if len(files) != 1 || files[0] != want {
+		t.Fatalf("%d report files, want one holding the held write's "+
+			"report, then the second one", len(files))
+	}
+}
+
 // TestReportsRefusedWhileWritesFail posts reports through the report
 // handler, as the page does. Once a write has failed, a report is
 // refused with 500 until a write succeeds again; the reports taken
@@ -890,9 +1074,54 @@ func TestReportsRefusedWhileWritesFail(t *testing.T) {
 	}
 }
 
+// TestHealthCheckFailsWhileWritesFail asks the health check handler
+// for its answer around a failed write. It answers 200 with
+// "status":"ok" at start, 503 with "status":"error" from the failed
+// write until a write succeeds, and 200 with "status":"ok" again after.
+func TestHealthCheckFailsWhileWritesFail(t *testing.T) {
+	dir := t.TempDir()
+	t.Setenv("DATA_DIR", dir)
+
+	buf, h := startBufferWithHandlers(t)
+
+	checkHealthCheck(t, h, "at start", http.StatusOK, "ok")
+
+	// A flush writes a file only when a report is waiting.
+	err := buf.Append(map[string]string{"id": "kept"})
+	if err != nil {
+		t.Fatalf("append: %v", err)
+	}
+
+	// With the data directory gone, the write has nowhere to go.
+	err = os.Remove(dir)
+	if err != nil {
+		t.Fatalf("remove data dir: %v", err)
+	}
+
+	err = buf.Flush()
+	if err == nil {
+		t.Fatal("flush succeeded with the data directory gone")
+	}
+
+	checkHealthCheck(t, h, "after the failed write",
+		http.StatusServiceUnavailable, "error")
+
+	err = os.Mkdir(dir, 0o750)
+	if err != nil {
+		t.Fatalf("make data dir: %v", err)
+	}
+
+	err = buf.Flush()
+	if err != nil {
+		t.Fatalf("flush with the data directory back: %v", err)
+	}
+
+	checkHealthCheck(t, h, "after a write succeeded", http.StatusOK, "ok")
+}
+
 // startBufferWithHandlers starts a Buffer and the Handlers that store
-// reports in it through fx, as main does, with the DATA_DIR the calling
+// reports in it through fx, as cli.RunServer does, with the DATA_DIR
-// test has set.
+// the calling test has set.
 func startBufferWithHandlers(
 	t *testing.T,
 ) (*reportbuf.Buffer, *handlers.Handlers) {
@@ -937,6 +1166,39 @@ func postReport(t *testing.T, h *handlers.Handlers, clientID string) int {
 	return rec.Code
 }
 
+// checkHealthCheck asks the health check handler for its answer, and
+// fails the test, saying when it asked, unless the answer has status
+// code wantCode and a "status" field of wantStatus.
+func checkHealthCheck(
+	t *testing.T,
+	h *handlers.Handlers,
+	when string,
+	wantCode int,
+	wantStatus string,
+) {
+	t.Helper()
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequestWithContext(t.Context(),
+		http.MethodGet, "/.well-known/healthcheck", http.NoBody)
+	h.HandleHealthCheck().ServeHTTP(rec, req)
+
+	var body struct {
+		Status string `json:"status"`
+	}
+
+	err := json.Unmarshal(rec.Body.Bytes(), &body)
+	if err != nil {
+		t.Fatalf("health check %s: body not JSON: %v (%q)",
+			when, err, rec.Body.String())
+	}
+
+	if rec.Code != wantCode || body.Status != wantStatus {
+		t.Fatalf("health check %s: %d with status %q, want %d with %q",
+			when, rec.Code, body.Status, wantCode, wantStatus)
+	}
+}
+
 // TestConcurrentAppendsStopAtCap appends from many goroutines at once
 // with room for exactly roomFor reports: exactly that many must be
 // taken, which holds only if Append checks and counts each report
@@ -1088,6 +1350,106 @@ func TestFlushAtSizeThreshold(t *testing.T) {
 	}
 }
 
