Footer no longer says IPv4 only (closes #111)

Each check is a fetch: the browser picks IPv4 or IPv6 for each WAN
host, and the local targets are IPv4 addresses, so "IPv4 only" was
wrong for the page as a whole. The footer's first line drops it and
its separator; the rest of the footer is unchanged. TODO.md moves this
to Completed Steps and the layout decision from Future Steps into Next
Step.

Model: opus-5-5

.prettierignore

@@ -1,6 +1,7 @@
-backend/
 dist/
 node_modules/
 tmp/
 yarn.lock
 .claude/
+# The org standard file, copied verbatim; backend/script/lint checks its sha256.
+backend/.golangci.yml

Dockerfile

@@ -1,6 +1,6 @@
 # The one image netwatch ships: nginx serves the built frontend and
-# passes /api/ and /.well-known/healthcheck to netwatch-server, the Go
+# passes /api/, /.well-known/healthcheck and /metrics to netwatch-server,
-# backend, which runs in the same container on loopback only.
+# the Go backend, which runs in the same container on loopback only.
 # bin/entrypoint.sh starts and watches both.
 
 # Lint stage — fast feedback on formatting and lint issues. The
@@ -59,20 +59,35 @@ RUN version="${VERSION:-$(git --git-dir=/git describe --tags --always)}"; \
     esac; \
     VERSION="$version" make build
 
+# Frontend lint stage — eslint over the JavaScript, as the lint stage
+# above lints the Go. The root make lint builds this stage alone too.
+# node:22-alpine as of 2026-02-22
+FROM node@sha256:e4bf2a82ad0a4037d28035ae71529873c069b13eb0455466ae0bc13363826e34 AS frontend-lint
+WORKDIR /app
+COPY package.json yarn.lock ./
+RUN yarn install --frozen-lockfile
+COPY . .
+RUN script/frontend-lint
+
 # Frontend stage
 # node:22-alpine as of 2026-02-22
 FROM node@sha256:e4bf2a82ad0a4037d28035ae71529873c069b13eb0455466ae0bc13363826e34 AS frontend
 WORKDIR /app
+
+# Force BuildKit to run the frontend-lint stage before proceeding, as
+# the builder stage does with the lint stage: without this no-op copy an
+# eslint failure would not gate the image.
+COPY --from=frontend-lint /app/yarn.lock /dev/null
+
 COPY package.json yarn.lock ./
 RUN yarn install --frozen-lockfile
 RUN apk add --no-cache git make
 COPY . .
-# make frontend-check is the frontend half of make check (test + lint +
+# make frontend-check runs the frontend tests and format check; its test
-# fmt-check); its test step runs the unit tests, then the production
+# step runs the unit tests, then the production yarn build, so this both
-# yarn build, so this both produces dist/ and gates the image on
+# produces dist/ and gates the image on test and formatting regressions.
-# lint/fmt-check/test regressions.
+# This node stage has neither Go nor Docker; the frontend-lint, lint and
-# This node stage has neither Go nor Docker; the lint and builder stages
+# builder stages above gate the rest.
-# above gate the backend half.
 RUN make frontend-check
 
 # Runtime stage

Makefile

@@ -1,5 +1,5 @@
-.PHONY: bootstrap setup dev test lint fmt fmt-check check frontend-check \
+.PHONY: bootstrap setup dev build test lint fmt fmt-check check \
-	frontend-viewport-test docker hooks
+	add-dependency tidy frontend-check frontend-viewport-test docker hooks
 
 # Standard targets are thin shims; the implementations live in script/
 # per the scripts-to-rule-them-all pattern (see the Entrypoints section
@@ -13,7 +13,11 @@ setup:
 	@script/setup
 
 dev:
-	yarn dev
+	@script/dev
+
+# The frontend only; backend/Makefile's build target builds the Go server.
+build:
+	@script/build
 
 test:
 	@script/test
@@ -30,8 +34,16 @@ fmt-check:
 check:
 	@script/check
 
-# The frontend half of check, for Dockerfile's node build stage, which
+# make add-dependency PACKAGE=<name>@<version>. PACKAGE reaches the
-# has neither Go nor Docker. Use check everywhere else.
+# script through the environment, so the shell never reads it as code.
+add-dependency:
+	@script/add-dependency "$$PACKAGE"
+
+tidy:
+	@script/tidy
+
+# The frontend tests and format check, for Dockerfile's frontend stage,
+# which has neither Go nor Docker. Use check everywhere else.
 frontend-check:
 	@script/frontend-check
 

README.md

@@ -1,30 +1,33 @@
 NetWatch is an MIT-licensed JavaScript single-page application by
 [@sneak](https://sneak.berlin) that provides real-time network latency
 monitoring to common internet hosts, displayed with color-coded figures and
-sparkline graphs, served from a static bucket or Docker container.
+sparkline graphs, served from a static bucket or from its Docker image, where a
+small Go backend stores the measurements the page reports.
 
 ## Getting Started
 
 ```bash
-# Install dependencies
+# Install the dependencies and the git pre-commit hook
-yarn install
+make setup
 
-# Development server
+# Run the page on the Vite dev server
-yarn dev
+make dev
 
-# Production build
+# Run the tests, both linters and the format check
-yarn build
+make check
 
-# Preview production build
+# Build the page into dist/
-yarn preview
+make build
 
-# Docker
+# Build the image and run it
-docker build -t netwatch .
+make docker
 docker run -p 8080:8080 netwatch
 ```
 
-`yarn dev` proxies `/api` to `http://127.0.0.1:8080`, so a locally running
+`make check` and `make docker` need Docker. `make dev` passes `/api` to
-`netwatch-server` (see `backend/`) receives the reports the page posts.
+`http://127.0.0.1:8080`, where `make run` in `backend/` starts `netwatch-server`
+with its defaults, so the reports the page posts are stored in
+`backend/data/reports`.
 
 ## Entrypoints
 
@@ -42,25 +45,42 @@ halves, so the root `make check` fails if either one is broken. We provide:
   as new as `backend/go.mod` asks for is installed, the Go modules, and gcc with
   the C library headers unless gcc is installed, for the race detector in
   `make test`), linking what it installs itself into `~/.local/bin`, which has
-  to be on `PATH`. It installs no Go linter and not Docker: `make lint` runs the
+  to be on `PATH`. It installs no Go linter and not Docker: `make lint` runs
-  linter in Docker
+  both linters 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
+  running `netwatch-server`
+- `script/build` — build the frontend for production into `dist/`;
+  `backend/script/build` builds the Go server
 - `script/projectname` — print the project name (used for the Docker image tag)
-- `script/test` — run `script/frontend-test`, then the backend's Go tests, each
+- `script/test` — run `script/frontend-test`, then `backend/script/test`, the
-  under its own 30-second timeout
+  backend's Go tests with the race detector and coverage
-- `script/lint` — run `script/frontend-lint`, then golangci-lint in Docker, by
+- `script/lint` — run eslint, then golangci-lint, both in Docker, by building
-  building the lint stage of `Dockerfile` without the cache
+  the `frontend-lint` and `lint` stages of `Dockerfile` without the cache
 - `script/fmt` — format all files (writes): prettier, then gofmt over `backend/`
 - `script/fmt-check` — check formatting (read-only): prettier, then gofmt
 - `script/check` — run test, lint, and fmt-check
+- `script/add-dependency` — add a frontend package, or move one to another
+  version: `make add-dependency PACKAGE=<name>@<version>` runs `yarn add --dev`,
+  which changes `package.json` and `yarn.lock` together, then
+  `yarn install --frozen-lockfile`
+- `script/tidy` — run `go mod tidy` in `backend/`: to add a Go module, import it
+  and run `make tidy`; to move one to another version, edit its `require` line
+  in `backend/go.mod`, then run `make tidy`
 - `script/frontend-test` — run the unit tests in `test/unit/` with Node's
-  built-in test runner, then the production build
+  built-in test runner, through the `test` script in `package.json`, and if any
-- `script/frontend-lint` — run prettier in check mode
+  fails, run them again listing every test, and fail; then the production build.
-- `script/frontend-fmt` — format everything prettier understands (writes)
+  Each run has a 30-second timeout
+- `script/frontend-lint` — run eslint with the rules in `eslint.config.js`; it
+  runs inside the `frontend-lint` stage of `Dockerfile`, which `make lint`
+  builds
+- `script/frontend-fmt` — format everything prettier understands (writes), the
+  markdown in `backend/` included
 - `script/frontend-fmt-check` — check prettier formatting (read-only)
-- `script/frontend-check` — the frontend half of `script/check`, for
+- `script/frontend-check` — run `script/frontend-test` and
-  `Dockerfile`, whose node build stage has neither Go nor Docker
+  `script/frontend-fmt-check`, for the frontend stage of `Dockerfile`, which has
+  neither Go nor Docker
 - `script/frontend-viewport-test` — responsive-layout verification of the built
   frontend in a containerised headless Chrome (see
   [test/viewport/README.md](test/viewport/README.md)). Not part of
@@ -77,25 +97,30 @@ halves, so the root `make check` fails if either one is broken. We provide:
 The narrow-viewport layout lives in the `max-width: 768px` media block in
 `src/styles.css`. It is verified automatically by `make frontend-viewport-test`,
 which drives a digest-pinned headless Chrome against the built `dist/` and
-asserts on computed layout at widths derived from that CSS — one pixel either
+asserts on computed layout at widths derived from that CSS — on every breakpoint
-side of every breakpoint it declares, plus a 320px floor, a desktop baseline and
+it declares and one pixel either side of it, plus a 320px floor, a desktop
-two landscape sizes. See [test/viewport/README.md](test/viewport/README.md) for
+baseline and two landscape sizes. See
-what it covers and what it genuinely cannot.
+[test/viewport/README.md](test/viewport/README.md) for what it covers and what
+it genuinely cannot.
 
 ## Rationale
 
 When debugging network issues, it's useful to have a persistent at-a-glance view
 of latency and reachability to multiple well-known internet endpoints. NetWatch
-provides this as a zero-dependency SPA that can be deployed anywhere static
+provides this as a single page that does all its measuring in the browser, so it
-files are served, with no backend required.
+can be served from anywhere static files are served. The backend in its Docker
+image only stores the measurements the page reports; without it, the page works
+the same and nothing is stored.
 
 ## Design
 
-The application is a single-page app built with Vite and Tailwind CSS v4. All
+The page is built with Vite and Tailwind CSS v4. Its code is all in
-code lives in `src/main.js` with a class-based architecture:
+`src/main.js`, with a class-based architecture:
 
-- **`CONFIG`**: Frozen configuration object (update interval, timeouts, axis
+- **`CONFIG`**: Configuration object (update interval, timeouts, axis ticks,
-  ticks, etc.)
+  etc.). The interval menu sets `updateInterval`, the one value the page writes
+  into `CONFIG`; the timeouts, the time the history spans and the x-axis ticks
+  are computed from it
 - **`HostState`**: Per-host state management — history buffer, latency tracking,
   status transitions
 - **`AppState`**: Top-level state container — WAN hosts, local hosts, pause
@@ -104,9 +129,12 @@ code lives in `src/main.js` with a class-based architecture:
   color-coded line segments, error regions, and DPR-aware scaling
 - **UI functions**: `buildUI()` constructs the DOM, `updateHostRow()` /
   `updateSummary()` / `updateHealthBox()` handle incremental updates
-- **`tick()`**: Main loop — measures all hosts in parallel via `Promise.all`,
+- **`tick()`**: Main loop — measures all hosts in parallel, pushing each host's
-  pushes samples, redraws UI. When paused, pushes blank markers (no probes, no
+  sample and redrawing its row as soon as its check ends, then redraws every
-  false outage)
+  row, the summary and the health box once the last check ends. The first round,
+  after loading or an interval change, is discarded. The rows are sorted when
+  the last check ends in round 2, the first one kept, and in rounds 11, 21, 31
+  and so on. When paused, pushes blank markers (no probes, no false outage)
 - **`Reporter`**: Posts collected samples to the backend
 
 ### Reporting
@@ -120,12 +148,35 @@ delivered report, and while paused nothing is sent. Delivery failure is quiet
 one debug-log line per outage, retried at the next interval, never blocking
 probing. The report-building step is a pure function of host state.
 
+### Backend
+
+`netwatch-server`, in `backend/`, is a small Go HTTP server that stores the
+reports the page posts. It keeps them in memory and writes them to `DATA_DIR` as
+zstd-compressed files of JSON lines: every minute, whenever 10 MiB are waiting,
+and when it stops. Its routes:
+
+- `POST /api/v1/reports` — takes a report, without credentials; each client
+  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
+- `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
+
+In the image, the `builder` stage of `Dockerfile` tests it and builds it with
+`backend/script/build`, and `bin/entrypoint.sh` runs it as user `netwatch` on
+`127.0.0.1:8081`, behind nginx. Outside the image, `make run` in `backend/`
+builds it and runs it on port 8080. Its settings, report storage and limits are
+in [backend/README.md](backend/README.md).
+
 ### Monitoring targets
 
-- **22 WAN hosts**: datavi.be, Anthropic API, OpenAI API, AWS Console, GCP
+- **26 WAN hosts**: datavi.be (pinned at start), Anthropic API, OpenAI API, AWS
-  Console, Azure, Cloudflare, Fastly, Akamai, GitHub, B2, 7 S3 regional
+  Console, Google Cloud Console, Microsoft Azure, Cloudflare, Fastly CDN,
-  endpoints (Cape Town, London, Bahrain, Tokyo, Sydney, Oregon, São Paulo), 4
+  Akamai, Google, GitHub, B2, 8 S3 regional endpoints (Cape Town, London,
-  GCS locational endpoints (Iowa, Belgium, Singapore, Sydney)
+  Bahrain, Tokyo, Singapore, Sydney, Oregon, São Paulo) and 6 Hetzner speed test
+  servers (Nuremberg, Falkenstein, Helsinki, Ashburn, Hillsboro, Singapore)
 - **Local CPE**: Cable modem at 192.168.100.1 (always monitored)
 - **Local Gateway**: Auto-detected on startup by probing common default gateway
   addresses (192.168.1.1, 192.168.0.1, 192.168.8.1, 10.0.0.1); first responder
@@ -137,15 +188,17 @@ Local hosts are tracked separately from WAN stats.
 
 ### Latency measurement
 
-HEAD requests with `mode: 'no-cors'` and `cache: 'no-store'`, timed with
+GET requests with `mode: 'no-cors'`, `cache: 'no-store'` and a cache-busting
-`performance.now()`. Each check times out after 80% of the refresh interval (24
+query parameter, timed with `performance.now()`. Each check times out after 80%
-seconds at 30 seconds) and is then recorded as a timeout, so a round's checks
+of the refresh interval (24 seconds at 30 seconds) and is then recorded as a
-have all finished before the next round is due. When no WAN host answers, a
+timeout, so a round's checks have all finished before the next round is due.
-recovery probe checks 4 random WAN hosts every half second, giving up the checks
+When no WAN host answers, a recovery probe checks 4 WAN hosts, picked at random
-it started half a second before. As soon as one answers, a new round starts at
+when it starts, every half second, giving up the checks it started half a second
-once, as it does after an interval change. A round started early gives up the
+before. As soon as one answers, a new round starts at once, as it does after an
-last round's checks if they are still waiting, and that round records nothing,
+interval change. A round started early gives up the last round's checks if they
-so rounds never overlap. IPv4 only.
+are still waiting, and that round records nothing more, so rounds never overlap.
+The browser chooses between IPv4 and IPv6 for each target, as for any request;
+the local targets are IPv4 addresses.
 
 ### Color coding
 
@@ -170,27 +223,44 @@ dist/
 
 ## Features
 
-- Real-time monitoring with 2s update interval and 300s history sparklines
+- A round of checks every 3 seconds by default; the interval menu sets 1, 2, 3,
-- Health indicator: green (HEALTHY) or red (DEGRADED) based on WAN reachability
+  5, 10, 15, 30 or 60 seconds and clears the history
-- Summary stats: reachable count, min/max/avg latency across WAN hosts only
+- Sparklines of each target's last 100 rounds: 300 seconds at 3 seconds
-- Fixed chart axes: Y-axis 0–1000ms, X-axis 0–300s
+- The first round after loading or an interval change is discarded, as DNS and
+  TLS setup inflate its latencies
+- Health indicator from the WAN hosts' latest results: OFFLINE (red) when more
+  than 10 fail and at most 4 answer, otherwise DEGRADED (orange) when more than
+  4 fail, otherwise SLOW (yellow) when more than 3 take over 1000ms, otherwise
+  HEALTHY (green)
+- Summary stats across WAN hosts only: how many answered, the min, median,
+  average and max of their latest latencies, the min and max over the whole
+  history, and the number of rounds run (`Checks`)
+- Fixed chart axes: Y-axis 0–1000ms, higher latencies drawn at the top; X-axis
+  the time the history spans
 - Color-coded latency figures and sparkline line segments
+- WAN host rows sorted by latest latency, unreachable last; pinned rows stay on
+  top, in name order
 - Play/pause: pause stops probes but history keeps scrolling (blank gaps, no
   false outage)
+- Debug log panel, behind a checkbox in the footer, with five levels (error,
+  warning, notice, info, debug) and the last 1000 lines
+- Local and UTC clocks
 - Clickable service URLs
+- A footer link to the commit the page was built from
 - Canvas-based sparkline rendering with devicePixelRatio scaling
 - Zero runtime dependencies: all resources bundled into build artifacts
 
 ## Deployment
 
-After running `yarn build`, deploy the contents of the `dist/` directory to any
+`make build` writes the page to `dist/`, which any static file host (S3, GCS,
-static file host (S3, GCS, Cloudflare Pages, Vercel, Netlify, GitHub Pages) or
+Cloudflare Pages, Vercel, Netlify, GitHub Pages) can serve; with no backend
-use the Docker image behind a reverse proxy.
+there, its reports fail quietly and nothing is stored. Or run the Docker image
+behind a reverse proxy.
 
