Stop the strict acceptance from waiting for itself

A scenario whose body ran past about 19 s idled until 45 s after its
start, and twice one never exited and hit the runner's 300 s timeout
after it had passed: something of Chrome outlived Browser.close and kept
its stdout and stderr open.

- scripts/qa/browser.mjs closes Chrome within 5 s, kills its process
  group and closes its pipes; the harness then finishes and exits with
  the scenario's code. QA_EXIT_DIAGNOSTICS=1 keeps the old ending and
  prints the open handles and Chrome's process tree 5 s after finish.
- Every strict page starts with its clock at the 18:30 UTC nearest to
  the real time, so Navara's sun is up over the test cities whenever the
  acceptance runs; a scenario can ask for its own instant or the real
  clock (Wildlife does, for its freshness fixtures).
- The runner records a scenario that passed but did not exit as hung,
  kills the whole process group on a timeout, runs a failed scenario
  once more and reports one that passes then as flaky, and appends each
  run to --history=FILE; scripts/report-acceptance-flake.mjs reads it.

scripts/qa/browser.mjs

@@ -0,0 +1,73 @@
+import { processTree, readProcesses } from './process.mjs';
+
+/** Bound shutdown even when Chrome leaves descendants holding its pipes. */
+export async function closeBrowser(browser, { deadlineMs = 5000 } = {}) {
+  const start = performance.now();
+  let child;
+  let closed = false;
+  let killed = false;
+  let timer;
+  try {
+    child = browser.process();
+  } catch {}
+  try {
+    closed = await Promise.race([
+      Promise.resolve()
+        .then(() => browser.close())
+        .then(
+          () => true,
+          () => false,
+        ),
+      new Promise((resolve) => {
+        timer = setTimeout(() => resolve(false), deadlineMs);
+      }),
+    ]);
+  } catch {}
+  clearTimeout(timer);
+  if (process.env.QA_EXIT_DIAGNOSTICS !== '1') {
+    try {
+      if (Number.isInteger(child?.pid) && child.pid > 0) {
+        // The group can outlive its leader, so probe the group itself.
+        process.kill(-child.pid, 0);
+        process.kill(-child.pid, 'SIGKILL');
+        killed = true;
+      }
+    } catch {}
+    for (const name of ['stdout', 'stderr']) {
+      try {
+        child?.[name]?.destroy();
+      } catch {}
+    }
+  }
+  return { closed, killed, ms: performance.now() - start };
+}
+
+export async function exitAfterFlush(code = process.exitCode ?? 0) {
+  process.exitCode = code;
+  if (process.env.QA_EXIT_DIAGNOSTICS === '1') return;
+  await Promise.all(
+    [process.stdout, process.stderr].map(
+      (stream) =>
+        new Promise((resolve) => {
+          if (stream.destroyed) return resolve();
+          stream.write('', resolve);
+        }),
+    ),
+  );
+  process.exit(code);
+}
+
+export function diagnoseExit(pids) {
+  if (process.env.QA_EXIT_DIAGNOSTICS !== '1') return;
+  setTimeout(() => {
+    console.error('QA EXIT resources', process.getActiveResourcesInfo());
+    const processes = readProcesses();
+    console.error('QA EXIT Chrome: pid ppid pgid elapsed command');
+    for (const pid of pids) {
+      const rows = processTree(pid, processes);
+      console.error(`Chrome ${pid}:`);
+      for (const row of rows) console.error(row.line);
+      if (!rows.length) console.error('(no remaining processes)');
+    }
+  }, 5000).unref();
+}