+// TestEachSizeThresholdStartsAWrite reaches FlushSizeThreshold twice,
+// with no call to Flush and the periodic flush a minute away. Append
+// must start a write each time, so that the reports end up in two
+// report files. The test runs in a synctest bubble, so that
+// synctest.Wait returns once each write is done.
+func TestEachSizeThresholdStartsAWrite(t *testing.T) {
+	dir := t.TempDir()
+	t.Setenv("DATA_DIR", dir)
+
+	synctest.Test(t, func(t *testing.T) {
+		buf := startBuffer(t)
+
+		var appended string
+
+		for range 2 {
+			appended = appendToThreshold(t, buf)
+
+			synctest.Wait()
+		}
+
+		files, err := readReportFiles(dir)
+		if err != nil {
+			t.Fatalf("read report files: %v", err)
+		}
+
+		if !slices.Equal(files, []string{appended, appended}) {
+			t.Fatalf("%d report files, want two, each holding the %d "+
+				"bytes appended to reach the threshold",
+				len(files), len(appended))
+		}
+	})
+}
+
+// TestAppendStartsOneWaitingWrite holds a write open, then appends two
+// reports of FlushSizeThreshold bytes each. Append starts a write at
+// the first, which waits for the held one, and none at the second,
+// since that write is still waiting. Every write fails, so a write
+// started at the second report would find the reports back in the
+// buffer and try them again: once the held write is released, two
+// writes must have been started, the held one and one more.
+func TestAppendStartsOneWaitingWrite(t *testing.T) {
+	t.Setenv("DATA_DIR", t.TempDir())
+
+	synctest.Test(t, func(t *testing.T) {
+		buf := startBuffer(t)
+
+		var writes atomic.Int64
+
+		release := make(chan struct{})
+
+		// Each write is held until release is closed, then fails, as its
+		// file is closed under it.
+		buf.OnFileCreated(func(f *os.File) {
+			writes.Add(1)
+
+			<-release
+
+			_ = f.Close()
+		})
+
+		err := buf.Append(map[string]string{"id": "one-waiting"})
+		if err != nil {
+			t.Fatalf("report of the held write: %v", err)
+		}
+
+		go func() { _ = buf.Flush() }()
+
+		// Wait returns once the flush's write is held.
+		synctest.Wait()
+
+		big := map[string]string{
+			"id": strings.Repeat("a", reportbuf.FlushSizeThreshold),
+		}
+
+		// Errorf, not Fatalf, until the held write is released, so that a
+		// failure here does not leave it held.
+		for range 2 {
+			err = buf.Append(big)
+			if err != nil {
+				t.Errorf("report of FlushSizeThreshold bytes: %v", err)
+			}
+		}
+
+		close(release)
+
+		// Wait returns once every write started is done.
+		synctest.Wait()
+
+		started := writes.Load()
+
+		// Writes from here on, the final one at stop included, succeed.
+		buf.OnFileCreated(func(*os.File) {})
+
+		if started != 2 {
+			t.Fatalf("%d writes started, want 2: the held one and one "+
+				"at the threshold", started)
+		}
+	})
+}
+
 // appendToThreshold appends reports to buf until it holds
 // FlushSizeThreshold bytes, so that the last append starts writing them
 // to a report file, and returns the lines appended.
@@ -1233,85 +1595,3 @@ func hasReportFile(t *testing.T, dir string) bool {
 