-The Docker image, built from `Dockerfile`, is the whole service in one
+The Docker image, built from `Dockerfile` by `make docker`, is the whole service
-container: nginx serves the built frontend and passes `/api/` and
+in one container: nginx serves the built frontend and passes `/api/`,
-`/.well-known/healthcheck` to the Go backend, `netwatch-server`, which listens
+`/.well-known/healthcheck` and `/metrics` to the Go backend, `netwatch-server`,
-only inside the container, on `127.0.0.1:8081`. The image:
+which listens only inside the container, on `127.0.0.1:8081`. The image:
 
 - Listens on port 8080 by default (override with `PORT` env var)
 - Takes the client address from `X-Forwarded-For` only on requests from the
@@ -238,6 +308,16 @@ What the [upaas](https://git.eeqj.de/sneak/upaas) app for netwatch needs:
       connects from one can write its own `X-Forwarded-For`, and through a port
       Docker publishes, every client may connect from the Docker network's
       gateway, such as `172.17.0.1`.
+    - `METRICS_USERNAME` and `METRICS_PASSWORD`, default empty: with both set,
+      the backend records Prometheus metrics of its requests and serves them at
+      `/metrics` on the container port, to requests with this user name and
+      password as their basic auth credentials. With neither set, there are no
+      metrics and `/metrics` is not found. One set without the other, or a user
+      name containing `:`, stops the container
+    - `SENTRY_DSN`, default empty: set to a Sentry project's DSN, the backend
+      sends its errors to that Sentry project: each request whose handling
+      crashes, which still gets a 500 response. A value Sentry does not accept
+      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
@@ -251,21 +331,19 @@ properties.
 
 ## Limitations
 
-- **CORS**: Some hosts may block cross-origin HEAD requests. The app uses
+- **CORS**: The checks are cross-origin requests in `no-cors` mode, so the page
-  `no-cors` mode which allows the request but provides opaque responses. Latency
+  cannot read the answer, only time it: any answer counts as reachable, an error
-  is still measurable based on request timing.
+  page included.
-- **Local gateway**: The 192.168.100.1 endpoint requires the host to be
+- **Local targets**: The cable modem at 192.168.100.1 and the detected gateway
-  accessible from your network.
+  answer only on a network that has them, and only when NetWatch is served from
+  localhost or a private address (see Monitoring targets).
 - **Network conditions**: Measurements reflect browser-to-endpoint latency,
   which includes your local network, ISP, and internet routing.
 
 ## TODO
 
-- Add unit tests
+The to-do list is [TODO.md](TODO.md): where the work stands, the next step, the
-- Add eslint for JS linting (currently lint target runs prettier only)
+open work, and what has been done.
-- Add configurable host list (environment variable or config file)
-- Add latency history export (CSV/JSON)
-- Add notification/alert when status changes to DEGRADED
 
 ## License
 

TODO.md

@@ -1,36 +1,163 @@
 # Workflow
 
-- branch (from `main`)
+- branch from `next`
 - do the work in Next Step
 - move Next Step to the top of Completed Steps
 - move the top item of Future Steps into Next Step
 - commit (`TODO.md` changes in the same commit as the work)
-- merge to `main` if the branch is not protected, otherwise open a PR
+- push the branch and open a PR against `next`
-- push
 
 # Status
 
-pre-1.0. No git tags. `feat/reportbuf-storage` is merged; the backend, the CI
+pre-1.0. No git tags. `main` is the stable branch and `next` the development
-workflow, and the backend repo standard files are all on `main`. Frontend and
+branch, which every PR targets. The frontend and the Go backend ship as one
-backend are both functional. Working toward the 1.0.0 milestone by closing the
+Docker image, and the Gitea workflow `.gitea/workflows/check.yml` runs
-remaining repo-compliance issues on the tracker.
+`script/cibuild` on every push. Working toward 1.0.0.
 
 # Next Step
 
-Confirm the `.gitea/workflows/check.yml` run is green (main always green
+Decide whether the repo moves to the layout `REPO_POLICIES.md` gives, with
-policy). The workflow file is already on `main`; what is unverified is that its
+`backend/` no longer repeating files from the root
-latest run passes.
+([#30](https://git.eeqj.de/sneak/netwatch/issues/30)).
 
 # Completed Steps
 
+- 2026-10-04: the page's footer no longer says "IPv4 only"
+  ([#111](https://git.eeqj.de/sneak/netwatch/issues/111)): each check is a
+  `fetch`, the browser picks IPv4 or IPv6 for each WAN host, and the local
+  targets are IPv4 addresses. The rest of the footer is unchanged
+- 2026-10-04: `README.md`, `TODO.md` and `test/viewport/README.md` say what the
+  tree does (issue #24). The README's Getting Started leads with `make` targets;
+  a new Backend section says what `netwatch-server` stores, its routes and how
+  the image builds and runs it, and points to `backend/README.md` for its
+  settings; the checks are GET requests; the 26 WAN hosts, the four health
+  states, the summary's figures and the features the list lacked are described
+  as the page has them; and its TODO section points here, as does the one in
+  `backend/README.md`, whose open items moved to Future Steps. This file's
+  Workflow branches from `next` and opens the PR against `next`, Status says
+  where the repo stands, and Next Step and Future Steps hold only open work,
+  linked to its issue where one exists. The viewport harness README names Node's
+  test runner, not `vitest`
+- 2026-10-04: in `src/main.js` (issue #102), a target's min, max, median and
+  average latency come from one list of its answers, through the same function
+  the summary's figures use, so the median is written once. The latency color
+  limits are one table in `CONFIG`, read by both the figure's and the
+  sparkline's color. The health thresholds, the debug log's length, the gateway
+  check's timeout, the recovery probe's number of hosts and interval, how often
+  the rows are sorted and the delay before the first sparkline resize are
+  `CONFIG` entries too. A unit test checks the summary's figures. Nothing the
+  page does or shows changed; the footer's color legend still writes the limits
+  out as text
+- 2026-10-04: password guesses at `/metrics` are rate limited (issue #104): each
+  client address, resolved through `TRUSTED_PROXIES` as for reports, may make 60
+  requests to `/metrics` a minute, counted by `go-chi/httprate` apart from its
+  reports; past that it gets 429 and its basic auth credentials are not checked.
+  The limit is a constant in `backend/internal/server/routes.go`. A test uses up
+  one client's allowance on wrong passwords, gets 429 with the right one, and
+  checks that another client behind the same nginx still gets in
+- 2026-10-04: the backend reports errors to Sentry (issue #95). With
+  `SENTRY_DSN` set, it sets up `sentry-go` with the release `netwatch-server-`
+  and its version, reports each panic in a handler through `sentryhttp`, the
+  last of the middleware every request goes through, which panics again so the
+  request still gets the 500 from the panic recovery, and waits up to 2 seconds
+  on shutdown for Sentry to finish sending. A DSN Sentry refuses stops the start
+  with an error naming `SENTRY_DSN`. With it empty, Sentry is not set up and
+  nothing is sent to it
+- 2026-10-04: a target's name and URL and a debug log message show as the
+  characters they are and are never read as HTML (issue #29): a host row escapes
+  the name and URL it writes into its markup, and the debug log sets each line
+  as text. A unit test checks a target whose name and URL hold `<`, `>`, `"`,
+  `&` and `'`. `README.md` no longer calls `CONFIG` frozen: the interval menu
+  sets its `updateInterval`, and the values computed from it follow. `AppState`
+  declares the recovery probe's two properties, the sparkline axis functions
+  lose the parameters they did not use, and the comment on a target's history
+  names both kinds of entry it holds. Nothing the page does changed
+- 2026-10-04: a dependency can be added without running yarn or go by hand
+  (issue #45): `make add-dependency PACKAGE=<name>@<version>` shims to the new
+  `script/add-dependency`, which runs `yarn add --dev`, so `package.json` and
+  `yarn.lock` change together, then `yarn install --frozen-lockfile`; the same
+  command moves a package to another version. `make tidy` shims to the new
+  `script/tidy`, which runs `go mod tidy` in `backend/`: a Go module is added by
+  importing it, or moved by editing its `require` line, then `make tidy`.
+  `script/bootstrap` still installs with `--frozen-lockfile`
+- 2026-10-04: the backend serves Prometheus metrics (issue #94). With
+  `METRICS_USERNAME` and `METRICS_PASSWORD` both set, it records request
+  duration and response size through `go-http-metrics` and serves them, with
+  Go's runtime and process metrics, at `GET /metrics` behind basic auth with
+  those credentials; nginx passes `/metrics` to it as it does `/api/`. With
+  neither set there are no metrics and `/metrics` is 404; one without the other
+  stops the start with an error naming both, and so does a `METRICS_USERNAME`
+  containing `:`, with an error naming it. Only requests that reach the health
+  check or `POST /api/v1/reports` are recorded, not `/metrics` itself and not
+  every request as `GO_HTTP_SERVER_CONVENTIONS.md` shows, because the labels are
+  the request's path and method, which clients can make up without end. For
+  that, `POST /api/v1/reports` is now registered by its full path instead of
+  inside a `/api/v1` route group; it answers as before
+- 2026-10-04: `script/` and `Makefile` follow the org models (issue #28):
+  `make dev` shims to the new `script/dev`, the Vite dev server, and the new
+  `make build` to `script/build`, the frontend production build.
+  `.prettierignore` no longer leaves out `backend/`, so `make fmt` and
+  `make fmt-check` cover `backend/README.md`; it leaves out
+  `backend/.golangci.yml` by name, the org standard file whose sha256
+  `backend/script/lint` checks. `script/install-precommit` and the date on
+  `script/bootstrap`'s pins are the org model again; `script/bootstrap`,
+  `script/fmt` and `script/fmt-check` each say in a comment why they differ from
+  it
+- 2026-10-04: a frontend build on Node 26 or newer, such as `make test` on a
+  host with Node 26, no longer prints Node's warning that `module.register()` is
+  deprecated (issue #32); the build in `Dockerfile` runs on Node 22, which never
+  printed it. The call was in `@tailwindcss/node`, which `@tailwindcss/vite`
+  brings in at its own exact version; tailwind 4.3.1 calls
+  `module.registerHooks()` instead where Node has it. `yarn.lock` now has
+  `@tailwindcss/vite` and `tailwindcss` at 4.3.3, inside the ranges
+  `package.json` already allowed, and tailwind's own dependencies moved with
+  them. The built CSS changes only in how it is written out, in tailwind's
+  Firefox focus-ring rule, which no longer applies to iframes (the page has
+  none, so nothing on it looks different), and in tailwind's default sans-serif
+  font list, which the page does not use: `body` sets a monospace font
+- 2026-10-03: the frontend's unit tests cover what the page computes (issue
+  #21): `humanDuration`, the latency colours of a figure and of a sparkline
+  either side of each boundary, a target's min, max, average and median latency
+  over an empty history, an all-unreachable one and a mixed one, and each of the
+  four health states either side of its thresholds. `package.json` has a `test`
+  script, so `yarn run test` and `npm run test` run them, and
+  `script/frontend-test` runs it: quietly, and if a test fails, again with every
+  test listed, and then fails. `src/main.js` now exports `humanDuration`,
+  `HostState`, `latencyClass` and `latencyHex` for the tests; nothing it does
+  changed
 - 2026-10-03: the backend's logs are one stream (issue #27): fx logs its own
   steps of starting and stopping through the backend's logger, so off a terminal
-  every line from start to stop is JSON, where fx used to write plain text to
+  every line the backend's own logger and fx write is JSON, where fx used to
-  stderr. The backend logs its name, version and architecture once at start. The
+  write plain text to stderr. A config file that is found but cannot be read now
-  health check's uptime keys are now `uptime_seconds` and `uptime_human`; its
+  stops the start with its error, logged as JSON like a bad setting, where it
-  path, content type, `"status":"ok"` and 200 are unchanged. `SENTRY_DSN`,
+  used to end in a Go panic. A test runs the server as a child process and
-  `METRICS_USERNAME` and `METRICS_PASSWORD` are still read and still unused, and
+  checks both. The backend logs its name, version and architecture once at
-  left out of `backend/README.md`, until issues #94 and #95 wire them up
+  start. The health check's uptime keys are now `uptime_seconds` and
+  `uptime_human`; its path, content type, `"status":"ok"` and 200 are unchanged.
+  `SENTRY_DSN`, `METRICS_USERNAME` and `METRICS_PASSWORD` are still read and
+  still unused, and left out of `backend/README.md`, until issues #94 and #95
+  wire them up
+- 2026-10-03: the frontend has a real linter (issue #47, and item 2 of issue
+  #28): `eslint` with its recommended rules, set in `eslint.config.js`, runs in
+  a new `frontend-lint` stage of `Dockerfile`, which the frontend stage waits
+  on, as the builder stage waits on the Go `lint` stage. `script/lint` builds
+  both stages without the cache and runs no linter on the host.
+  `script/frontend-lint` runs eslint where it used to repeat the prettier check
+  that `script/fmt-check` runs on the host, and `script/frontend-check`, run by
+  the frontend stage, is now the tests and the format check. `script/bootstrap`
+  wants node 22.13.0 or newer, as eslint 10 does
+- 2026-10-03: the tap-target check in `make frontend-viewport-test` expects one
+  visible pin button per WAN host row (issue #46), where it expected at least 10
+  of the 26, so pin buttons missing from only some rows now fail it. The host
+  row count the harness gathers, which the `app-rendered` check also reads, now
+  counts only the WAN host rows: the local host rows have no pin button
+- 2026-10-03: each target's row shows its result as soon as its check ends
+  (issue #91), where every row waited for the round's slowest check, up to 24
+  seconds at a 30-second interval. Every row is still redrawn, and sorting, the
+  summary, the health box and offline detection still run, once, when the
+  round's last check ends, so no row reads "paused" after a pause and resume
+  during the round. A check that ends after the user pauses or after its round
+  is given up shows nothing, and the first round is still discarded as a whole
 - 2026-10-03: root no longer acts outside `/data` when it prepares `DATA_DIR`
   (issue #80): `bin/entrypoint.sh` runs `netwatch-server prepare-data-dir`,
   which refuses a `DATA_DIR` that is not `/data` or a path below it written in
@@ -238,9 +365,11 @@ latest run passes.
 - 2026-08-09: automated responsive-layout harness
   (`make frontend-viewport-test`): digest-pinned headless Chrome driven over CDP
   against the built `dist/`, viewport widths derived from the breakpoints in
-  `src/styles.css` ([#13](https://git.eeqj.de/sneak/netwatch/issues/13)). Every
+  `src/styles.css` ([#13](https://git.eeqj.de/sneak/netwatch/issues/13)). The
-  check carries a presence guard so none of them can pass against a page it is
+  tap-target and host-row checks each fail when they measured nothing; the
-  not actually measuring. Found two real layout defects, filed as
+  overflow, viewport-edge and clipped-text checks have no such guard of their
+  own and rely on the `app-rendered` check, which fails the run when the app did
+  not render. Found two real layout defects, filed as
   [#42](https://git.eeqj.de/sneak/netwatch/issues/42) and
   [#43](https://git.eeqj.de/sneak/netwatch/issues/43)
 - 2026-07-07 Adopted scripts-to-rule-them-all: `script/` entrypoints, Makefile
@@ -260,12 +389,14 @@ latest run passes.
 
 # Future Steps
 
-- Wire `script/frontend-viewport-test` into CI as its own step (deliberately not
+- Run `make frontend-viewport-test` in CI as its own step; it is not part of
-  part of `make check` today; the decision has real CI-runtime cost and is
+  `make check`, as it needs Docker and takes minutes
-  tracked separately)
+- A backend test that posts a report to `POST /api/v1/reports` and checks the
-- Compliance top-up as one small commit: add .editorconfig and add the hooks
+  compressed file it is written to
-  target to the Makefile
+- A backend route that decompresses the stored reports and answers queries on
-- After merge, confirm .gitea/workflows/check.yml is on main and CI is green
+  them
-  (main always green policy)
+- Prometheus metrics for the backend's in-memory buffer: its size, the number of
-- Decide what to do with untracked resume.sh: commit it, gitignore it, or delete
+  flushes and the number of reports
-  it
+- 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

backend/README.md

@@ -32,17 +32,16 @@ pattern as the repo root: the targets in `backend/Makefile` are thin shims over
 `test`, `fmt` and `fmt-check`:
 
 - `script/build` — compile the static `netwatch-server` binary with its version
-  stamped in. The version is `VERSION` from the environment;
+  stamped in. The version is `VERSION` from the environment; when that is unset
-  when that is unset or empty, it falls back to `git describe` inside a git
+  or empty, it falls back to `git describe` inside a git checkout, then to `dev`
-  checkout, then to `dev`
 - `script/test` — run the Go tests with the race detector and coverage. Go's
   `-timeout 30s` bounds the tests, not their compile. If they fail, they run
   again with `-v` for the details, and the script fails. The race detector needs
   a C compiler
 - `script/lint` — check `.golangci.yml` against its pinned sha256, then run
-  golangci-lint. It runs inside the golangci-lint image of the lint stage of
+  golangci-lint. It runs inside the golangci-lint image of the lint stage of the
-  the root `Dockerfile`; from a checkout, run `make lint` at the repo root,
+  root `Dockerfile`; from a checkout, run `make lint` at the repo root, which
-  which builds that stage
+  builds that stage
 - `script/fmt` — format the Go sources (writes)
 - `script/fmt-check` — check Go formatting (read-only)
 - `script/run` — build and run the server locally
@@ -61,8 +60,9 @@ flushes them to compressed files on disk for later analysis.
 