 	return false
 }
-
-// TestFilesDeletedOldestFirstWhenWritesOverlap holds the write of an
-// older report file open until a newer one's write completes, then
-// releases it. When room is needed, the older file is deleted first,
-// though its write was the last to complete.
-func TestFilesDeletedOldestFirstWhenWritesOverlap(t *testing.T) {
-	const maxBytes = 1000
-
-	report := map[string]string{"id": "overlap"}
-
-	t.Setenv("DATA_DIR", t.TempDir())
-	t.Setenv("DATA_DIR_MAX_BYTES", strconv.Itoa(maxBytes))
-
-	buf := startBuffer(t)
-
-	created := make(chan string)
-	release := make(chan struct{})
-
-	buf.OnFileCreated(func(f *os.File) {
-		created <- f.Name()
-
-		<-release
-	})
-
-	err := buf.Append(report)
-	if err != nil {
-		t.Fatalf("older report: %v", err)
-	}
-
-	flushed := make(chan error)
-
-	go func() { flushed <- buf.Flush() }()
-
-	older := <-created
-
-	// Only the older write is held; the newer one goes through, and
-	// so do the writes after it, the final one at stop included.
-	var newer string
-
-	buf.OnFileCreated(func(f *os.File) { newer = f.Name() })
-
-	// Errorf, not Fatalf, until the older write is released, so that a
-	// failure here does not leave it held.
-	err = buf.Append(report)
-	if err != nil {
-		t.Errorf("newer report: %v", err)
-	}
-
-	err = buf.Flush()
-	if err != nil {
-		t.Errorf("flush of the newer report: %v", err)
-	}
-
-	close(release)
-
-	err = <-flushed
-	if err != nil {
-		t.Fatalf("flush of the older report: %v", err)
-	}
-
-	info, err := os.Stat(newer)
-	if err != nil {
-		t.Fatalf("stat %s: %v", newer, err)
-	}
-
-	// A report that fits beside the newer file alone, so deleting the
-	// older one makes exactly the room it needs.
-	pad := maxBytes - int(info.Size()) - lineBytes(t, map[string]string{"id": ""})
-
-	err = buf.Append(map[string]string{"id": strings.Repeat("a", pad)})
-	if err != nil {
-		t.Fatalf("report that needs room: %v", err)
-	}
-
-	if exists(t, older) {
-		t.Fatal("older report file kept when room was needed")
-	}
-
-	if !exists(t, newer) {
-		t.Fatal("newer report file deleted before the older one")
-	}
-}

internal/server/routes_test.go

@@ -22,9 +22,9 @@ import (
 	"go.uber.org/fx/fxtest"
 )
 
-// newServer builds a Server from the same constructors as main,
+// newServer builds a Server from the same constructors as
-// configured from the environment. It is never started, so nothing
+// cli.RunServer, configured from the environment. It is never started,
-// listens.
+// so nothing listens.
 func newServer(t *testing.T) *server.Server {
 	t.Helper()
 

internal/server/server.go

@@ -79,7 +79,7 @@ func New(
 			s.SetupRoutes()
 			s.httpServer = s.newHTTPServer()
 
-			go s.listenAndServe()
+			go func() { s.listenAndServe() }()
 
 			return nil
 		},

package.json

@@ -1,6 +1,6 @@
 {
     "name": "netwatch",
-    "version": "1.0.0",
+    "private": true,
     "description": "Real-time network latency monitor SPA",
     "type": "module",
     "scripts": {

script/add-dependency

@@ -23,8 +23,8 @@ main() {
     "" | -* | *[[:space:]]*) usage ;;
     esac
     cd "$ROOT"
-    yarn add --dev "$1"
+    "$ROOT/script/yarn" add --dev "$1"
-    yarn install --frozen-lockfile
+    "$ROOT/script/yarn" install --frozen-lockfile
 }
 
 main "$@"

script/bootstrap

@@ -3,11 +3,14 @@
 # this repo. Idempotent: every install is guarded by a check so already
 # installed tools are skipped. Base tooling comes from nix, apt, brew,
 # or apk (detected in that order); assumes nothing is present. Node is
-# used directly if it is at least NODE_MIN_VERSION; otherwise it is
+# used directly only if it has the major version of NODE_VERSION and is
-# installed at a pinned version via nvm (installing nvm itself first,
+# at least NODE_MIN_VERSION; otherwise NODE_VERSION is installed via nvm
-# from a hash-verified release archive, never curl | sh). Go, with its
+# (installing nvm itself first, from a hash-verified release archive,
-# gofmt, is used directly only if it is exactly GO_VERSION; otherwise
+# never curl | sh). Yarn is used directly only if it is exactly
-# that release is installed from its hash-verified archive.
+# YARN_VERSION; otherwise that version is prepared through corepack,
+# or installed with npm where there is no corepack.
+# Go, with its gofmt, is used directly only if it is exactly GO_VERSION;
+# otherwise that release is installed from its hash-verified archive.
 #
 # What this script installs outside the system package manager lives
 # under $HOME and is linked into ~/.local/bin, where make and the git
@@ -142,32 +145,63 @@ ensure_nvm() {
     rm -rf "$tmp"
 }
 