 ## Design
 
-The server is structured as an `fx`-wired Go application under `cmd/netwatch-server/`.
+The server is structured as an `fx`-wired Go application under
-Internal packages in `internal/` follow standard Go project layout:
+`cmd/netwatch-server/`. Internal packages in `internal/` follow standard Go
+project layout:
 
 - **`config`**: Loads configuration from environment variables and config files
   via Viper.
@@ -88,12 +88,16 @@ Internal packages in `internal/` follow standard Go project layout:
 | `TRUSTED_PROXIES`      | loopback + RFC1918   | 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)                                  |
+| `METRICS_USERNAME`     | empty                | Basic auth user name for `/metrics`; see [Metrics](#metrics)                                             |
+| `METRICS_PASSWORD`     | empty                | Basic auth password for `/metrics`; see [Metrics](#metrics)                                              |
+| `SENTRY_DSN`           | empty                | DSN of the Sentry project to send errors to; see [Sentry](#sentry)                                       |
 
-`TRUSTED_PROXIES` defaults to `127.0.0.1/32,::1/128,10.0.0.0/8,172.16.0.0/12,192.168.0.0/16`.
+`TRUSTED_PROXIES` defaults to
-The loopback entries cover a reverse proxy on the same host. A request whose
+`127.0.0.1/32,::1/128,10.0.0.0/8,172.16.0.0/12,192.168.0.0/16`. The loopback
-direct peer is outside this set has its forwarded headers ignored, and the
+entries cover a reverse proxy on the same host. A request whose direct peer is
-direct peer is logged and rate-limited instead. The container image does not use
+outside this set has its forwarded headers ignored, and the direct peer is
-this default; see [Container image](#container-image).
+logged and rate-limited instead. The container image does not use this default;
+see [Container image](#container-image).
 
 A variable set to a value the server cannot use, such as `PORT=abc`,
 `DEBUG=maybe` or a `BIND_ADDRESS` that is not an IP address, stops it from
@@ -102,16 +106,16 @@ starting, with an error naming the variable. An empty variable counts as unset.
 ### Container image
 
 The root `Dockerfile` builds one image in which nginx listens on the public port
-8080, serves the frontend, and proxies `/api/` and `/.well-known/healthcheck` to
+8080, serves the frontend, and proxies `/api/`, `/.well-known/healthcheck` and
-this server. The image's entrypoint, `bin/entrypoint.sh`, starts the server as
+`/metrics` to this server. The image's entrypoint, `bin/entrypoint.sh`, starts
-user `netwatch` (uid 1000) with `BIND_ADDRESS=127.0.0.1` and `PORT=8081`, so
+the server as user `netwatch` (uid 1000) with `BIND_ADDRESS=127.0.0.1` and
-only nginx reaches it, and with `TRUSTED_PROXIES=127.0.0.1/32`, so it takes the
+`PORT=8081`, so only nginx reaches it, and with `TRUSTED_PROXIES=127.0.0.1/32`,
-client address nginx passes on and no other. `DATA_DIR` is `/data/reports`, on
+so it takes the client address nginx passes on and no other. `DATA_DIR` is
-the `/data` volume; before starting the server, the entrypoint creates it and
+`/data/reports`, on the `/data` volume; before starting the server, the
-gives it and `/data` to `netwatch` with `netwatch-server prepare-data-dir`,
+entrypoint creates it and gives it and `/data` to `netwatch` with
-which acts on nothing outside `/data`. nginx replaces the security headers
+`netwatch-server prepare-data-dir`, which acts on nothing outside `/data`. nginx
-this server sets with those in the root `security-headers.conf`, so those are
+replaces the security headers this server sets with those in the root
-what clients of the image see.
+`security-headers.conf`, so those are what clients of the image see.
 
 The container's own `TRUSTED_PROXIES` goes to nginx instead: IP addresses or
 CIDRs, separated by commas, of the reverse proxies in front of the container.
@@ -162,8 +166,7 @@ credentials, so it is bounded instead. Both refusals below answer with the same
   to be written fill the cap on their own, and then no file is deleted. At
   start, report files past the cap, as after lowering it, are deleted the same
   way. So the cap is 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
+  GiB is small enough for any host; set it to the space you can give `DATA_DIR`.
-  `DATA_DIR`.
 
 ### CORS
 
@@ -176,11 +179,48 @@ origin, `scheme://host` with an optional `:port`, as browsers send it: no path,
 not even a trailing `/`, and no `*`. Any other entry stops the server from
 starting, with an error naming `CORS_ALLOWED_ORIGINS`.
 
+### Metrics
+
+With both `METRICS_USERNAME` and `METRICS_PASSWORD` set, the server serves
+Prometheus metrics at `GET /metrics` to requests with those as their basic auth
+credentials, and answers any other with 401. For each request that reaches the
+health check or `POST /api/v1/reports`, those the rate limit refuses included,
+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
+for a method its route does not take, or 413 for declaring a body length over
+the 1 MiB limit. A report whose body goes over the limit without declaring its
+length reaches the route, is answered 413 there, and is recorded with that
+status. 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. With neither set, nothing
+is recorded and `/metrics` answers 404. One without the other stops the server
+from starting, with an error naming both; so does a `METRICS_USERNAME`
+containing `:`, which basic auth cannot carry, with an error naming it.
+
+`/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
+minute, whatever their credentials. Past that it gets 429 with
+`Retry-After: 60`, and its credentials are not checked. The minute slides as it
+does for reports (see [Report limits](#report-limits)), so a scraper polling
+every 2 seconds or less often is never refused. This allowance is apart from the
+one for reports.
+
+### Sentry
+
+With `SENTRY_DSN` set, the server sends its errors to that Sentry project: each
+panic in a handler is reported there, under the release `netwatch-server-`
+followed by the server's version, and the request still gets 500 from the
+server's panic recovery. On shutdown the server waits up to 2 seconds for Sentry
+to finish sending. A DSN Sentry refuses stops the server from starting, with an
+error naming `SENTRY_DSN`. With it empty, Sentry is not set up, and nothing is
+sent to it.
+
 ## TODO
 
-- Add integration test that POSTs a report and verifies the compressed output
+The to-do list, this backend's open work included, is [TODO.md](../TODO.md) at
-- Add report decompression/query endpoint
+the repo root.
-- Add metrics (Prometheus) for buffer size, flush count, report count
 
 ## License
 

backend/cmd/netwatch-server/main_test.go

@@ -0,0 +1,249 @@
+package main
+
+import (
+	"bytes"
+	"context"
+	"encoding/json"
+	"net"
+	"net/http"
+	"os"
+	"os/exec"
+	"os/signal"
+	"path/filepath"
+	"strings"
+	"syscall"
+	"testing"
+	"time"
+)
+
+// The tests run main() in a child process, this test binary started
+// again with runMainEnv set, because main() can exit its process and
+// takes its settings from the environment.
+const runMainEnv = "NETWATCH_SERVER_RUN_MAIN"
+
+// childTimeout bounds each child's whole run; it is killed after it.
+const childTimeout = 10 * time.Second
+
+func TestMain(m *testing.M) {
+	if os.Getenv(runMainEnv) != "" {
+		// A SIGTERM that comes before fx catches it is dropped, not fatal.
+		signal.Notify(make(chan os.Signal, 1), syscall.SIGTERM)
+		main()
+
+		return
+	}
+
+	os.Exit(m.Run())
+}
+
+// TestOutputIsJSON: off a terminal, every line the server writes from
+// start to stop is JSON, fx's own lines included.
+func TestOutputIsJSON(t *testing.T) {
+	t.Parallel()
+
+	ctx, cancel := context.WithTimeout(t.Context(), childTimeout)
+	defer cancel()
+
+	port := freePort(ctx, t)
+	child, stdout, stderr := startServer(ctx, t, t.TempDir(), port)
+
+	waitForHealthcheck(ctx, t, port)
+
+	// The child drops a SIGTERM that comes before fx catches it (see
+	// TestMain), so send one every 100ms until the test ends. ctx
+	// bounds the wait: when it ends, the child is killed.
+	stop := make(chan struct{})
+	defer close(stop)
+
+	go func() {
+		for {
+			_ = child.Process.Signal(syscall.SIGTERM)
+
+			select {
+			case <-stop:
+				return
+			case <-time.After(100 * time.Millisecond):
+			}
+		}
+	}()
+
+	err := child.Wait()
+	if err != nil {
+		t.Fatalf("server exit = %v, want success", err)
+	}
+
+	requireJSONLines(t, stdout, stderr)
+
+	if !strings.Contains(stdout.String(), `"msg":"starting"`) {
+		t.Fatalf("no startup line in stdout:\n%s", stdout)
+	}
+}
+
+// TestMalformedConfigFileStopsTheStart: a config file the server finds
+// but cannot read stops the start, and the error is logged as JSON.
+func TestMalformedConfigFileStopsTheStart(t *testing.T) {
+	t.Parallel()
+
+	ctx, cancel := context.WithTimeout(t.Context(), childTimeout)
+	defer cancel()
+
+	home := t.TempDir()
+	dir := filepath.Join(home, ".config", "netwatch-server")
+
+	err := os.MkdirAll(dir, 0o750)
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	err = os.WriteFile(filepath.Join(dir, "netwatch-server.yaml"),
+		[]byte("PORT: [8080\n"), 0o600)
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	child, stdout, stderr := startServer(ctx, t, home, freePort(ctx, t))
+
+	err = child.Wait()
+	if child.ProcessState.ExitCode() != 1 {
+		t.Fatalf("server exit = %v, want exit status 1", err)
+	}
+
+	requireJSONLines(t, stdout, stderr)
+
+	if !strings.Contains(stdout.String(), "netwatch-server.yaml") {
+		t.Fatalf("no error naming the config file in stdout:\n%s", stdout)
+	}
+}
+
+// TestRefusedSentryDSNStopsTheStart: a SENTRY_DSN that Sentry refuses
+// stops the start, and the error, naming SENTRY_DSN, is logged as JSON.
+func TestRefusedSentryDSNStopsTheStart(t *testing.T) {
+	t.Setenv("SENTRY_DSN", "not-a-dsn")
+
+	ctx, cancel := context.WithTimeout(t.Context(), childTimeout)
+	defer cancel()
+
+	child, stdout, stderr := startServer(ctx, t, t.TempDir(), freePort(ctx, t))
+
+	err := child.Wait()
+	if child.ProcessState.ExitCode() != 1 {
+		t.Fatalf("server exit = %v, want exit status 1", err)
+	}
+
+	requireJSONLines(t, stdout, stderr)
+
+	if !strings.Contains(stdout.String(), "SENTRY_DSN") {
+		t.Fatalf("no error naming SENTRY_DSN in stdout:\n%s", stdout)
+	}
+}
+
+// startServer runs main() in a child process listening on
+// 127.0.0.1:port, with home as its HOME and working directory and its
+// data directory in home, so it touches nothing outside home. Its
+// stdout and stderr go to the two buffers returned, which hold all of
+// it once child.Wait returns. The child is killed when ctx ends, and
+// killed and reaped when the test ends if nothing waited for it.
+func startServer(
+	ctx context.Context,
+	t *testing.T,
+	home, port string,
+) (*exec.Cmd, *bytes.Buffer, *bytes.Buffer) {
+	t.Helper()
+
+	self, err := os.Executable()
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	var stdout, stderr bytes.Buffer
+
+	child := exec.CommandContext(ctx, self) //nolint:gosec // this test binary
+	child.Dir = home
+	child.Env = append(os.Environ(),
+		runMainEnv+"=1",
+		"HOME="+home,
+		"DATA_DIR="+filepath.Join(home, "data"),
+		"BIND_ADDRESS=127.0.0.1",
+		"PORT="+port,
+	)
+	child.Stdout = &stdout
+	child.Stderr = &stderr
+
+	err = child.Start()
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	t.Cleanup(func() {
+		if child.ProcessState == nil {
+			_ = child.Process.Kill()
+			_ = child.Wait()
+		}
+	})
+
+	return child, &stdout, &stderr
+}
+
+// freePort returns a TCP port on 127.0.0.1 that was free a moment ago.
+func freePort(ctx context.Context, t *testing.T) string {
+	t.Helper()
+
+	var lc net.ListenConfig
+
+	l, err := lc.Listen(ctx, "tcp", "127.0.0.1:0")
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	_ = l.Close()
+
+	_, port, err := net.SplitHostPort(l.Addr().String())
+	if err != nil {
+		t.Fatal(err)
+	}
+
+	return port
+}
+
+// waitForHealthcheck returns once the health check on port answers 200.
+func waitForHealthcheck(ctx context.Context, t *testing.T, port string) {
+	t.Helper()
+
+	url := "http://127.0.0.1:" + port + "/.well-known/healthcheck"
+
+	for {
+		req, err := http.NewRequestWithContext(ctx, http.MethodGet, url, nil)
+		if err != nil {
+			t.Fatal(err)
+		}
+
+		resp, err := http.DefaultClient.Do(req)
+		if err == nil {
+			_ = resp.Body.Close()
+
+			if resp.StatusCode == http.StatusOK {
+				return
+			}
+		}
+
+		select {
+		case <-ctx.Done():
+			t.Fatalf("health check never answered: %v", err)
+		case <-time.After(50 * time.Millisecond):
+		}
+	}
+}
+
+// requireJSONLines fails the test on each line of outs that is not
+// JSON.
+func requireJSONLines(t *testing.T, outs ...*bytes.Buffer) {
+	t.Helper()
+
+	for _, out := range outs {
+		for line := range strings.Lines(out.String()) {
+			if !json.Valid([]byte(line)) {
+				t.Errorf("line is not JSON: %s", line)
+			}
+		}
+	}
+}

backend/go.mod

@@ -3,20 +3,30 @@ module sneak.berlin/go/netwatch
 go 1.25.5
 
 require (
+	github.com/99designs/basicauth-go v0.0.0-20230316000542-bf6f9cbbf0f8
+	github.com/getsentry/sentry-go v0.49.0
 	github.com/go-chi/chi/v5 v5.2.5
 	github.com/go-chi/cors v1.2.2
 	github.com/go-chi/httprate v0.16.0
 	github.com/joho/godotenv v1.5.1
-	github.com/klauspost/compress v1.18.4
+	github.com/klauspost/compress v1.19.1
+	github.com/prometheus/client_golang v1.24.1
+	github.com/slok/go-http-metrics v0.13.0
 	github.com/spf13/viper v1.21.0
 	go.uber.org/fx v1.24.0
 )
 
 require (
+	github.com/beorn7/perks v1.0.1 // indirect
+	github.com/cespare/xxhash/v2 v2.3.0 // indirect
 	github.com/fsnotify/fsnotify v1.9.0 // indirect
 	github.com/go-viper/mapstructure/v2 v2.4.0 // indirect
 	github.com/klauspost/cpuid/v2 v2.2.10 // indirect
+	github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822 // indirect
 	github.com/pelletier/go-toml/v2 v2.2.4 // indirect
+	github.com/prometheus/client_model v0.6.2 // indirect
+	github.com/prometheus/common v0.70.1 // indirect
+	github.com/prometheus/procfs v0.21.1 // indirect
 	github.com/sagikazarmark/locafero v0.11.0 // indirect
 	github.com/sourcegraph/conc v0.3.1-0.20240121214520-5f936abd7ae8 // indirect
 	github.com/spf13/afero v1.15.0 // indirect
@@ -28,6 +38,7 @@ require (
 	go.uber.org/multierr v1.10.0 // indirect
 	go.uber.org/zap v1.26.0 // indirect
 	go.yaml.in/yaml/v3 v3.0.4 // indirect
-	golang.org/x/sys v0.30.0 // indirect
+	golang.org/x/sys v0.47.0 // indirect
-	golang.org/x/text v0.28.0 // indirect
+	golang.org/x/text v0.40.0 // indirect
+	google.golang.org/protobuf v1.36.11 // indirect
 )