-# node_ok: the node on PATH is at least NODE_MIN_VERSION. node itself
+# node_ok: the node on PATH has the major version of NODE_VERSION and is
-# compares the two: major, then minor, then patch.
+# at least NODE_MIN_VERSION. Only the major is compared with the pin,
+# as REPO_POLICIES.md asks. node itself compares the versions: major,
+# then minor, then patch. A node that fails does not pass.
+# script/bootstrap and script/yarn carry this function unchanged.
 node_ok() {
-    if missing node; then return 1; fi
+    if ! command -v node >/dev/null 2>&1; then return 1; fi
     node -e '
         const have = process.versions.node.split(".").map(Number);
         const want = process.argv[1].split(".").map(Number);
+        const pinned = process.argv[2].split(".").map(Number);
+        if (have[0] !== pinned[0]) process.exit(1);
         for (let i = 0; i < 3; i++) {
             if (have[i] !== want[i]) process.exit(have[i] > want[i] ? 0 : 1);
         }
-    ' "$NODE_MIN_VERSION"
+    ' "$NODE_MIN_VERSION" "$NODE_VERSION"
 }
 
 # ensure_node: unless node_ok, install NODE_VERSION and link its node.
+# Afterwards node is looked up through PATH again and must pass node_ok.
 ensure_node() {
-    if node_ok; then return 0; fi
+    if node_ok; then
+        echo "bootstrap: using node $(node --version)"
+        return 0
+    fi
     ensure_nvm
     nvm_sh "nvm install $NODE_VERSION"
     link_bin "$HOME/.nvm/versions/node/v$NODE_VERSION/bin/node" node
+    hash -r
+    if ! node_ok; then
+        echo "bootstrap: after installing node $NODE_VERSION through nvm," >&2
+        echo "  the node on PATH, $(command -v node), is not it" >&2
+        exit 1
+    fi
+    echo "bootstrap: installed node $(node --version)"
 }
 
-# ensure_yarn: corepack writes its shims (pnpm and yarnpkg as well as
+# yarn_ok: yarn --version on PATH prints exactly YARN_VERSION, compared
-# yarn) into $TOOLCHAIN rather than next to itself, and the npm fallback
+# over the whole version. A yarn that fails or prints anything else does
-# installs there too; only yarn is linked.
+# not pass.
+yarn_ok() {
+    if missing yarn; then return 1; fi
+    version="$(yarn --version 2>/dev/null)" || return 1
+    [ "$version" = "$YARN_VERSION" ]
+}
+
+# ensure_yarn: unless yarn_ok, prepare YARN_VERSION and link its yarn.
+# corepack writes its shims (pnpm and yarnpkg as well as yarn) into
+# $TOOLCHAIN rather than next to itself, and the npm fallback installs
+# there too; only yarn is linked. Afterwards yarn is looked up through
+# PATH again and must pass yarn_ok, so another yarn that hides the link
+# stops bootstrap.
 ensure_yarn() {
-    if ! missing yarn; then return 0; fi
+    if yarn_ok; then
+        echo "bootstrap: using yarn $(yarn --version)"
+        return 0
+    fi
     shims="$TOOLCHAIN/corepack-shims"
     mkdir -p "$shims"
     if ! missing corepack; then
@@ -182,6 +216,13 @@ ensure_yarn() {
         shims="$TOOLCHAIN/npm-global/bin"
     fi
     link_bin "$shims/yarn" yarn
+    hash -r
+    if ! yarn_ok; then
+        echo "bootstrap: after installing yarn $YARN_VERSION in $shims," >&2
+        echo "  the yarn on PATH, $(command -v yarn), does not report it" >&2
+        exit 1
+    fi
+    echo "bootstrap: installed yarn $(yarn --version)"
 }
 
 # go_ok: the go on PATH has its gofmt beside it (a Go release ships the

script/build

@@ -8,11 +8,12 @@ ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    yarn build
+    "$ROOT/script/yarn" build
 
-    # VERSION comes from the environment. Unset or empty, it is git
+    # VERSION comes from the environment. Unset or empty, it is
-    # describe, or "dev" where there is no git or no repository history.
+    # `git describe --tags --always --dirty`, as in script/docker, or
-    version="${VERSION:-$(git describe --always --dirty 2>/dev/null || echo dev)}"
+    # "dev" where there is no git or no repository history.
+    version="${VERSION:-$(git describe --tags --always --dirty 2>/dev/null || echo dev)}"
 
     CGO_ENABLED=0 go build -trimpath \
         -ldflags "-s -w -X main.Version=$version" \

script/cibuild

@@ -2,9 +2,12 @@
 # script/cibuild: run the CI build. It bootstraps first: a CI runner
 # checks out and runs this and nothing else, and script/fmt-check runs
 # the formatter on the host, which a pristine checkout cannot do.
-# --no-cache for the same reason as script/docker: the gate phases the
+# A new CACHEBUST value reruns every RUN step after the ARG CACHEBUST
-# final stage depends on are RUN steps, and a cached one is a check that
+# line, since a cached gate phase is a check that did not run; unlike
-# did not run.
+# --no-cache, it keeps the layers above that line and the cache mounts.
+# This relies on every stage of the Dockerfile declaring ARG CACHEBUST
+# before every step that lints, tests or compiles; a stage without that
+# line is served from cache on an unchanged tree, with no warning.
 set -eu
 
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd -P)"
@@ -14,15 +17,17 @@ main() {
     cd "$ROOT"
     "$SCRIPT_DIR/bootstrap"
     "$SCRIPT_DIR/check"
-    # Own line: a failing command substitution inside an argument does
+    # The version and the tag each get their own line: a failing
-    # not trip `set -e`, so the inline form degrades silently to an
+    # command substitution inside an argument does not trip `set -e`,
-    # empty constant. The VERSION build argument takes precedence over
+    # so the inline form degrades silently to an empty constant. The
-    # the version a build stage derives from the .git in the context.
+    # VERSION build argument takes precedence over the version a build
+    # stage derives from the .git in the context.
     version="$(git describe --tags --always --dirty 2>/dev/null || true)"
     [ -n "$version" ] || version="unknown"
-    docker build --no-cache \
+    tag="$("$SCRIPT_DIR/projectname")"
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
         --build-arg VERSION="$version" \
-        -t "$("$SCRIPT_DIR/projectname")" .
+        -t "$tag" .
 }
 
 main "$@"

script/dev

@@ -7,7 +7,7 @@ ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    yarn dev
+    "$ROOT/script/yarn" dev
 }
 
 main "$@"

script/docker

@@ -1,8 +1,12 @@
 #!/bin/sh
 # script/docker: build the Docker image tagged with the project name.
 # Identical in all repos; the tag comes from script/projectname.
-# --no-cache because the gate phases the final stage depends on are RUN
+# A new CACHEBUST value reruns every RUN step after the ARG CACHEBUST
-# steps, and a cached one is a check that did not run.
+# line, since a cached gate phase is a check that did not run; unlike
+# --no-cache, it keeps the layers above that line and the cache mounts.
+# This relies on every stage of the Dockerfile declaring ARG CACHEBUST
+# before every step that lints, tests or compiles; a stage without that
+# line is served from cache on an unchanged tree, with no warning.
 set -eu
 