backend/go.sum

@@ -1,37 +1,65 @@
-github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
+github.com/99designs/basicauth-go v0.0.0-20230316000542-bf6f9cbbf0f8 h1:nMpu1t4amK3vJWBibQ5X/Nv0aXL+b69TQf2uK5PH7Go=
-github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
+github.com/99designs/basicauth-go v0.0.0-20230316000542-bf6f9cbbf0f8/go.mod h1:3cARGAK9CfW3HoxCy1a0G4TKrdiKke8ftOMEOHyySYs=
+github.com/beorn7/perks v1.0.1 h1:VlbKKnNfV8bJzeqoa4cOKqO6bYr3WgKZxO8Z16+hsOM=
+github.com/beorn7/perks v1.0.1/go.mod h1:G2ZrVWU2WbWT9wwq4/hrbKbnv/1ERSJQ0ibhJ6rlkpw=
+github.com/cespare/xxhash/v2 v2.3.0 h1:UL815xU9SqsFlibzuggzjXhog7bL6oX9BbNZnL2UFvs=
+github.com/cespare/xxhash/v2 v2.3.0/go.mod h1:VGX0DQ3Q6kWi7AoAeZDth3/j3BFtOZR5XLFGgcrjCOs=
+github.com/davecgh/go-spew v1.1.2-0.20180830191138-d8f796af33cc h1:U9qPSI2PIWSS1VwoXQT9A3Wy9MM3WgvqSxFWenqJduM=
+github.com/davecgh/go-spew v1.1.2-0.20180830191138-d8f796af33cc/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/frankban/quicktest v1.14.6 h1:7Xjx+VpznH+oBnejlPUj8oUpdxnVs4f8XU8WnHkI4W8=
 github.com/frankban/quicktest v1.14.6/go.mod h1:4ptaffx2x8+WTWXmUCuVU6aPUX1/Mz7zb5vbUoiM6w0=
 github.com/fsnotify/fsnotify v1.9.0 h1:2Ml+OJNzbYCTzsxtv8vKSFD9PbJjmhYF14k/jKC7S9k=
 github.com/fsnotify/fsnotify v1.9.0/go.mod h1:8jBTzvmWwFyi3Pb8djgCCO5IBqzKJ/Jwo8TRcHyHii0=
+github.com/getsentry/sentry-go v0.49.0 h1:Ehejknu1l023Ub7QoRBVLAI7g3Jnhqku4oWx4B4Sh5s=
+github.com/getsentry/sentry-go v0.49.0/go.mod h1:nuMJAoCfe1u0Bts2ocyNI+TW8HT84vRMqwA5Qq/SKUI=
 github.com/go-chi/chi/v5 v5.2.5 h1:Eg4myHZBjyvJmAFjFvWgrqDTXFyOzjj7YIm3L3mu6Ug=
 github.com/go-chi/chi/v5 v5.2.5/go.mod h1:X7Gx4mteadT3eDOMTsXzmI4/rwUpOwBHLpAfupzFJP0=
 github.com/go-chi/cors v1.2.2 h1:Jmey33TE+b+rB7fT8MUy1u0I4L+NARQlK6LhzKPSyQE=
 github.com/go-chi/cors v1.2.2/go.mod h1:sSbTewc+6wYHBBCW7ytsFSn836hqM7JxpglAy2Vzc58=
 github.com/go-chi/httprate v0.16.0 h1:8V5DH9j6pSK6UQoBsTpvMyFxycqaKEIToyPKzHJjUa8=
 github.com/go-chi/httprate v0.16.0/go.mod h1:A8lo+qRhk+s9LiuP5saS7XCGDXRXMcrueq0NfIuCa/I=
+github.com/go-errors/errors v1.4.2 h1:J6MZopCL4uSllY1OfXM374weqZFFItUbrImctkmUxIA=
+github.com/go-errors/errors v1.4.2/go.mod h1:sIVyrIiJhuEF+Pj9Ebtd6P/rEYROXFi3BopGUQ5a5Og=
 github.com/go-viper/mapstructure/v2 v2.4.0 h1:EBsztssimR/CONLSZZ04E8qAkxNYq4Qp9LvH92wZUgs=
 github.com/go-viper/mapstructure/v2 v2.4.0/go.mod h1:oJDH3BJKyqBA2TXFhDsKDGDTlndYOZ6rGS0BRZIxGhM=
-github.com/google/go-cmp v0.6.0 h1:ofyhxvXcZhMsU5ulbFiLKl/XBFqE1GSq7atu8tAmTRI=
+github.com/google/go-cmp v0.7.0 h1:wk8382ETsv4JYUZwIsn6YpYiWiBsYLSJiTsyBybVuN8=
-github.com/google/go-cmp v0.6.0/go.mod h1:17dUlkBOakJ0+DkrSSNjCkIjxS6bF9zb3elmeNGIjoY=
+github.com/google/go-cmp v0.7.0/go.mod h1:pXiqmnSA92OHEEa9HXL2W4E7lf9JzCmGVUdgjX3N/iU=
 github.com/joho/godotenv v1.5.1 h1:7eLL/+HRGLY0ldzfGMeQkb7vMd0as4CfYvUVzLqw0N0=
 github.com/joho/godotenv v1.5.1/go.mod h1:f4LDr5Voq0i2e/R5DDNOoa2zzDfwtkZa6DnEwAbqwq4=
-github.com/klauspost/compress v1.18.4 h1:RPhnKRAQ4Fh8zU2FY/6ZFDwTVTxgJ/EMydqSTzE9a2c=
+github.com/klauspost/compress v1.19.1 h1:VsB4HPswih7mmZ8WleSFQ75c/Ui1M4trX5oAsJnhSlk=
-github.com/klauspost/compress v1.18.4/go.mod h1:R0h/fSBs8DE4ENlcrlib3PsXS61voFxhIs2DeRhCvJ4=
+github.com/klauspost/compress v1.19.1/go.mod h1:cwPg85FWrGar70rWktvGQj8/hthj3wpl0PGDogxkrSQ=
 github.com/klauspost/cpuid/v2 v2.2.10 h1:tBs3QSyvjDyFTq3uoc/9xFpCuOsJQFNPiAhYdw2skhE=
 github.com/klauspost/cpuid/v2 v2.2.10/go.mod h1:hqwkgyIinND0mEev00jJYCxPNVRVXFQeu1XKlok6oO0=
 github.com/kr/pretty v0.3.1 h1:flRD4NNwYAUpkphVc1HcthR4KEIFJ65n8Mw5qdRn3LE=
 github.com/kr/pretty v0.3.1/go.mod h1:hoEshYVHaxMs3cyo3Yncou5ZscifuDolrwPKZanG3xk=
 github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
 github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
+github.com/kylelemons/godebug v1.1.0 h1:RPNrshWIDI6G2gRW9EHilWtl7Z6Sb1BR0xunSBf0SNc=
+github.com/kylelemons/godebug v1.1.0/go.mod h1:9/0rRGxNHcop5bhtWyNeEfOS8JIWk580+fNqagV/RAw=
+github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822 h1:C3w9PqII01/Oq1c1nUAm88MOHcQC9l5mIlSMApZMrHA=
+github.com/munnerz/goautoneg v0.0.0-20191010083416-a7dc8b61c822/go.mod h1:+n7T8mK8HuQTcFwEeznm/DIxMOiR9yIdICNftLE1DvQ=
 github.com/pelletier/go-toml/v2 v2.2.4 h1:mye9XuhQ6gvn5h28+VilKrrPoQVanw5PMw/TB0t5Ec4=
 github.com/pelletier/go-toml/v2 v2.2.4/go.mod h1:2gIqNv+qfxSVS7cM2xJQKtLSTLUE9V8t9Stt+h56mCY=
-github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
+github.com/pingcap/errors v0.11.4 h1:lFuQV/oaUMGcD2tqt+01ROSmJs75VG1ToEOkZIZ4nE4=
-github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
+github.com/pingcap/errors v0.11.4/go.mod h1:Oi8TUi2kEtXXLMJk9l1cGmz20kV3TaQ0usTwv5KuLY8=
-github.com/rogpeppe/go-internal v1.9.0 h1:73kH8U+JUqXU8lRuOHeVHaa/SZPifC7BkcraZVejAe8=
+github.com/pkg/errors v0.9.1 h1:FEBLx1zS214owpjy7qsBeixbURkuhQAwrK5UwLGTwt4=
-github.com/rogpeppe/go-internal v1.9.0/go.mod h1:WtVeX8xhTBvf0smdhujwtBcq4Qrzq/fJaraNFVN+nFs=
+github.com/pkg/errors v0.9.1/go.mod h1:bwawxfHBFNV+L2hUp1rHADufV3IMtnDRdf1r5NINEl0=
+github.com/pmezard/go-difflib v1.0.1-0.20181226105442-5d4384ee4fb2 h1:Jamvg5psRIccs7FGNTlIRMkT8wgtp5eCXdBlqhYGL6U=
+github.com/pmezard/go-difflib v1.0.1-0.20181226105442-5d4384ee4fb2/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
+github.com/prometheus/client_golang v1.24.1 h1:JnJkREXzWxUdCuPFpIWZiPispT9xVV59uiuyR2bPlnU=
+github.com/prometheus/client_golang v1.24.1/go.mod h1:F+oSRECHg4sse5ucfYpYDeIv/hu68Zo0uoHKetWnzcE=
+github.com/prometheus/client_model v0.6.2 h1:oBsgwpGs7iVziMvrGhE53c/GrLUsZdHnqNwqPLxwZyk=
+github.com/prometheus/client_model v0.6.2/go.mod h1:y3m2F6Gdpfy6Ut/GBsUqTWZqCUvMVzSfMLjcu6wAwpE=
+github.com/prometheus/common v0.70.1 h1:1HvjP4D5oL3t8RsPlwxA9onvvStjtIHYE5XuuwOi/PY=
+github.com/prometheus/common v0.70.1/go.mod h1:VdFUQDMZK3VLkurFUVhia6uys/0suUp86TJz5qbJRhc=
+github.com/prometheus/procfs v0.21.1 h1:GljZCt+zSTS+NZq88cyQ1LjZ+RCHp3uVuabBWA5+OJI=
+github.com/prometheus/procfs v0.21.1/go.mod h1:aB55Cww9pdSJVHk0hUf0inxWyyjPogFIjmHKYgMKmtY=
+github.com/rogpeppe/go-internal v1.14.1 h1:UQB4HGPB6osV0SQTLymcB4TgvyWu6ZyliaW0tI/otEQ=
+github.com/rogpeppe/go-internal v1.14.1/go.mod h1:MaRKkUm5W0goXpeCfT7UZI6fk/L7L7so1lCWt35ZSgc=
 github.com/sagikazarmark/locafero v0.11.0 h1:1iurJgmM9G3PA/I+wWYIOw/5SyBtxapeHDcg+AAIFXc=
 github.com/sagikazarmark/locafero v0.11.0/go.mod h1:nVIGvgyzw595SUSUE6tvCp3YYTeHs15MvlmU87WwIik=
+github.com/slok/go-http-metrics v0.13.0 h1:lQDyJJx9wKhmbliyUsZ2l6peGnXRHjsjoqPt5VYzcP8=
+github.com/slok/go-http-metrics v0.13.0/go.mod h1:HIr7t/HbN2sJaunvnt9wKP9xoBBVZFo1/KiHU3b0w+4=
 github.com/sourcegraph/conc v0.3.1-0.20240121214520-5f936abd7ae8 h1:+jumHNA0Wrelhe64i8F6HNlS8pkoyMv5sreGx2Ry5Rw=
 github.com/sourcegraph/conc v0.3.1-0.20240121214520-5f936abd7ae8/go.mod h1:3n1Cwaq1E1/1lhQhtRK2ts/ZwZEhjcQeJQ1RuC6Q/8U=
 github.com/spf13/afero v1.15.0 h1:b/YBCLWAJdFWJTN9cLhiXXcD7mzKn9Dm86dNnfyQw1I=
@@ -42,6 +70,8 @@ github.com/spf13/pflag v1.0.10 h1:4EBh2KAYBwaONj6b2Ye1GiHfwjqyROoF4RwYO+vPwFk=
 github.com/spf13/pflag v1.0.10/go.mod h1:McXfInJRrz4CZXVZOBLb0bTZqETkiAhM9Iw0y3An2Bg=
 github.com/spf13/viper v1.21.0 h1:x5S+0EU27Lbphp4UKm1C+1oQO+rKx36vfCoaVebLFSU=
 github.com/spf13/viper v1.21.0/go.mod h1:P0lhsswPGWD/1lZJ9ny3fYnVqxiegrlNrEmgLjbTCAY=
+github.com/stretchr/objx v0.5.2 h1:xuMeJ0Sdp5ZMRXx/aWO6RZxdr3beISkG5/G/aIRr3pY=
+github.com/stretchr/objx v0.5.2/go.mod h1:FRsXN1f5AsAjCGJKqEizvkpNtU+EGNCLh3NxZ/8L+MA=
 github.com/stretchr/testify v1.11.1 h1:7s2iGBzp5EwR7/aIZr8ao5+dra3wiQyKjjFuvgVKu7U=
 github.com/stretchr/testify v1.11.1/go.mod h1:wZwfW3scLgRK+23gO65QZefKpKQRnfz6sD981Nm4B6U=
 github.com/subosito/gotenv v1.6.0 h1:9NlTDc1FTs4qu0DDq7AEtTPNw6SVm7uBMsUCUjABIf8=
@@ -54,20 +84,24 @@ go.uber.org/dig v1.19.0 h1:BACLhebsYdpQ7IROQ1AGPjrXcP5dF80U3gKoFzbaq/4=
 go.uber.org/dig v1.19.0/go.mod h1:Us0rSJiThwCv2GteUN0Q7OKvU7n5J4dxZ9JKUXozFdE=
 go.uber.org/fx v1.24.0 h1:wE8mruvpg2kiiL1Vqd0CC+tr0/24XIB10Iwp2lLWzkg=
 go.uber.org/fx v1.24.0/go.mod h1:AmDeGyS+ZARGKM4tlH4FY2Jr63VjbEDJHtqXTGP5hbo=
-go.uber.org/goleak v1.2.0 h1:xqgm/S+aQvhWFTtR0XK3Jvg7z8kGV8P4X14IzwN3Eqk=
+go.uber.org/goleak v1.3.0 h1:2K3zAYmnTNqV73imy9J1T3WC+gmCePx2hEGkimedGto=
-go.uber.org/goleak v1.2.0/go.mod h1:XJYK+MuIchqpmGmUSAzotztawfKvYLUIgg7guXrwVUo=
+go.uber.org/goleak v1.3.0/go.mod h1:CoHD4mav9JJNrW/WLlf7HGZPjdw8EucARQHekz1X6bE=
 go.uber.org/multierr v1.10.0 h1:S0h4aNzvfcFsC3dRF1jLoaov7oRaKqRGC/pUEJ2yvPQ=
 go.uber.org/multierr v1.10.0/go.mod h1:20+QtiLqy0Nd6FdQB9TLXag12DsQkrbs3htMFfDN80Y=
 go.uber.org/zap v1.26.0 h1:sI7k6L95XOKS281NhVKOFCUNIvv9e0w4BF8N3u+tCRo=
 go.uber.org/zap v1.26.0/go.mod h1:dtElttAiwGvoJ/vj4IwHBS/gXsEu/pZ50mUIRWuG0so=
+go.yaml.in/yaml/v2 v2.4.4 h1:tuyd0P+2Ont/d6e2rl3be67goVK4R6deVxCUX5vyPaQ=
+go.yaml.in/yaml/v2 v2.4.4/go.mod h1:gMZqIpDtDqOfM0uNfy0SkpRhvUryYH0Z6wdMYcacYXQ=
 go.yaml.in/yaml/v3 v3.0.4 h1:tfq32ie2Jv2UxXFdLJdh3jXuOzWiL1fo0bu/FbuKpbc=
 go.yaml.in/yaml/v3 v3.0.4/go.mod h1:DhzuOOF2ATzADvBadXxruRBLzYTpT36CKvDb3+aBEFg=
-golang.org/x/sys v0.30.0 h1:QjkSwP/36a20jFYWkSue1YwXzLmsV5Gfq7Eiy72C1uc=
+golang.org/x/sys v0.47.0 h1:o7XGOvZQCADBQQ4Y7VNq2dRWQR7JmOUW8Kxx4ZsNgWs=
-golang.org/x/sys v0.30.0/go.mod h1:/VUhepiaJMQUp4+oa/7Zr1D23ma6VTLIYjOOTFZPUcA=
+golang.org/x/sys v0.47.0/go.mod h1:4GL1E5IUh+htKOUEOaiffhrAeqysfVGipDYzABqnCmw=
-golang.org/x/text v0.28.0 h1:rhazDwis8INMIwQ4tpjLDzUhx6RlXqZNPEM0huQojng=
+golang.org/x/text v0.40.0 h1:Ub2Z6/xjgF1WrYQz2nuITOEegKFtiIy+rieRJ5lHZKs=
-golang.org/x/text v0.28.0/go.mod h1:U8nCwOR8jO/marOQ0QbDiOngZVEBB7MAiitBuMjXiNU=
+golang.org/x/text v0.40.0/go.mod h1:hpnzDAfGV753zIKo+wk3u1bVKCGPbrnF7+7LBF/UHVY=
+google.golang.org/protobuf v1.36.11 h1:fV6ZwhNocDyBLK0dj+fg8ektcVegBBuEolpbTQyBNVE=
+google.golang.org/protobuf v1.36.11/go.mod h1:HTf+CrKn2C3g5S8VImy6tdcUvCska2kB7j23XfzDpco=
 gopkg.in/check.v1 v0.0.0-20161208181325-20d25e280405/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
-gopkg.in/check.v1 v1.0.0-20190902080502-41f04d3bba15 h1:YR8cESwS4TdDjEe65xsg0ogRM/Nc3DYOhEAlW+xobZo=
+gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c h1:Hei/4ADfdWqJk1ZMxUNpqntNwaWcugrBjAiHlqqRiVk=
-gopkg.in/check.v1 v1.0.0-20190902080502-41f04d3bba15/go.mod h1:Co6ibVJAznAaIkqp8huTwlJQCZ016jof/cbN4VW5Yz0=
+gopkg.in/check.v1 v1.0.0-20201130134442-10cb98267c6c/go.mod h1:JHkPIbrfpd72SG/EVd6muEfDQjcINNoR0C8j2r3qZ4Q=
 gopkg.in/yaml.v3 v3.0.1 h1:fxVm/GzAzEWqLHuvctI91KS9hhNmmWOoWu0XTYJS7CA=
 gopkg.in/yaml.v3 v3.0.1/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=

backend/internal/config/config.go

@@ -44,6 +44,15 @@ var (
 	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(
+		"METRICS_USERNAME and METRICS_PASSWORD must be set together, " +
+			"or neither",
+	)
+	errMetricsUsernameColon = errors.New(
+		"METRICS_USERNAME must not contain \":\", " +
+			"which basic auth cannot carry in a user name",
+	)
 )
 
 // Params defines the dependencies for Config.
@@ -73,7 +82,8 @@ type Config struct {
 
 // New loads configuration from env, .env files, and config
 // files, returning a fully resolved Config. It fails, with an error
-// naming the setting, on a value the server cannot use.
+// naming the setting, on a value the server cannot use, and on a
+// config file it finds but cannot read.
 func New(
 	_ fx.Lifecycle,
 	params Params,
@@ -106,8 +116,8 @@ func New(
 	if err != nil {
 		var notFound viper.ConfigFileNotFoundError
 		if !errors.As(err, &notFound) {
-			log.Error("config file malformed", "error", err)
+			return nil, fmt.Errorf("config file %s: %w",
-			panic(err)
+				viper.ConfigFileUsed(), err)
 		}
 	}
 
@@ -177,6 +187,18 @@ func (s *Config) check() error {
 		}
 	}
 
+	// The server records and serves metrics only with both set, so
+	// one alone is a mistake that would otherwise go unnoticed.
+	if (s.MetricsUsername == "") != (s.MetricsPassword == "") {
+		return errMetricsCredentials
+	}
+
+	// Basic auth splits the credentials at the first ":", so with one
+	// in the user name every request to /metrics would get 401.
+	if strings.Contains(s.MetricsUsername, ":") {
+		return errMetricsUsernameColon
+	}
+
 	return checkOrigins(s.CORSAllowedOrigins)
 }
 

backend/internal/config/config_test.go

@@ -103,6 +103,32 @@ func TestDataDirMaxBytesMustBeANumber(t *testing.T) {
 	requireConfigError(t, "DATA_DIR_MAX_BYTES")
 }
 
+// TestMetricsCredentialsGoTogether: with only one of the two set, the
+// server would quietly serve no metrics, so the start fails, naming
+// both.
+func TestMetricsCredentialsGoTogether(t *testing.T) {
+	for _, set := range []string{"METRICS_USERNAME", "METRICS_PASSWORD"} {
+		t.Run(set, func(t *testing.T) {
+			t.Setenv("METRICS_USERNAME", "")
+			t.Setenv("METRICS_PASSWORD", "")
+			t.Setenv(set, "prometheus")
+
+			requireConfigError(t, "METRICS_USERNAME")
+			requireConfigError(t, "METRICS_PASSWORD")
+		})
+	}
+}
+
+// TestMetricsUsernameMustNotContainColon: basic auth splits the
+// credentials at the first ":", so such a user name would get 401 on
+// every request to /metrics.
+func TestMetricsUsernameMustNotContainColon(t *testing.T) {
+	t.Setenv("METRICS_USERNAME", "prom:etheus")
+	t.Setenv("METRICS_PASSWORD", "secret")
+
+	requireConfigError(t, "METRICS_USERNAME")
+}
+
 // TestCORSAllowedOriginsMustBeOrigins: "*" would let every origin in,
 // and an entry that is not a plain origin would match no page.
 func TestCORSAllowedOriginsMustBeOrigins(t *testing.T) {

backend/internal/middleware/middleware.go

@@ -18,9 +18,14 @@ import (
 	"sneak.berlin/go/netwatch/internal/globals"
 	"sneak.berlin/go/netwatch/internal/logger"
 
+	basicauth "github.com/99designs/basicauth-go"
 	"github.com/go-chi/chi/v5/middleware"
 	"github.com/go-chi/cors"
 	"github.com/go-chi/httprate"
+	"github.com/prometheus/client_golang/prometheus"
+	metrics "github.com/slok/go-http-metrics/metrics/prometheus"
+	ghmm "github.com/slok/go-http-metrics/middleware"
+	"github.com/slok/go-http-metrics/middleware/std"
 	"go.uber.org/fx"
 )
 
@@ -364,3 +369,25 @@ func (s *Middleware) RateLimit(
 		),
 	)
 }
+
+// Metrics returns middleware that records each request's duration and
+// response size, and the requests in progress, in registry. They are
+// labelled by the request path.
+func (s *Middleware) Metrics(
+	registry prometheus.Registerer,
+) func(http.Handler) http.Handler {
+	mdlw := ghmm.New(ghmm.Config{
+		Recorder: metrics.NewRecorder(metrics.Config{Registry: registry}),
+	})
+
+	return std.HandlerProvider("", mdlw)
+}
+
+// MetricsAuth returns middleware that lets a request through only with
+// METRICS_USERNAME and METRICS_PASSWORD as its basic auth credentials,
+// and answers any other with 401.
+func (s *Middleware) MetricsAuth() func(http.Handler) http.Handler {
+	return basicauth.New("metrics", map[string][]string{
+		s.params.Config.MetricsUsername: {s.params.Config.MetricsPassword},
+	})
+}

backend/internal/server/export_test.go

@@ -1,9 +1,21 @@
 package server
 
+import "github.com/go-chi/chi/v5"
+
+// Router exposes the router to the external tests, which add routes
+// of their own to it after SetupRoutes.
+func (s *Server) Router() *chi.Mux {
+	return s.router
+}
+
 // MaxRequestBodyBytes exposes the router-wide body limit to the
 // external tests.
 const MaxRequestBodyBytes = maxRequestBodyBytes
 
+// MetricsRequestsPerMinute exposes the /metrics rate limit to the
+// external tests.
+const MetricsRequestsPerMinute = metricsRequestsPerMinute
+
 // ListenAddr exposes the address the server listens on to the
 // external tests.
 func (s *Server) ListenAddr() string {

backend/internal/server/routes.go

@@ -3,8 +3,12 @@ package server
 import (
 	"time"
 
+	sentryhttp "github.com/getsentry/sentry-go/http"
 	"github.com/go-chi/chi/v5"
 	"github.com/go-chi/chi/v5/middleware"
+	"github.com/prometheus/client_golang/prometheus"
+	"github.com/prometheus/client_golang/prometheus/collectors"
+	"github.com/prometheus/client_golang/prometheus/promhttp"
 )
 
 const (
@@ -14,6 +18,12 @@ const (
 	// can mount s.mw.MaxBodyBytes with a smaller value to lower
 	// its bound, but cannot raise it: this cap runs first.
 	maxRequestBodyBytes int64 = 1 << 20 // 1 MiB
+
+	// metricsRequestsPerMinute is how many requests to /metrics each
+	// client address may make a minute, whatever their credentials. A
+	// scraper polling every 2 seconds sends half of it, which httprate
+	// never refuses.
+	metricsRequestsPerMinute = 60
 )
 
 // SetupRoutes configures the chi router with middleware and
@@ -29,13 +39,47 @@ func (s *Server) SetupRoutes() {
 	s.router.Use(s.mw.MaxBodyBytes(maxRequestBodyBytes))
 	s.router.Use(middleware.Timeout(requestTimeout))
 
-	s.router.Get(
+	// Sentry reports a panic, then panics again, so that s.mw.Recoverer
-		"/.well-known/healthcheck",
+	// still answers 500.
-		s.h.HandleHealthCheck(),
+	if s.params.Config.SentryDSN != "" {
+		s.router.Use(sentryhttp.New(sentryhttp.Options{Repanic: true}).Handle)
+	}
+
+	// The metrics go in a registry of this server's own, not in
+	// Prometheus' default one, which takes them only once per process.
+	registry := prometheus.NewRegistry()
+	registry.MustRegister(
+		collectors.NewGoCollector(),
+		collectors.NewProcessCollector(collectors.ProcessCollectorOpts{}),
 	)
 
-	s.router.Route("/api/v1", func(r chi.Router) {
+	// Requests are measured only once chi has matched them to one of
+	// these routes, by path and method. The metrics are labelled with
+	// both, which any client can make up, so measuring every request
+	// would let clients add labels without bound. A Route here would
+	// be matched by its path prefix alone, so each path is given in
+	// full.
+	s.router.Group(func(r chi.Router) {
+		// config.New refuses one of the two credentials without the
+		// other.
+		if s.params.Config.MetricsUsername != "" {
+			r.Use(s.mw.Metrics(registry))
+		}
+
+		r.Get("/.well-known/healthcheck", s.h.HandleHealthCheck())
+
 		r.With(s.mw.RateLimit(s.params.Config.ReportsPerMinute)).
-			Post("/reports", s.h.HandleReport())
+			Post("/api/v1/reports", s.h.HandleReport())
 	})
+
+	// The rate limit comes before the basic auth, so a client past it
+	// gets 429 and its password is not checked.
+	if s.params.Config.MetricsUsername != "" {
+		s.router.With(
+			s.mw.RateLimit(metricsRequestsPerMinute),
+			s.mw.MetricsAuth(),
+		).Get("/metrics", promhttp.HandlerFor(
+			registry, promhttp.HandlerOpts{},
+		).ServeHTTP)
+	}
 }

backend/internal/server/routes_test.go

@@ -1,10 +1,12 @@
 package server_test
 
 import (
+	"io"
 	"net/http"
 	"net/http/httptest"
 	"strings"
 	"testing"
+	"time"
 
 	"sneak.berlin/go/netwatch/internal/config"
 	"sneak.berlin/go/netwatch/internal/globals"
@@ -15,6 +17,7 @@ import (
 	"sneak.berlin/go/netwatch/internal/reportbuf"
 	"sneak.berlin/go/netwatch/internal/server"
 
+	"github.com/getsentry/sentry-go"
 	"go.uber.org/fx"
 	"go.uber.org/fx/fxtest"
 )
@@ -107,6 +110,203 @@ func TestCORSAllowedOriginsReachTheRouter(t *testing.T) {
 	}
 }
 
+// TestNoMetricsWithoutCredentials: with neither metrics setting set,
+// there is no /metrics.
+func TestNoMetricsWithoutCredentials(t *testing.T) {
+	t.Setenv("METRICS_USERNAME", "")
+	t.Setenv("METRICS_PASSWORD", "")
+
+	srv := newServer(t)
+	srv.SetupRoutes()
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequestWithContext(t.Context(),
+		http.MethodGet, "/metrics", http.NoBody)
+	srv.ServeHTTP(rec, req)
+
+	if rec.Code != http.StatusNotFound {
+		t.Fatalf("status = %d, want %d", rec.Code, http.StatusNotFound)
+	}
+}
+
+// TestMetricsBehindBasicAuth: with both metrics settings set, /metrics
+// answers only with them as basic auth credentials, and shows a
+// request to a route but not one to a path no route has.
+func TestMetricsBehindBasicAuth(t *testing.T) {
+	t.Setenv("METRICS_USERNAME", "prometheus")
+	t.Setenv("METRICS_PASSWORD", "right")
+
+	srv := newServer(t)
+	srv.SetupRoutes()
+
+	get := func(path, username, password string) *httptest.ResponseRecorder {
+		rec := httptest.NewRecorder()
+		req := httptest.NewRequestWithContext(t.Context(),
+			http.MethodGet, path, http.NoBody)
+
+		if username != "" {
+			req.SetBasicAuth(username, password)
+		}
+
+		srv.ServeHTTP(rec, req)
+
+		return rec
+	}
+
+	get("/.well-known/healthcheck", "", "")
+	get("/api/v1/no-such-route", "", "")
+
+	for _, creds := range [][2]string{
+		{"", ""},
+		{"prometheus", "wrong"},
+		{"someone", "right"},
+	} {
+		rec := get("/metrics", creds[0], creds[1])
+		if rec.Code != http.StatusUnauthorized {
+			t.Errorf("credentials %q: status = %d, want %d",
+				creds, rec.Code, http.StatusUnauthorized)
+		}
+	}
+
+	rec := get("/metrics", "prometheus", "right")
+	if rec.Code != http.StatusOK {
+		t.Fatalf("right credentials: status = %d, want %d",
+			rec.Code, http.StatusOK)
+	}
+
+	body := rec.Body.String()
+	if !strings.Contains(body, `handler="/.well-known/healthcheck"`) {
+		t.Errorf("metrics show no health check request:\n%s", body)
+	}
+
+	if strings.Contains(body, "no-such-route") {
+		t.Errorf("metrics show a request to a path no route has:\n%s", body)
+	}
+
+	if !strings.Contains(body, "go_goroutines") {
+		t.Errorf("metrics show no Go runtime metrics:\n%s", body)
+	}
+}
+
+// TestMetricsAreRateLimited: a client that has used up its /metrics
+// allowance on wrong passwords gets 429 even with the right one, which
+// is then not checked, while another client behind the same nginx
+// still gets in.
+func TestMetricsAreRateLimited(t *testing.T) {
+	t.Setenv("METRICS_USERNAME", "prometheus")
+	t.Setenv("METRICS_PASSWORD", "right")
+	// As in the container: nginx connects from loopback and names the
+	// client in X-Forwarded-For.
+	t.Setenv("TRUSTED_PROXIES", "127.0.0.1/32")
+
+	srv := newServer(t)
+	srv.SetupRoutes()
+
+	get := func(client, password string) int {
+		rec := httptest.NewRecorder()
+		req := httptest.NewRequestWithContext(t.Context(),
+			http.MethodGet, "/metrics", http.NoBody)
+		req.RemoteAddr = "127.0.0.1:40000"
+		req.Header.Set("X-Forwarded-For", client)
+		req.SetBasicAuth("prometheus", password)
+		srv.ServeHTTP(rec, req)
+
+		return rec.Code
+	}
+
+	for i := range server.MetricsRequestsPerMinute {
+		if code := get("203.0.113.7", "wrong"); code != http.StatusUnauthorized {
+			t.Fatalf("guess %d: status = %d, want %d",
+				i+1, code, http.StatusUnauthorized)
+		}
+	}
+
+	if code := get("203.0.113.7", "right"); code != http.StatusTooManyRequests {
+		t.Fatalf("right password past the limit: status = %d, want %d",
+			code, http.StatusTooManyRequests)
+	}
+
+	if code := get("203.0.113.8", "right"); code != http.StatusOK {
+		t.Fatalf("another client: status = %d, want %d",
+			code, http.StatusOK)
+	}
+}
+
+// TestMetricsInTwoServers: two servers in one process can both have
+// metrics on.
+func TestMetricsInTwoServers(t *testing.T) {
+	t.Setenv("METRICS_USERNAME", "prometheus")
+	t.Setenv("METRICS_PASSWORD", "right")
+
+	for range 2 {
+		newServer(t).SetupRoutes()
+	}
+}
+
+// TestSentry: with SENTRY_DSN empty there is no Sentry client. With it
+// pointing at a local server standing in for Sentry, a panic in a
+// handler reaches that server, and the request still gets the 500 from
+// the panic recovery.
+func TestSentry(t *testing.T) {
+	const panicMessage = "handler panic for TestSentry"
+
+	// sentry.Init sets the client for the whole process; take it away
+	// again so that no other test reports to Sentry.
+	t.Cleanup(func() { sentry.CurrentHub().BindClient(nil) })
+
+	t.Setenv("SENTRY_DSN", "")
+	newServer(t)
+
+	if sentry.CurrentHub().Client() != nil {
+		t.Fatal("a Sentry client exists with SENTRY_DSN empty")
+	}
+
+	// The body of the first request the stand-in for Sentry receives.
+	received := make(chan string, 1)
+
+	sentryServer := httptest.NewServer(http.HandlerFunc(
+		func(_ http.ResponseWriter, r *http.Request) {
+			body, _ := io.ReadAll(r.Body)
+
+			select {
+			case received <- string(body):
+			default:
+			}
+		},
+	))
+	defer sentryServer.Close()
+
+	t.Setenv("SENTRY_DSN",
+		"http://key@"+sentryServer.Listener.Addr().String()+"/1")
+
+	srv := newServer(t)
+	srv.SetupRoutes()
+	srv.Router().Get("/panic", func(http.ResponseWriter, *http.Request) {
+		panic(panicMessage)
+	})
+
+	rec := httptest.NewRecorder()
+	req := httptest.NewRequestWithContext(t.Context(),
+		http.MethodGet, "/panic", http.NoBody)
+	srv.ServeHTTP(rec, req)
+
+	if rec.Code != http.StatusInternalServerError {
+		t.Fatalf("status = %d, want %d",
+			rec.Code, http.StatusInternalServerError)
+	}
+
+	// Sentry sends from a goroutine of its own.
+	select {
+	case body := <-received:
+		if !strings.Contains(body, panicMessage) {
+			t.Fatalf("the Sentry server received no report of the panic:\n%s",
+				body)
+		}
+	case <-time.After(5 * time.Second):
+		t.Fatal("nothing reached the Sentry server")
+	}
+}
+
 // TestHealthCheckRejectsOversizeBody sends the health check, which
 // never reads its body, a body one byte over the limit. Only the
 // router-wide body limit can reject it.

backend/internal/server/server.go

@@ -7,8 +7,10 @@ package server
 
 import (
 	"context"
+	"fmt"
 	"log/slog"
 	"net/http"
+	"time"
 
 	"sneak.berlin/go/netwatch/internal/config"
 	"sneak.berlin/go/netwatch/internal/globals"
@@ -16,10 +18,15 @@ import (
 	"sneak.berlin/go/netwatch/internal/logger"
 	"sneak.berlin/go/netwatch/internal/middleware"
 
+	"github.com/getsentry/sentry-go"
 	"github.com/go-chi/chi/v5"
 	"go.uber.org/fx"
 )
 
+// sentryFlushTimeout is how long shutdown waits for Sentry to send
+// what it still holds.
+const sentryFlushTimeout = 2 * time.Second
+
 // Params defines the dependencies for Server.
 type Params struct {
 	fx.In
@@ -56,6 +63,11 @@ func New(
 	s.log = params.Logger.Get()
 	s.shutdowner = params.Shutdowner
 
+	err := s.enableSentry()
+	if err != nil {
+		return nil, err
+	}
+
 	lc.Append(fx.Hook{
 		OnStart: func(_ context.Context) error {
 			// Build the router and http.Server synchronously
@@ -87,10 +99,37 @@ func (s *Server) ServeHTTP(
 	s.router.ServeHTTP(w, r)
 }
 