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd -P)"
@@ -10,15 +14,17 @@ ROOT="$(cd "$SCRIPT_DIR/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    # Own line: a failing command substitution inside an argument does
+    # The version and the tag each get their own line: a failing
-    # not trip `set -e`, so the inline form degrades silently to an
+    # command substitution inside an argument does not trip `set -e`,
-    # empty constant. The VERSION build argument takes precedence over
+    # so the inline form degrades silently to an empty constant. The
-    # the version a build stage derives from the .git in the context.
+    # VERSION build argument takes precedence over the version a build
+    # stage derives from the .git in the context.
     version="$(git describe --tags --always --dirty 2>/dev/null || true)"
     [ -n "$version" ] || version="unknown"
-    docker build --no-cache \
+    tag="$("$SCRIPT_DIR/projectname")"
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
         --build-arg VERSION="$version" \
-        -t "$("$SCRIPT_DIR/projectname")" .
+        -t "$tag" .
 }
 
 main "$@"

script/frontend-fmt

@@ -10,7 +10,7 @@ ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    yarn prettier --write '**/*.{js,css,html,md}'
+    "$ROOT/script/yarn" prettier --write '**/*.{js,css,html,md}'
 }
 
 main "$@"

script/frontend-fmt-check

@@ -7,7 +7,7 @@ ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    yarn prettier --check '**/*.{js,css,html,md}'
+    "$ROOT/script/yarn" prettier --check '**/*.{js,css,html,md}'
 }
 
 main "$@"

script/frontend-test

@@ -14,12 +14,14 @@ ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    NODE_OPTIONS=--test-reporter=dot timeout 90 yarn --silent run test || {
+    NODE_OPTIONS=--test-reporter=dot \
+        timeout 90 "$ROOT/script/yarn" --silent run test || {
         echo "--- Rerunning with every test listed for details ---"
-        NODE_OPTIONS=--test-reporter=spec timeout 90 yarn --silent run test
+        NODE_OPTIONS=--test-reporter=spec \
+            timeout 90 "$ROOT/script/yarn" --silent run test
         exit 1
     }
-    timeout 30 yarn build
+    timeout 30 "$ROOT/script/yarn" build
 }
 
 main "$@"

script/frontend-viewport-test

@@ -77,16 +77,23 @@ main() {
         -v "$ROOT/configs/security-headers.conf:/etc/nginx/security-headers.conf:ro" \
         "$SERVER_IMAGE" > /dev/null
 
-    # The page, and a path with no file of its own, which falls back to
+    # The page, and /no-such-page, which location / answers with the
-    # it, must go out with Cache-Control: no-cache. That add_header in
+    # page through its try_files fallback, must go out with
-    # its location drops the server's own unless the location includes
+    # Cache-Control: no-cache. That add_header in its location drops the
-    # the security headers again; only Content-Security-Policy is
+    # server's own unless the location includes the security headers
-    # checked, as it comes from the same file as the rest. curl tries up
+    # again; only Content-Security-Policy is checked, as it comes from
-    # to ten more times, a second apart, while nginx starts.
+    # the same file as the rest. curl tries up to ten more times, a
+    # second apart, while nginx starts. It runs without -S, which would
+    # print an error for every failed attempt, even when a later one
+    # succeeds.
     for path in / /no-such-page; do
-        headers="$(docker exec "$SERVER" curl -sS -o /dev/null -D - \
+        if ! headers="$(docker exec "$SERVER" curl -s -o /dev/null -D - \
             --retry 10 --retry-connrefused --retry-delay 1 \
-            "http://127.0.0.1:8080$path" | tr -d '\r')"
+            "http://127.0.0.1:8080$path")"; then
+            echo "frontend-viewport-test: nginx did not answer $path" >&2
+            exit 1
+        fi
+        headers="$(printf '%s\n' "$headers" | tr -d '\r')"
         if ! printf '%s\n' "$headers" | grep -qix 'Cache-Control: no-cache' ||
             ! printf '%s\n' "$headers" | grep -qi '^Content-Security-Policy:'; then
             echo "frontend-viewport-test: $path needs" \

script/lint

@@ -5,9 +5,14 @@
 # make its answer untrustworthy.
 #
 # The phase is not the last stage in the file, so it is built only when
-# --target names it. --no-cache because a cached lint layer is a lint
+# --target names it. A new CACHEBUST value reruns every RUN step after
-# that did not run. The tag makes each build replace the previous image
+# the ARG CACHEBUST line, since a cached lint layer is a lint that did
-# instead of leaving a dangling one behind.
+# not run; unlike --no-cache, it keeps the layers above that line and
+# the cache mounts. This relies on every stage of the Dockerfile
+# declaring ARG CACHEBUST before every step that lints, tests or
+# compiles; a stage without that line is served from cache on an
+# unchanged tree, with no warning.
+# --output type=cacheonly writes no image, since nothing uses one.
 set -eu
 