+// enableSentry sets Sentry up when SENTRY_DSN is set, so that
+// SetupRoutes can report panics to it. With SENTRY_DSN empty it does
+// nothing. A DSN Sentry refuses stops the start.
+func (s *Server) enableSentry() error {
+	if s.params.Config.SentryDSN == "" {
+		return nil
+	}
+
+	err := sentry.Init(sentry.ClientOptions{
+		Dsn:     s.params.Config.SentryDSN,
+		Release: s.params.Globals.Appname + "-" + s.params.Globals.Version,
+	})
+	if err != nil {
+		return fmt.Errorf("SENTRY_DSN: %w", err)
+	}
+
+	s.log.Info("sentry error reporting activated")
+
+	return nil
+}
+
 // shutdown gracefully stops the HTTP server within the
-// deadline of the context fx provides for OnStop.
+// deadline of the context fx provides for OnStop, then gives
+// Sentry, if set up, time to send what it still holds.
 func (s *Server) shutdown(ctx context.Context) error {
 	err := s.httpServer.Shutdown(ctx)
+
+	if s.params.Config.SentryDSN != "" {
+		sentry.Flush(sentryFlushTimeout)
+	}
+
 	if err != nil {
 		s.log.Error("server clean shutdown failed", "error", err)
 

eslint.config.js

@@ -0,0 +1,27 @@
+// eslint's recommended rules over every JavaScript file in the repo.
+// script/frontend-lint runs it, in the frontend-lint stage of Dockerfile.
+import js from "@eslint/js";
+import globals from "globals";
+import { defineConfig } from "eslint/config";
+
+export default defineConfig([
+    js.configs.recommended,
+    {
+        // The page, and facts.js, which the viewport test runs inside it.
+        files: ["src/**/*.js", "test/viewport/facts.js"],
+        languageOptions: {
+            globals: {
+                ...globals.browser,
+                // vite.config.js defines these when it builds the page.
+                __COMMIT_HASH__: "readonly",
+                __COMMIT_FULL__: "readonly",
+            },
+        },
+    },
+    {
+        // Run by node: the tests, the viewport harness and these configs.
+        files: ["test/**/*.js", "*.js"],
+        ignores: ["test/viewport/facts.js"],
+        languageOptions: { globals: globals.node },
+    },
+]);

nginx.conf

@@ -68,4 +68,10 @@ server {
     location = /.well-known/healthcheck {
         proxy_pass http://127.0.0.1:8081;
     }
+
+    # The backend's Prometheus metrics, behind its own basic auth. Unless
+    # METRICS_USERNAME and METRICS_PASSWORD are set, it answers 404.
+    location = /metrics {
+        proxy_pass http://127.0.0.1:8081;
+    }
 }

package.json

@@ -6,12 +6,16 @@
     "scripts": {
         "dev": "vite",
         "build": "vite build",
-        "preview": "vite preview"
+        "preview": "vite preview",
+        "test": "node --test test/unit/*.test.js"
     },
     "license": "MIT",
     "devDependencies": {
+        "@eslint/js": "^10.0.1",
         "@tailwindcss/vite": "^4.1.18",
         "autoprefixer": "^10.4.23",
+        "eslint": "^10.12.0",
+        "globals": "^17.13.0",
         "postcss": "^8.5.6",
         "prettier": "^3.8.1",
         "puppeteer-core": "25.5.0",

script/add-dependency

@@ -0,0 +1,30 @@
+#!/bin/sh
+# script/add-dependency: add a frontend package, or move one to another
+# version, with yarn add, which changes package.json and yarn.lock
+# together; then install from yarn.lock with --frozen-lockfile, as
+# script/bootstrap does, to show it installs as written. --dev because
+# no frontend package is needed when the page runs: it ships as the
+# built dist/.
+set -eu
+
+ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
+
+usage() {
+    echo "usage: make add-dependency PACKAGE=<name>@<version>" >&2
+    exit 2
+}
+
+main() {
+    # Exactly one package. A value beginning with - would reach yarn as
+    # an option; yarn would quietly drop all but the first of several
+    # packages given in one value.
+    [ "$#" -eq 1 ] || usage
+    case "$1" in
+    "" | -* | *[[:space:]]*) usage ;;
+    esac
+    cd "$ROOT"
+    yarn add --dev "$1"
+    yarn install --frozen-lockfile
+}
+
+main "$@"

script/bootstrap

@@ -17,16 +17,18 @@
 #
 # golangci-lint is not installed: make lint runs it in Docker, which
 # this script does not install either.
+#
+# Unlike the org model: Go and gcc for backend/, a newer node for eslint.
 set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
-# Pinned versions, 2026-07-07
+# Pinned versions, 2026-07-06
 NODE_VERSION="22.17.0"
 # The oldest node the frontend's dependencies accept: the "engines"
-# field of puppeteer-core 25.5.0, the most demanding of them, asks for
+# field of eslint 10.12.0, the most demanding of them, asks for 22.13.0
-# 22.12.0 or newer, 2026-09-29. An older installed node is not used.
+# or newer, 2026-10-03. An older installed node is not used.
-NODE_MIN_VERSION="22.12.0"
+NODE_MIN_VERSION="22.13.0"
 NVM_VERSION="0.40.3"
 # sha256 of https://github.com/nvm-sh/nvm/archive/refs/tags/v0.40.3.tar.gz
 NVM_SHA256="5f4d6aaa04a177dc93c985e31dbc411ab6b8c6e1e21d8015dbc1372625fcd1d0"
@@ -270,7 +272,7 @@ main() {
 
     if missing docker; then
         echo "bootstrap: docker not found; make lint, and so make check" >&2
-        echo "  and the pre-commit hook, need it to run the Go linter" >&2
+        echo "  and the pre-commit hook, need it to run the linters" >&2
     fi
     if [ -n "$path_hint" ] && [ -d "$BIN_DIR" ]; then
         echo "bootstrap: add $BIN_DIR to the front of your PATH, e.g." >&2

script/build

@@ -0,0 +1,13 @@
+#!/bin/sh
+# script/build: build the frontend for production into dist/. The Go
+# backend is built by backend/script/build.
+set -eu
+
+ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
+
+main() {
+    cd "$ROOT"
+    yarn build
+}
+
+main "$@"

script/dev

@@ -0,0 +1,13 @@
+#!/bin/sh
+# script/dev: run the frontend's Vite dev server. It proxies /api to a
+# netwatch-server running locally (see vite.config.js).
+set -eu
+
+ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
+
+main() {
+    cd "$ROOT"
+    yarn dev
+}
+
+main "$@"

script/fmt

@@ -1,6 +1,7 @@
 #!/bin/sh
 # script/fmt: format the whole repo (writes): prettier over everything
 # it understands, then gofmt over the Go backend.
+# The org model formats only markdown; this repo also has JS and Go.
 set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"

script/fmt-check

@@ -1,6 +1,7 @@
 #!/bin/sh
 # script/fmt-check: check formatting across the whole repo (read-only).
 # Same scope as script/fmt, but fails instead of writing.
+# The org model checks only markdown; this repo also has JS and Go.
 set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"

script/frontend-check

@@ -1,9 +1,10 @@
 #!/bin/sh
-# script/frontend-check: run the frontend half of the checks only (test,
+# script/frontend-check: run the frontend tests and format check only.
-# lint, fmt-check). This exists for the frontend stage of Dockerfile, a
+# This exists for the frontend stage of Dockerfile, a node image with
-# node image with neither Go nor Docker; the Dockerfile's lint and
+# neither Go nor Docker; the Dockerfile's frontend-lint stage runs the
-# backend build stages gate the backend half. Everywhere else, use
+# frontend linter, and its lint and builder stages gate the backend.
-# script/check, which covers the whole repo. Must not modify any files.
+# Everywhere else, use script/check, which covers the whole repo. Must
+# not modify any files.
 set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
@@ -11,7 +12,6 @@ ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 main() {
     cd "$ROOT"
     "$ROOT/script/frontend-test"
-    "$ROOT/script/frontend-lint"
     "$ROOT/script/frontend-fmt-check"
 }
 

script/frontend-fmt

@@ -1,7 +1,7 @@
 #!/bin/sh
 # script/frontend-fmt: format the frontend and every other file prettier
-# understands, repo-wide (writes). backend/ is in .prettierignore; Go
+# understands, repo-wide (writes), the markdown in backend/ included.
-# sources are formatted by backend/script/fmt.
+# Prettier does not read Go; backend/script/fmt formats the Go sources.
 set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"

script/frontend-lint

@@ -1,12 +1,15 @@
 #!/bin/sh
-# script/frontend-lint: run the frontend linter (prettier in check mode).
+# script/frontend-lint: run eslint over the frontend. This runs inside
+# the frontend-lint stage of Dockerfile, the digest-pinned node image
+# with the packages from yarn.lock. From a checkout, run `make lint`,
+# which builds that stage.
 set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    yarn prettier --check .
+    yarn eslint .
 }
 
 main "$@"

script/frontend-test

@@ -1,14 +1,22 @@
 #!/bin/sh
 # script/frontend-test: run the frontend test suite: the unit tests in
-# test/unit/ with Node's built-in test runner, then the production
+# test/unit/, through the test script in package.json, then the
-# build, which fails on broken code.
+# production build, which fails on broken code. The tests print a dot
+# each; if any fails, they run again with every test listed, and the
+# script fails even if that run passes. NODE_OPTIONS chooses the
+# reporter because yarn adds its arguments after the test files, where
+# node would take a reporter option for one more file.
 set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    timeout 30 node --test test/unit/*.test.js
+    NODE_OPTIONS=--test-reporter=dot timeout 30 yarn --silent run test || {
+        echo "--- Rerunning with every test listed for details ---"
+        NODE_OPTIONS=--test-reporter=spec timeout 30 yarn --silent run test
+        exit 1
+    }
     timeout 30 yarn build
 }
 

script/install-precommit

@@ -8,8 +8,8 @@ ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 main() {
     cd "$ROOT"
     hook=".git/hooks/pre-commit"
-    printf '#!/bin/sh\nset -e\nscript/precommit\n' > "$hook"
+    printf '#!/bin/sh\nset -e\nscript/precommit\n' > .git/hooks/pre-commit
-    chmod +x "$hook"
+    chmod +x .git/hooks/pre-commit
     echo "pre-commit hook installed: runs script/precommit"
 }
 

script/lint

@@ -1,19 +1,21 @@
 #!/bin/sh
-# script/lint: lint the whole repo: prettier over the frontend, then the
+# script/lint: lint the whole repo: eslint over the frontend, then the Go
-# Go linter over backend/.
+# linter over backend/.
 #
-# The Go linter runs only in Docker: this builds the lint stage of
+# No linter runs on the host: this builds the frontend-lint and lint
-# Dockerfile, the digest-pinned golangci-lint image, which runs the
+# stages of Dockerfile, the digest-pinned node and golangci-lint images.
-# backend's fmt-check and lint targets. --no-cache makes the linter
+# The first runs eslint; the second runs the backend's fmt-check and
-# really run every time rather than reuse an earlier result, and the
+# lint targets. --no-cache makes each linter really run every time
-# stage is built for its checks alone, so no image is kept.
+# rather than reuse an earlier result, and each stage is built for its
+# checks alone, so no image is kept.
 set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
-    "$ROOT/script/frontend-lint"
+    timeout 300 docker build --no-cache --target frontend-lint \
+        --output type=cacheonly .
     timeout 300 docker build --no-cache --target lint \
         --output type=cacheonly .
 }

script/tidy

@@ -0,0 +1,13 @@
+#!/bin/sh
+# script/tidy: run go mod tidy in backend/, which adds the modules the
+# Go sources import, drops those they no longer do, and updates go.sum.
+set -eu
+
+ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
+
+main() {
+    cd "$ROOT/backend"
+    go mod tidy
+}
+
+main "$@"

src/main.js

@@ -8,6 +8,8 @@
 // display their real value in the latency figure. The history buffer holds
 // maxHistoryPoints samples (historyDuration / updateInterval).
 // reportInterval is how often collected samples are POSTed to the backend.
+// The interval menu changes updateInterval while the page runs; the
+// getters compute their values from it each time they are read.
 export const CONFIG = {
     updateInterval: 3000,
     maxHistoryPoints: 100,
@@ -28,6 +30,39 @@ export const CONFIG = {
         return [0, 1, 2, 3, 4, 5].map((i) => Math.round((d * i) / 5));
     },
     canvasHeight: 96,
+    // A latency figure and its sparkline take the color of the first entry
+    // whose limit, in ms, the latency is below.
+    latencyColors: [
+        { below: 50, hex: "#22c55e", className: "text-green-500" },
+        { below: 100, hex: "#84cc16", className: "text-lime-500" },
+        { below: 200, hex: "#eab308", className: "text-yellow-500" },
+        { below: 500, hex: "#f97316", className: "text-orange-500" },
+        { below: Infinity, hex: "#ef4444", className: "text-red-500" },
+    ],
+    // The health is offline when more than offlineTimeouts WAN hosts timed
+    // out or were unreachable and at most offlineReachable answered;
+    // otherwise degraded when more than degradedTimeouts timed out or were
+    // unreachable; otherwise slow when more than slowHosts answered after
+    // more than slowLatency ms.
+    offlineTimeouts: 10,
+    offlineReachable: 4,
+    degradedTimeouts: 4,
+    slowHosts: 3,
+    slowLatency: 1000,
+    // The debug log keeps its last maxLogEntries lines.
+    maxLogEntries: 1000,
+    // A gateway candidate that has not answered after gatewayTimeout ms is
+    // passed over.
+    gatewayTimeout: 1500,
+    // When no WAN host answers, the recovery probe checks recoveryProbeHosts
+    // random ones every recoveryProbeInterval ms.
+    recoveryProbeHosts: 4,
+    recoveryProbeInterval: 500,
+    // The rows are sorted after the first round that is not discarded, then
+    // every roundsPerSort rounds.
+    roundsPerSort: 10,
+    // The sparklines are sized and drawn again resizeDelay ms after start.
+    resizeDelay: 100,
 };
 
 // WAN endpoints to monitor. These are used for the aggregate health/stats
@@ -112,7 +147,8 @@ const debugLog = [];
 const log = (() => {
     function append(level, message) {
         debugLog.push({ timestamp: new Date(), level, message });
-        if (debugLog.length > 1000) debugLog.splice(0, debugLog.length - 1000);
+        if (debugLog.length > CONFIG.maxLogEntries)
+            debugLog.splice(0, debugLog.length - CONFIG.maxLogEntries);
         const panel = document.getElementById("debug-panel");
         if (panel && !panel.classList.contains("hidden")) renderDebugLog();
     }
@@ -151,7 +187,7 @@ function formatUTCTimestamp(date) {
 
 // --- Duration Formatting -----------------------------------------------------
 
-function humanDuration(seconds) {
+export function humanDuration(seconds) {
     const h = Math.floor(seconds / 3600);
     const m = Math.floor((seconds % 3600) / 60);
     const s = seconds % 60;
@@ -172,7 +208,10 @@ async function detectGateway() {
         const result = await Promise.any(
             GATEWAY_CANDIDATES.map(async (url) => {
                 const controller = new AbortController();
-                const timeoutId = setTimeout(() => controller.abort(), 1500);
+                const timeoutId = setTimeout(
+                    () => controller.abort(),
+                    CONFIG.gatewayTimeout,
+                );
                 try {
                     await fetch(url, {
                         method: "GET",
@@ -197,11 +236,35 @@ async function detectGateway() {
 
 // --- App State ---------------------------------------------------------------
 
-class HostState {
+// The min, max, median and average of latencies, a list of numbers, or all
+// null when it is empty. The median of an even count is the mean of the
+// middle two; it and the average are rounded.
+function latencyStats(latencies) {
+    if (latencies.length === 0)
+        return { min: null, max: null, med: null, avg: null };
+    const sorted = [...latencies].sort((a, b) => a - b);
+    const mid = Math.floor(sorted.length / 2);
+    return {
+        min: sorted[0],
+        max: sorted[sorted.length - 1],
+        med:
+            sorted.length % 2
+                ? sorted[mid]
+                : Math.round((sorted[mid - 1] + sorted[mid]) / 2),
+        avg: Math.round(
+            latencies.reduce((a, b) => a + b, 0) / latencies.length,
+        ),
+    };
+}
+
+export class HostState {
     constructor(host, pinned = false) {
         this.name = host.name;
         this.url = host.url;
-        this.history = []; // { timestamp, latency, paused }
+        // Each entry is either a check's result, { timestamp, latency,
+        // error }, or a round skipped while paused, { timestamp,
+        // latency: null, paused: true }.
+        this.history = [];
         this.lastLatency = null;
         this.status = "pending"; // 'online' | 'offline' | 'error' | 'pending'
         this.pinned = pinned;
@@ -225,45 +288,23 @@ class HostState {
         this._trim();
     }
 
-    averageLatency() {
+    // The min, max, median and average latency of the checks in the history
-        const valid = this.history.filter((p) => p.latency !== null);
+    // that got an answer.
-        if (valid.length === 0) return null;
+    historyStats() {
-        return Math.round(
+        return latencyStats(
-            valid.reduce((s, p) => s + p.latency, 0) / valid.length,
+            this.history
+                .filter((p) => p.latency !== null)
+                .map((p) => p.latency),
         );
     }
 
-    minLatency() {
-        const valid = this.history.filter((p) => p.latency !== null);
-        if (valid.length === 0) return null;
-        return Math.min(...valid.map((p) => p.latency));
-    }
-
-    maxLatency() {
-        const valid = this.history.filter((p) => p.latency !== null);
-        if (valid.length === 0) return null;
-        return Math.max(...valid.map((p) => p.latency));
-    }
-
-    medianLatency() {
-        const sorted = this.history
-            .filter((p) => p.latency !== null)
-            .map((p) => p.latency)
-            .sort((a, b) => a - b);
-        if (sorted.length === 0) return null;
-        const mid = Math.floor(sorted.length / 2);
-        return sorted.length % 2
-            ? sorted[mid]
-            : Math.round((sorted[mid - 1] + sorted[mid]) / 2);
-    }
-
     _trim() {
         while (this.history.length > CONFIG.maxHistoryPoints)
             this.history.shift();
     }
 }
 
-class AppState {
+export class AppState {
     constructor(localHosts) {
         this.wan = WAN_HOSTS.map(
             (h) => new HostState(h, h.name === "datavi.be"),
@@ -271,6 +312,10 @@ class AppState {
         this.local = localHosts.map((h) => new HostState(h));
         this.paused = false;
         this.tickCount = 0;
+        // The recovery probe's timer, null while it is not running, and the
+        // checks it started last.
+        this._recoveryProbeId = null;
+        this._recoveryProbeChecks = null;
     }
 
     get allHosts() {
@@ -279,33 +324,13 @@ class AppState {
 
     /** WAN-only stats from latest sample (excludes local) */
     wanStats() {
-        const reachable = this.wan.filter((h) => h.lastLatency !== null);
+        const latencies = this.wan
-        const latencies = reachable.map((h) => h.lastLatency);
+            .filter((h) => h.lastLatency !== null)
-        const total = this.wan.length;
+            .map((h) => h.lastLatency);
-        if (latencies.length === 0)
-            return {
-                reachable: 0,
-                total,
-                min: null,
-                max: null,
-                med: null,
-                avg: null,
-            };
-        const sorted = [...latencies].sort((a, b) => a - b);
-        const mid = Math.floor(sorted.length / 2);
-        const med =
-            sorted.length % 2
-                ? sorted[mid]
-                : Math.round((sorted[mid - 1] + sorted[mid]) / 2);
         return {
             reachable: latencies.length,
-            total,
+            total: this.wan.length,
-            min: Math.min(...latencies),
+            ...latencyStats(latencies),
-            max: Math.max(...latencies),
-            med,
-            avg: Math.round(
-                latencies.reduce((a, b) => a + b, 0) / latencies.length,
-            ),
         };
     }
 