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd -P)"
@@ -15,9 +20,9 @@ ROOT="$(cd "$SCRIPT_DIR/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    docker build --no-cache \
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
         --target lint \
-        -t "$("$SCRIPT_DIR/projectname")-lint" .
+        --output type=cacheonly .
 }
 
 main "$@"

script/test

@@ -2,8 +2,13 @@
 # script/test: run the test suite. Testing is a phase of the Dockerfile
 # and this builds that phase alone, on the same terms as script/lint:
 # --target because a phase that is not the last stage is built only when
-# named, --no-cache because a cached test layer is a test that did not
+# named, and a new CACHEBUST value because a cached test layer is a test
-# run, and a tag so each build replaces the previous image.
+# that did not run; unlike --no-cache, it keeps the layers above the ARG
+# CACHEBUST line and the cache mounts. This relies on every stage of the
+# Dockerfile declaring ARG CACHEBUST before every step that lints, tests
+# or compiles; a stage without that line is served from cache on an
+# unchanged tree, with no warning. --output type=cacheonly writes no
+# image, since nothing uses one.
 set -eu
 
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd -P)"
@@ -11,9 +16,9 @@ ROOT="$(cd "$SCRIPT_DIR/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    docker build --no-cache \
+    docker build --build-arg CACHEBUST="$(date +%s)-$$" \
         --target test \
-        -t "$("$SCRIPT_DIR/projectname")-test" .
+        --output type=cacheonly .
 }
 
 main "$@"

script/yarn

@@ -0,0 +1,55 @@
+#!/bin/sh
+# script/yarn: run yarn with the given arguments under the node
+# script/bootstrap would use. Every script that runs yarn on the host
+# runs it through this one, except script/bootstrap, which picks its
+# own node and yarn.
+set -eu
+
+ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
+
+# Must match the pins in script/bootstrap.
+NODE_VERSION="22.17.0"
+NODE_MIN_VERSION="22.13.0"
+
+# node_ok: the node on PATH has the major version of NODE_VERSION and is
+# at least NODE_MIN_VERSION. Only the major is compared with the pin,
+# as REPO_POLICIES.md asks. node itself compares the versions: major,
+# then minor, then patch. A node that fails does not pass.
+# script/bootstrap and script/yarn carry this function unchanged.
+node_ok() {
+    if ! command -v node >/dev/null 2>&1; then return 1; fi
+    node -e '
+        const have = process.versions.node.split(".").map(Number);
+        const want = process.argv[1].split(".").map(Number);
+        const pinned = process.argv[2].split(".").map(Number);
+        if (have[0] !== pinned[0]) process.exit(1);
+        for (let i = 0; i < 3; i++) {
+            if (have[i] !== want[i]) process.exit(have[i] > want[i] ? 0 : 1);
+        }
+    ' "$NODE_MIN_VERSION" "$NODE_VERSION"
+}
+
+# Run the yarn script/bootstrap installed: the yarn on PATH when the
+# node on PATH passes node_ok, and otherwise yarn under the pinned node
+# in nvm. nvm is a bash script, hence the subshell.
+run_yarn() {
+    if node_ok && command -v yarn >/dev/null 2>&1; then
+        exec yarn "$@"
+    fi
+    if node_ok || [ ! -s "$HOME/.nvm/nvm.sh" ]; then
+        echo "script/yarn: no yarn; run script/bootstrap first" >&2
+        exit 1
+    fi
+    exec bash -c '. "$HOME/.nvm/nvm.sh" && nvm use "$1" >/dev/null &&
+        shift && exec yarn "$@"' bash "$NODE_VERSION" "$@"
+}
+
+main() {
+    cd "$ROOT"
+    # script/bootstrap links the node and yarn it installs into
+    # ~/.local/bin, which the shell that called it may not have on PATH.
+    PATH="$HOME/.local/bin:$PATH"
+    run_yarn "$@"
+}
+
+main "$@"

test/unit/main.test.js

@@ -83,6 +83,11 @@ function statusText(state, host) {
     return elements[`.status-text[data-host="${index}"]`]?.innerHTML;
 }
 
+// What this test wrote to the browser console with console[method].
+function consoleLines(method) {
+    return console[method].mock.calls.map((call) => call.arguments[0]);
+}
+
 // Mocks the clock for test t, so that a check lasting seconds takes no real
 // time, and replaces fetch with targets that each answer after
 // answerAfter(url) milliseconds of that clock, or never when that is
@@ -202,9 +207,10 @@ test("at a 30000ms interval, a target answering after 1000ms shows in its row wh
 });
 
 // In the next three tests, the answering target's check is still waiting
-// when something happens after which its result must not show.
+// when something happens after which its result must not be recorded or
+// show in its row, and the check must write nothing to the console.
 
-test("at a 30000ms interval, a check still waiting when its round is given up does not show in its row", async (t) => {
+test("at a 30000ms interval, a check still waiting when its round is given up is not recorded and does not show in its row", async (t) => {
     CONFIG.updateInterval = 30000;
     const state = new AppState([
         { name: "Answering", url: "https://answering.test" },
@@ -220,9 +226,13 @@ test("at a 30000ms interval, a check still waiting when its round is given up do
     roundChecks.abort();
     assert.notEqual(await settled(round), "still waiting");
     assert.equal(latencyFigure(state, answering), undefined);
+    assert.deepEqual(answering.history, []);
+    assert.equal(answering.status, "pending");
+    assert.deepEqual(consoleLines("error"), []);
+    assert.deepEqual(consoleLines("info"), []);
 });
 
-test("at a 30000ms interval, a check still waiting when the user pauses does not show in its row", async (t) => {
+test("at a 30000ms interval, a check still waiting when the user pauses is not recorded and does not show in its row", async (t) => {
     CONFIG.updateInterval = 30000;
     const state = new AppState([
         { name: "Answering", url: "https://answering.test" },
@@ -237,9 +247,13 @@ test("at a 30000ms interval, a check still waiting when the user pauses does not
     t.mock.timers.tick(500);
     assert.equal(await settled(round), "still waiting");
     assert.equal(latencyFigure(state, answering), undefined);
+    assert.deepEqual(answering.history, []);
+    assert.equal(answering.status, "pending");
+    assert.deepEqual(consoleLines("error"), []);
+    assert.deepEqual(consoleLines("info"), []);
 });
 
-test("at a 30000ms interval, a check in the first round does not show in its row", async (t) => {
+test("at a 30000ms interval, a check in the first round is not recorded and does not show in its row", async (t) => {
     CONFIG.updateInterval = 30000;
     const state = new AppState([
         { name: "Answering", url: "https://answering.test" },
@@ -251,6 +265,10 @@ test("at a 30000ms interval, a check in the first round does not show in its row
     t.mock.timers.tick(1000);
     assert.equal(await settled(round), "still waiting");
     assert.equal(latencyFigure(state, answering), undefined);
+    assert.deepEqual(answering.history, []);
+    assert.equal(answering.status, "pending");
+    assert.deepEqual(consoleLines("error"), []);
+    assert.deepEqual(consoleLines("info"), []);
 });
 
 test("at a 30000ms interval, after the user pauses and resumes during a round, no row reads paused once its last check ends", async (t) => {
@@ -300,11 +318,6 @@ function stateWithOneTarget() {
     return { state, target: state.local[0] };
 }
 
-// What this test wrote to the browser console with console[method].
-function consoleLines(method) {
-    return console[method].mock.calls.map((call) => call.arguments[0]);
-}
-
 test("a check that fails with a network error writes one console line with the target, the time, the error and how long the request took", async (t) => {
     const { state, target } = stateWithOneTarget();
     mockTargets(t, () => 23, target.url);