@@ -331,12 +356,16 @@ class AppState {
         const timeouts = this.wan.filter(
             (h) => h.status === "error" || h.status === "offline",
         ).length;
-        if (timeouts > 10 && reachable <= 4) return "offline";
+        if (
-        if (timeouts > 4) return "degraded";
+            timeouts > CONFIG.offlineTimeouts &&
+            reachable <= CONFIG.offlineReachable
+        )
+            return "offline";
+        if (timeouts > CONFIG.degradedTimeouts) return "degraded";
         const slow = this.wan.filter(
-            (h) => h.lastLatency !== null && h.lastLatency > 1000,
+            (h) => h.lastLatency !== null && h.lastLatency > CONFIG.slowLatency,
         ).length;
-        if (slow > 3) return "slow";
+        if (slow > CONFIG.slowHosts) return "slow";
         return "healthy";
     }
 
@@ -546,23 +575,15 @@ export async function measureLatency(url, signal) {
 
 // --- Color Helpers -----------------------------------------------------------
 
-function latencyHex(latency) {
+export function latencyHex(latency) {
     if (latency === null) return "#6b7280";
-    if (latency < 50) return "#22c55e";
+    return CONFIG.latencyColors.find((c) => latency < c.below).hex;
-    if (latency < 100) return "#84cc16";
-    if (latency < 200) return "#eab308";
-    if (latency < 500) return "#f97316";
-    return "#ef4444";
 }
 
-function latencyClass(latency, status) {
+export function latencyClass(latency, status) {
     if (status === "offline" || status === "error" || latency === null)
         return "text-gray-500";
-    if (latency < 50) return "text-green-500";
+    return CONFIG.latencyColors.find((c) => latency < c.below).className;
-    if (latency < 100) return "text-lime-500";
-    if (latency < 200) return "text-yellow-500";
-    if (latency < 500) return "text-orange-500";
-    return "text-red-500";
 }
 
 // --- Sparkline Renderer ------------------------------------------------------
@@ -580,8 +601,8 @@ class SparklineRenderer {
         const ch = h - m.top - m.bottom;
 
         ctx.clearRect(0, 0, w, h);
-        SparklineRenderer._drawYAxis(ctx, w, h, m, ch);
+        SparklineRenderer._drawYAxis(ctx, w, m, ch);
-        SparklineRenderer._drawXAxis(ctx, w, h, m, cw);
+        SparklineRenderer._drawXAxis(ctx, h, m, cw);
 
         const len = history.length;
         const pw = cw / (CONFIG.maxHistoryPoints - 1);
@@ -597,7 +618,7 @@ class SparklineRenderer {
         SparklineRenderer._drawTip(ctx, history, getX, getY);
     }
 
-    static _drawYAxis(ctx, w, h, m, ch) {
+    static _drawYAxis(ctx, w, m, ch) {
         ctx.font = "300 12px monospace";
         ctx.textAlign = "right";
         ctx.textBaseline = "middle";
@@ -614,7 +635,7 @@ class SparklineRenderer {
         }
     }
 
-    static _drawXAxis(ctx, w, h, m, cw) {
+    static _drawXAxis(ctx, h, m, cw) {
         ctx.textAlign = "center";
         ctx.textBaseline = "top";
         for (const tick of CONFIG.xAxisTicks) {
@@ -702,7 +723,18 @@ class SparklineRenderer {
 // horizontally.
 const STATUS_TEXT_CLASS = "status-text text-xs text-right col-span-2 mt-5";
 
-function hostRowHTML(host, index, showPin = true) {
+// Escapes text for HTML, so it shows as written inside an element or a
+// quoted attribute and is never read as markup.
+function escapeHTML(text) {
+    return text
+        .replaceAll("&", "&amp;")
+        .replaceAll("<", "&lt;")
+        .replaceAll(">", "&gt;")
+        .replaceAll('"', "&quot;")
+        .replaceAll("'", "&#39;");
+}
+
+export function hostRowHTML(host, index, showPin = true) {
     const pinColor = host.pinned
         ? "text-blue-500"
         : "text-gray-600 hover:text-gray-400";
@@ -721,12 +753,12 @@ function hostRowHTML(host, index, showPin = true) {
         <div class="w-[420px] flex-shrink-0 grid grid-cols-[minmax(0,1fr)_auto] items-center">
           <div class="flex items-center gap-2 min-w-[200px]">
             <div class="w-3 h-3 rounded-full flex-shrink-0 bg-[#6b7280]"></div>
-            <span class="font-medium text-white truncate">${host.name}</span>
+            <span class="font-medium text-white truncate">${escapeHTML(host.name)}</span>
           </div>
           <div class="latency-value text-4xl font-bold tabular-nums text-right mt-3" data-host="${index}">
             <span class="text-gray-500">---</span>
           </div>
-          <a href="${host.url}" target="_blank" rel="noopener" class="text-xs text-gray-500 truncate block col-span-2 -mt-2">${host.url}</a>
+          <a href="${escapeHTML(host.url)}" target="_blank" rel="noopener" class="text-xs text-gray-500 truncate block col-span-2 -mt-2">${escapeHTML(host.url)}</a>
           <div class="${STATUS_TEXT_CLASS} text-gray-500" data-host="${index}">waiting...</div>
         </div>
         <div class="flex-grow sparkline-container rounded overflow-hidden border border-gray-700/30">
@@ -825,7 +857,7 @@ function buildUI(state) {
       </div>
 
       <footer class="mt-8 text-center text-gray-600 text-xs">
-        <p>Latency measured via GET requests | IPv4 only | CORS restrictions may affect some measurements</p>
+        <p>Latency measured via GET requests | CORS restrictions may affect some measurements</p>
         <p class="mt-2">
           <span class="inline-block w-3 h-3 rounded-full bg-green-500 mr-1 align-middle"></span>&lt;50ms
           <span class="inline-block w-3 h-3 rounded-full bg-lime-500 mr-1 ml-3 align-middle"></span>&lt;100ms
@@ -892,10 +924,7 @@ function updateHostRow(host, index) {
         latencyEl.innerHTML = `<span class="text-gray-500">---</span>`;
     }
 
-    const avg = host.averageLatency();
+    const { min, med, avg, max } = host.historyStats();
-    const med = host.medianLatency();
-    const min = host.minLatency();
-    const max = host.maxLatency();
     if (host.status === "online" && avg !== null) {
         statusEl.innerHTML = statusStatsHTML([
             ["min", min],
@@ -1042,14 +1071,16 @@ function renderDebugLog() {
         info: "text-gray-300",
         debug: "text-gray-500",
     };
-    el.innerHTML = debugLog
+    el.replaceChildren(
-        .map((entry) => {
+        ...debugLog.map((entry) => {
             const ts = formatUTCTimestamp(entry.timestamp);
-            const cls = levelColors[entry.level] || "text-gray-400";
             const lvl = entry.level.toUpperCase().padEnd(7);
-            return `<div class="${cls}">${ts} ${lvl} ${entry.message}</div>`;
+            const line = document.createElement("div");
-        })
+            line.className = levelColors[entry.level] || "text-gray-400";
-        .join("");
+            line.textContent = `${ts} ${lvl} ${entry.message}`;
+            return line;
+        }),
+    );
     el.scrollTop = el.scrollHeight;
 }
 
@@ -1105,7 +1136,7 @@ function sortAndRebuildWAN(state) {
 
 // --- Main Loop ---------------------------------------------------------------
 
-async function tick(state, signal, onOffline) {
+export async function tick(state, signal, onOffline) {
     const ts = Date.now();
 
     if (state.paused) {
@@ -1126,12 +1157,26 @@ async function tick(state, signal, onOffline) {
 
     log.debug(`Tick #${state.tickCount + 1} started`);
 
-    const results = await Promise.all(
+    // Each host's row shows its result as soon as its check ends. The
-        state.allHosts.map((h) => measureLatency(h.url, signal)),
+    // result is discarded if by then the user has paused or the next round
+    // has given up this one's checks, and in the first tick (tickCount is
+    // still 0), which is discarded as a whole below. The row is looked up
+    // when the check ends, as a pin click may have re-sorted the rows since
+    // the round started.
+    await Promise.all(
+        state.allHosts.map(async (host) => {
+            const r = await measureLatency(host.url, signal);
+            if (state.paused || signal.aborted || state.tickCount === 0) {
+                return;
+            }
+            host.pushSample(ts, r);
+            updateHostRow(host, state.allHosts.indexOf(host));
+            log.debug(`${host.name}: ${r.error ? r.error : r.latency + "ms"}`);
+        }),
     );
 
     // User may have paused, or the next round may have given up this
-    // one's checks, while awaiting results — discard them
+    // one's checks, while awaiting results — skip the rest of the round
     if (state.paused || signal.aborted) return;
 
     state.tickCount++;
@@ -1142,15 +1187,13 @@ async function tick(state, signal, onOffline) {
         return;
     }
 
-    state.allHosts.forEach((host, i) => {
+    // Redraw every row: if the user paused and resumed during this round,
-        const r = results[i];
+    // rows whose check ended before the resume still read "paused"
-        host.pushSample(ts, r);
+    state.allHosts.forEach((host, i) => updateHostRow(host, i));
-        updateHostRow(host, i);
-        log.debug(`${host.name}: ${r.error ? r.error : r.latency + "ms"}`);
-    });
 
-    // Sort after the first real check, then every 10 ticks thereafter
+    // Sort after the first real check, then every CONFIG.roundsPerSort
-    if (state.tickCount === 2 || state.tickCount % 10 === 1) {
+    // ticks thereafter
+    if (state.tickCount === 2 || state.tickCount % CONFIG.roundsPerSort === 1) {
         sortAndRebuildWAN(state);
     }
 
@@ -1173,9 +1216,10 @@ async function tick(state, signal, onOffline) {
 
 // --- Recovery Probe ----------------------------------------------------------
 
-// When offline, check 4 random WAN hosts every 500ms, giving up the checks
+// When offline, check CONFIG.recoveryProbeHosts random WAN hosts every
-// started 500ms before, so at most 4 are ever waiting. As soon as one
+// CONFIG.recoveryProbeInterval ms, giving up the checks started one interval
-// answers, stop probing and start a new round at once.
+// before, so at most that many are ever waiting. As soon as one answers,
+// stop probing and start a new round at once.
 function startRecoveryProbe(state, startRounds) {
     if (state._recoveryProbeId) return; // already running
     const candidates = [...state.wan];
@@ -1183,7 +1227,7 @@ function startRecoveryProbe(state, startRounds) {
         const j = Math.floor(Math.random() * (i + 1));
         [candidates[i], candidates[j]] = [candidates[j], candidates[i]];
     }
-    const canaries = candidates.slice(0, 4);
+    const canaries = candidates.slice(0, CONFIG.recoveryProbeHosts);
     log.notice(
         `Recovery probe started (${canaries.map((h) => h.name).join(", ")})`,
     );
@@ -1200,7 +1244,7 @@ function startRecoveryProbe(state, startRounds) {
                 startRounds();
             });
         }
-    }, 500);
+    }, CONFIG.recoveryProbeInterval);
 }
 
 function stopRecoveryProbe(state) {
@@ -1213,7 +1257,7 @@ function stopRecoveryProbe(state) {
 
 // --- Pause / Resume ----------------------------------------------------------
 
-function greyOutUI(state) {
+export function greyOutUI(state) {
     // Grey out all host rows
     state.allHosts.forEach((host, i) => {
         const latencyEl = document.querySelector(
@@ -1385,10 +1429,10 @@ async function init() {
     setInterval(updateClocks, 1000);
 
     // Rounds never overlap: a round first gives up the last round's checks
-    // if they are still waiting, and the last round then records nothing.
+    // if they are still waiting, and the last round then records nothing
-    // At a steady interval they never are, as they time out at 80% of it;
+    // more. At a steady interval they never are, as they time out at 80% of
-    // they can be when a round starts early, after an interval change or
+    // it; they can be when a round starts early, after an interval change
-    // when the recovery probe finds a target answering.
+    // or when the recovery probe finds a target answering.
     let roundChecks = new AbortController();
     function doTick() {
         roundChecks.abort();
@@ -1464,7 +1508,7 @@ async function init() {
         });
 
     window.addEventListener("resize", () => handleResize(state));
-    setTimeout(() => handleResize(state), 100);
+    setTimeout(() => handleResize(state), CONFIG.resizeDelay);
 }
 
 // Bootstrap only when loaded as the page: a real DOM containing the #app

test/unit/main.test.js

@@ -1,19 +1,76 @@
-// Unit tests for src/main.js, run by script/frontend-test with Node's
+// Unit tests for src/main.js, run with Node's built-in test runner by the
-// built-in test runner. Importing the module does not start the page.
+// test script in package.json. Importing the module does not start the
+// page.
 
-import { test } from "node:test";
+import { beforeEach, test } from "node:test";
 import assert from "node:assert/strict";
-import { CONFIG, measureLatency } from "../../src/main.js";
+import {
+    AppState,
+    CONFIG,
+    greyOutUI,
+    hostRowHTML,
+    HostState,
+    humanDuration,
+    latencyClass,
+    latencyHex,
+    measureLatency,
+    tick,
+} from "../../src/main.js";
 
-// measureLatency writes timeouts to the debug log, which looks for its
+// There is no page here, so the tests stand in for it. The debug log looks
-// panel in the page. There is no page here.
+// for its panel by id and finds none. Each element of a host's row that
-globalThis.document = { getElementById: () => null };
+// tick or greyOutUI draws into is a plain object, made the first time a
+// test looks it up and kept in elements under its selector until the next
+// test starts. As on a page, writing its text replaces its markup; the
+// status dot greyOutUI looks for in it is not there. Drawing a sparkline
+// does nothing; it looks for the pixel ratio on window and finds none.
+let elements;
+beforeEach(() => {
+    elements = {};
+});
+const doNothing = () => {};
+const canvasContext = {
+    clearRect: doNothing,
+    beginPath: doNothing,
+    moveTo: doNothing,
+    lineTo: doNothing,
+    stroke: doNothing,
+    fill: doNothing,
+    fillRect: doNothing,
+    fillText: doNothing,
+    arc: doNothing,
+};
+globalThis.window = {};
+globalThis.document = {
+    getElementById: () => null,
+    querySelector: (selector) =>
+        (elements[selector] ??= {
+            getContext: () => canvasContext,
+            querySelector: () => null,
+            set textContent(text) {
+                this.innerHTML = text;
+            },
+        }),
+};
+
+// What tick last wrote into the latency figure in host's row, or undefined
+// if it has written nothing there.
+function latencyFigure(state, host) {
+    const index = state.allHosts.indexOf(host);
+    return elements[`.latency-value[data-host="${index}"]`]?.innerHTML;
+}
+
+// What was last written into the status text in host's row.
+function statusText(state, host) {
+    const index = state.allHosts.indexOf(host);
+    return elements[`.status-text[data-host="${index}"]`]?.innerHTML;
+}
 
 // Mocks the clock for test t, so that a check lasting seconds takes no real
-// time, and replaces fetch with a target that answers after answerAfter
+// time, and replaces fetch with targets that each answer after
-// milliseconds of that clock, or never when answerAfter is Infinity. Both
+// answerAfter(url) milliseconds of that clock, or never when that is
-// are restored when the test ends.
+// Infinity. Both are restored when the test ends.
-function mockTarget(t, answerAfter) {
+function mockTargets(t, answerAfter) {
     t.mock.timers.enable({ apis: ["setTimeout", "Date"] });
     t.mock.method(performance, "now", () => Date.now());
     t.mock.method(
@@ -21,7 +78,9 @@ function mockTarget(t, answerAfter) {
         "fetch",
         (url, { signal }) =>
             new Promise((resolve, reject) => {
-                if (answerAfter !== Infinity) setTimeout(resolve, answerAfter);
+                if (answerAfter(url) !== Infinity) {
+                    setTimeout(resolve, answerAfter(url));
+                }
                 signal.addEventListener("abort", () => reject(signal.reason));
             }),
     );
@@ -42,7 +101,7 @@ for (const interval of [10000, 30000]) {
 
     test(`at a ${interval}ms interval, an answer after ${slowAnswer}ms is recorded with its real time`, async (t) => {
         CONFIG.updateInterval = interval;
-        mockTarget(t, slowAnswer);
+        mockTargets(t, () => slowAnswer);
         const check = measureLatency("https://target.test");
         t.mock.timers.tick(slowAnswer);
         assert.deepEqual(await settled(check), {
@@ -53,7 +112,7 @@ for (const interval of [10000, 30000]) {
 
     test(`at a ${interval}ms interval, a target that never answers is recorded as a timeout after ${timeout}ms`, async (t) => {
         CONFIG.updateInterval = interval;
-        mockTarget(t, Infinity);
+        mockTargets(t, () => Infinity);
         const check = measureLatency("https://target.test");
         t.mock.timers.tick(timeout - 1);
         assert.equal(await settled(check), "still waiting");
@@ -64,3 +123,310 @@ for (const interval of [10000, 30000]) {
         });
     });
 }
+
+test("at a 30000ms interval, a target answering after 1000ms shows in its row while another target's check is still waiting", async (t) => {
+    CONFIG.updateInterval = 30000;
+    const state = new AppState([
+        { name: "Answering", url: "https://answering.test" },
+    ]);
+    const answering = state.local[0];
+    const waiting = state.wan[0];
+    // No target but the answering one ever answers.
+    mockTargets(t, (url) => (url.startsWith(answering.url) ? 1000 : Infinity));
+    // The third tick: the first is discarded as a whole, and the second ends
+    // by sorting the rows, which rebuilds a page that is not here.
+    state.tickCount = 2;
+
+    const round = tick(state, new AbortController().signal);
+    t.mock.timers.tick(1000);
+    assert.equal(await settled(round), "still waiting");
+    assert.match(latencyFigure(state, answering), />1000</);
+    assert.equal(latencyFigure(state, waiting), undefined);
+    assert.equal(state.tickCount, 2);
+
+    // The round ends, once, when the last check times out.
+    t.mock.timers.tick(CONFIG.requestTimeout - 1000);
+    assert.notEqual(await settled(round), "still waiting");
+    assert.equal(state.tickCount, 3);
+});
+
+// In the next three tests, the answering target's check is still waiting
+// when something happens after which its result must not show.
+
+test("at a 30000ms interval, a check still waiting when its round is given up does not show in its row", async (t) => {
+    CONFIG.updateInterval = 30000;
+    const state = new AppState([
+        { name: "Answering", url: "https://answering.test" },
+    ]);
+    const answering = state.local[0];
+    mockTargets(t, (url) => (url.startsWith(answering.url) ? 1000 : Infinity));
+    state.tickCount = 2;
+
+    const roundChecks = new AbortController();
+    const round = tick(state, roundChecks.signal);
+    t.mock.timers.tick(500);
+    // As a round started early does to the last round's checks.
+    roundChecks.abort();
+    assert.notEqual(await settled(round), "still waiting");
+    assert.equal(latencyFigure(state, answering), undefined);
+});
+
+test("at a 30000ms interval, a check still waiting when the user pauses does not show in its row", async (t) => {
+    CONFIG.updateInterval = 30000;
+    const state = new AppState([
+        { name: "Answering", url: "https://answering.test" },
+    ]);
+    const answering = state.local[0];
+    mockTargets(t, (url) => (url.startsWith(answering.url) ? 1000 : Infinity));
+    state.tickCount = 2;
+
+    const round = tick(state, new AbortController().signal);
+    t.mock.timers.tick(500);
+    state.paused = true;
+    t.mock.timers.tick(500);
+    assert.equal(await settled(round), "still waiting");
+    assert.equal(latencyFigure(state, answering), undefined);
+});
+
+test("at a 30000ms interval, a check in the first round does not show in its row", async (t) => {
+    CONFIG.updateInterval = 30000;
+    const state = new AppState([
+        { name: "Answering", url: "https://answering.test" },
+    ]);
+    const answering = state.local[0];
+    mockTargets(t, (url) => (url.startsWith(answering.url) ? 1000 : Infinity));
+
+    const round = tick(state, new AbortController().signal);
+    t.mock.timers.tick(1000);
+    assert.equal(await settled(round), "still waiting");
+    assert.equal(latencyFigure(state, answering), undefined);
+});
+
+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) => {
+    CONFIG.updateInterval = 30000;
+    const state = new AppState([
+        { name: "Answering", url: "https://answering.test" },
+    ]);
+    const answering = state.local[0];
+    mockTargets(t, (url) => (url.startsWith(answering.url) ? 1000 : Infinity));
+    state.tickCount = 2;
+
+    const round = tick(state, new AbortController().signal);
+    t.mock.timers.tick(1000);
+    assert.equal(await settled(round), "still waiting");
+    // The user pauses, which greys out every row, and resumes, which leaves
+    // the rows as they are, as togglePause does. The answering target's
+    // check has already ended, so only the redraw of every row at the end
+    // of the round can take "paused" out of its row.
+    state.paused = true;
+    greyOutUI(state);
+    state.paused = false;
+    assert.equal(statusText(state, answering), "paused");
+
+    // The round ends when the last check times out.
+    t.mock.timers.tick(CONFIG.requestTimeout - 1000);
+    assert.notEqual(await settled(round), "still waiting");
+    for (const host of state.allHosts) {
+        assert.notEqual(statusText(state, host), "paused", host.name);
+    }
+});
+
+// The page shows &lt; &gt; &quot; &amp; and &#39; in a row's markup as
+// < > " & and '.
+test(`a target whose name and URL hold < > " & and ' shows those characters in its row`, () => {
+    const host = new HostState({
+        name: `<b>"x" & 'y'</b>`,
+        url: `https://x.test/<b>?a="x"&b='y'`,
+    });
+    const row = hostRowHTML(host, 0);
+    assert.doesNotMatch(row, /<b>/);
+    assert.ok(
+        row.includes(
+            ">&lt;b&gt;&quot;x&quot; &amp; &#39;y&#39;&lt;/b&gt;</span>",
+        ),
+    );
+    assert.ok(
+        row.includes(
+            'href="https://x.test/&lt;b&gt;?a=&quot;x&quot;&amp;b=&#39;y&#39;"',
+        ),
+    );
+    assert.ok(
+        row.includes(
+            ">https://x.test/&lt;b&gt;?a=&quot;x&quot;&amp;b=&#39;y&#39;</a>",
+        ),
+    );
+});
+
+for (const [seconds, text] of [
+    [0, "0s"],
+    [1, "1s"],
+    [59, "59s"],
+    [60, "1m"],
+    [61, "1m1s"],
+    [3599, "59m59s"],
+    [3600, "1h"],
+    [3601, "1h1s"],
+    [3660, "1h1m"],
+    [3661, "1h1m1s"],
+]) {
+    test(`humanDuration writes ${seconds} seconds as ${text}`, () => {
+        assert.equal(humanDuration(seconds), text);
+    });
+}
+
+// The colour of a target's latency figure, as a class, and of its sparkline
+// for an answer after latency ms, as the colour coding in README.md gives
+// them. The rows sit either side of each boundary.
+for (const [latency, figure, sparkline] of [
+    [0, "text-green-500", "#22c55e"],
+    [49, "text-green-500", "#22c55e"],
+    [50, "text-lime-500", "#84cc16"],
+    [99, "text-lime-500", "#84cc16"],
+    [100, "text-yellow-500", "#eab308"],
+    [199, "text-yellow-500", "#eab308"],
+    [200, "text-orange-500", "#f97316"],
+    [499, "text-orange-500", "#f97316"],
+    [500, "text-red-500", "#ef4444"],
+]) {
+    test(`an answer after ${latency}ms has a ${figure} figure and a ${sparkline} sparkline`, () => {
+        assert.equal(latencyClass(latency, "online"), figure);
+        assert.equal(latencyHex(latency), sparkline);
+    });
+}
+
+test("a check that timed out or found its target unreachable has a grey figure and sparkline", () => {
+    assert.equal(latencyClass(null, "error"), "text-gray-500");
+    assert.equal(latencyClass(null, "offline"), "text-gray-500");
+    assert.equal(latencyHex(null), "#6b7280");
+});
+
+// A target whose checks, in turn, answered after each of latencies ms, or,
+// for null, found it unreachable.
+function hostAfter(latencies) {
+    const host = new HostState({ name: "Target", url: "https://target.test" });
+    for (const latency of latencies) {
+        host.pushSample(
+            Date.now(),
+            latency === null
+                ? { latency: null, error: "unreachable" }
+                : { latency, error: null },
+        );
+    }
+    return host;
+}
+
+for (const { history, latencies, statistics } of [
+    {
+        history: "no checks",
+        latencies: [],
+        statistics: { min: null, max: null, average: null, median: null },
+    },
+    {
+        history: "only unreachable checks",
+        latencies: [null, null, null],
+        statistics: { min: null, max: null, average: null, median: null },
+    },
+    {
+        history: "three answers and an unreachable check",
+        latencies: [30, null, 10, 20],
+        statistics: { min: 10, max: 30, average: 20, median: 20 },
+    },
+    {
+        // The median of an even number of answers is the mean of the
+        // middle two. It and the average, 23.75, are rounded.
+        history: "four answers",
+        latencies: [10, 40, 20, 25],
+        statistics: { min: 10, max: 40, average: 24, median: 23 },
+    },
+    {
+        // Sorted as text rather than as numbers, these answers would put
+        // 100 in the middle. The average, 39.67, is rounded.
+        history: "answers with different numbers of digits",
+        latencies: [100, 9, 10],
+        statistics: { min: 9, max: 100, average: 40, median: 10 },
+    },
+]) {
+    test(`a target's min, max, average and median latency over ${history}`, () => {
+        const { min, max, avg, med } = hostAfter(latencies).historyStats();
+        assert.deepEqual({ min, max, average: avg, median: med }, statistics);
+    });
+}
+
+// The summary's figures come from each WAN target's last check, by the same
+// rules as a target's own: here four answered, one was found unreachable
+// and the rest have not been checked yet. The median, 22.5, and the
+// average, 21.25, are rounded.
+test("the summary's min, max, median and average latency over the WAN targets' last checks", () => {
+    const state = new AppState([]);
+    [30, 10, null, 25, 20].forEach((latency, i) =>
+        state.wan[i].pushSample(
+            Date.now(),
+            latency === null
+                ? { latency: null, error: "unreachable" }
+                : { latency, error: null },
+        ),
+    );
+    assert.deepEqual(state.wanStats(), {
+        reachable: 4,
+        total: state.wan.length,
+        min: 10,
+        max: 30,
+        med: 23,
+        avg: 21,
+    });
+});
+
+// An app state in which, of the WAN targets, the first timedOut timed out,
+// the next unreachable were found unreachable, the next answered answered
+// after latency ms, and the rest have not been checked yet.
+function stateAfter({ timedOut, unreachable, answered, latency }) {
+    const state = new AppState([]);
+    const results = [
+        ...Array(timedOut).fill({ latency: null, error: "timeout" }),
+        ...Array(unreachable).fill({ latency: null, error: "unreachable" }),
+        ...Array(answered).fill({ latency, error: null }),
+    ];
+    results.forEach((result, i) => state.wan[i].pushSample(Date.now(), result));
+    return state;
+}
+
+// For each number the health is decided by, the rows put it one under, at
+// and one over its threshold.
+for (const { timedOut, unreachable = 0, answered, latency, health } of [
+    // Offline: more than 10 timed out and at most 4 answered.
+    { timedOut: 9, answered: 4, latency: 30, health: "degraded" },
+    { timedOut: 10, answered: 4, latency: 30, health: "degraded" },
+    { timedOut: 11, answered: 4, latency: 30, health: "offline" },
+    { timedOut: 11, answered: 3, latency: 30, health: "offline" },
+    { timedOut: 11, answered: 5, latency: 30, health: "degraded" },
+    // Otherwise degraded: more than 4 timed out.
+    { timedOut: 3, answered: 10, latency: 30, health: "healthy" },
+    { timedOut: 4, answered: 10, latency: 30, health: "healthy" },
+    { timedOut: 5, answered: 10, latency: 30, health: "degraded" },
+    // Otherwise slow: more than 3 answered after more than 1000ms.
+    { timedOut: 0, answered: 4, latency: 999, health: "healthy" },
+    { timedOut: 0, answered: 4, latency: 1000, health: "healthy" },
+    { timedOut: 0, answered: 4, latency: 1001, health: "slow" },
+    { timedOut: 0, answered: 2, latency: 1001, health: "healthy" },
+    { timedOut: 0, answered: 3, latency: 1001, health: "healthy" },
+    // A target found unreachable counts as one that timed out.
+    {
+        timedOut: 5,
+        unreachable: 6,
+        answered: 4,
+        latency: 30,
+        health: "offline",
+    },
+    {
+        timedOut: 0,
+        unreachable: 5,
+        answered: 10,
+        latency: 30,
+        health: "degraded",
+    },
+]) {
+    test(`with ${timedOut} WAN targets timed out, ${unreachable} found unreachable and ${answered} answering after ${latency}ms, the health is ${health}`, () => {
+        const state = stateAfter({ timedOut, unreachable, answered, latency });
+        assert.equal(state.healthStatus(), health);
+    });
+}

test/viewport/README.md

@@ -49,10 +49,11 @@ sizes straddling the breakpoint for the rotation case.
   excluded: it is a design choice, not breakage.
 - **tap-targets-44px** — every interactive control is at least 44x44 CSS px on
   touch viewports, _and_ each selector in the control list matched at least the
-  number of visible elements it declares. The second half is what stops the
+  number of visible elements it declares: one of each single control, and one
-  check passing vacuously: with size alone, a renamed class would take its
+  pin button per WAN host row. The second half is what stops the check passing
-  controls out of the measured set and the check would report "all 0 controls
+  vacuously: with size alone, a renamed class would take its controls out of the
-  are at least 44x44" and pass. See below.
+  measured set and the check would report "all 0 controls are at least 44x44"
+  and pass. See below.
 - **host-rows-stacked / host-rows-side-by-side** — the rows genuinely reflow.
   Computed `flex-direction` _and_ the actual geometry are checked, and in the
   narrow layout the info block and the sparkline must each occupy essentially
@@ -76,7 +77,7 @@ the internet, so the app's latency probes cannot reach anything real. The
 harness answers them itself from a fixed delay table, with a deterministic
 fraction failed outright, so the rows render a realistic spread of one-, two-
 and three-digit latencies plus some unreachable rows. That spread is what the
-layout has to survive; 24 identical `---` placeholders would not exercise it.
+layout has to survive; a `---` placeholder in every row would not exercise it.
 
 ## What this cannot verify
 
@@ -103,10 +104,11 @@ Everything else this issue was actually about — does the layout reflow, does
 anything overflow, is content clipped, are the controls big enough — is a
 function of viewport width and CSS, and is covered above.
 
-## Relation to the unit test framework (#21)
+## Relation to the unit tests
 
-Complementary layers, not two stacks. `vitest` (#21) will exercise module-level
+Complementary layers, not two stacks. The unit tests in `test/unit/`, which
-logic in-process with no browser. This harness exercises rendered layout in a
+`make test` runs with Node's built-in test runner, exercise the functions
-real engine and is the only thing here that can see a media query. Neither
+`src/main.js` exports in-process with no browser. This harness exercises
-replaces the other; assertions about computed styles and element geometry belong
+rendered layout in a real engine and is the only thing here that can see a media
-here, assertions about functions belong in `vitest`.
+query. Neither replaces the other; assertions about computed styles and element
+geometry belong here, assertions about functions belong in `test/unit/`.

test/viewport/checks.js

@@ -15,7 +15,8 @@ export const MIN_TAP_TARGET_PX = 44;
 
 // The controls named in the definition of done, plus the pause button.
 // Each carries the smallest number of *visible* instances the page has to
-// contain for the tap-target oracle to be measuring anything at all.
+// contain, worked out from the facts gathered from that page, for the
+// tap-target oracle to be measuring every control it should.
 //
 // Without those floors the check is inert: `undersized` is empty both when
 // every control is large enough and when the selectors have gone stale and
@@ -24,13 +25,12 @@ export const MIN_TAP_TARGET_PX = 44;
 // for all three singleton controls vanishing at once — so the floor is per
 // selector, and one stale selector out of four fails the check.
 export const INTERACTIVE_CONTROLS = [
-    { selector: "#pause-btn", minCount: 1 },
+    { selector: "#pause-btn", minCount: () => 1 },
-    { selector: "#interval-select", minCount: 1 },
+    { selector: "#interval-select", minCount: () => 1 },
-    // One per pinnable host row. `app-rendered` already requires at least
+    // One per WAN host row, so pin buttons missing from even one row fail
-    // 10 host rows, so a count below that means the pin buttons stopped
+    // the check rather than only a drop below some fixed number.
-    // being rendered per row rather than that there were fewer hosts.
+    { selector: ".pin-btn", minCount: (facts) => facts.wanRowCount },
-    { selector: ".pin-btn", minCount: 10 },
+    { selector: "#debug-toggle", minCount: () => 1 },
-    { selector: "#debug-toggle", minCount: 1 },
 ];
 
 export const INTERACTIVE_SELECTORS = INTERACTIVE_CONTROLS.map(
@@ -105,8 +105,8 @@ export function evaluateChecks(facts, viewport, probes) {
     // never rendered. Everything below is only meaningful if this holds.
     check(
         "app-rendered",
-        facts.rowCount >= 10 && facts.numericLatencies >= 5,
+        facts.wanRowCount >= 10 && facts.numericLatencies >= 5,
-        `${facts.rowCount} host rows, ${facts.numericLatencies} showing a numeric latency`,
+        `${facts.wanRowCount} WAN host rows, ${facts.numericLatencies} showing a numeric latency`,
     );
 
     const viewportWidth = Math.min(facts.innerWidth, facts.documentClientWidth);
@@ -162,7 +162,8 @@ export function evaluateChecks(facts, viewport, probes) {
             seen.set(target.selector, (seen.get(target.selector) ?? 0) + 1);
         }
         const missing = INTERACTIVE_CONTROLS.filter(
-            (control) => (seen.get(control.selector) ?? 0) < control.minCount,
+            (control) =>
+                (seen.get(control.selector) ?? 0) < control.minCount(facts),
         );
 
         const undersized = facts.tapTargets.filter(
@@ -184,11 +185,11 @@ export function evaluateChecks(facts, viewport, probes) {
         const detail = [];
         if (missing.length > 0) {
             detail.push(
-                "oracle is not measuring the page: " +
+                "oracle is not measuring every control: " +
                     summarise(
                         missing,
                         (c) =>
-                            `${c.selector} matched ${seen.get(c.selector) ?? 0} visible element(s), expected at least ${c.minCount}`,
+                            `${c.selector} matched ${seen.get(c.selector) ?? 0} visible element(s), expected at least ${c.minCount(facts)}`,
                         4,
                     ),
             );

test/viewport/facts.js

@@ -173,7 +173,9 @@ export function collectLayoutFacts(options) {
         clipped,
         tapTargets,
         rows,
-        rowCount: document.querySelectorAll(".host-row").length,
+        // WAN host rows only: each has a pin button, and the tap-target
+        // check expects one per row. The local host rows have none.
+        wanRowCount: document.querySelectorAll("#wan-hosts .host-row").length,
         numericLatencies: Array.from(
             document.querySelectorAll(".latency-value"),
         ).filter((el) => /\d/.test(el.textContent)).length,

test/viewport/harness.js

@@ -16,6 +16,10 @@ import { collectLayoutFacts } from "./facts.js";
 import { evaluateChecks, INTERACTIVE_SELECTORS } from "./checks.js";
 import { deriveViewports } from "./viewports.js";
 
+// page.waitForFunction and page.evaluate run the functions given them in
+// the page, where these are defined.
+/* global document, requestAnimationFrame */
+
 function required(name) {
     const value = process.env[name];
     if (!value) {
@@ -62,7 +66,6 @@ function stableHash(text) {
 
 async function connectBrowser() {
     const deadline = Date.now() + BROWSER_TIMEOUT_MS;
-    let lastError;
     for (;;) {
         try {
             const response = await fetch(`${CDP_URL}/json/version`);
@@ -77,9 +80,10 @@ async function connectBrowser() {
             });
             return { browser, version: info.Browser };
         } catch (error) {
-            lastError = error;
             if (Date.now() > deadline) {
-                throw new Error(`browser never came up: ${lastError}`);
+                throw new Error(`browser never came up: ${error}`, {
+                    cause: error,
+                });
             }
             await sleep(250);
         }