pages: the wall paragraph counted to 27; the reader gate makes it 28

html/cubelinux-builds-to-date.html

@@ -256,7 +256,7 @@ write&rsquo;s two <code>fsync</code>s into one, and the box has not been re-meas
   <tr><td>a reader refuses a device it cannot hold</td><td><code>verify-store-readers</code></td><td class="ok">pass — today: a store device refused in <b>0&nbsp;s</b> with the reason and the remedy, byte-identical afterwards; a 16&nbsp;GB device-shaped file refused under a 2&nbsp;GB cap rather than mapped; and the readers whose subject <em>is</em> a device still reading it</td></tr>
   <tr><td>a store on the USB dongle, written and read back</td><td><code>verify-dongle</code></td><td class="ok">pass — today, run on its own with the device attached: four records written through <code>cube(2)</code>, the machine gone, and a later boot walking <b>4 space(s), 8 record(s)</b> off the device — the records, the device's <b>own space</b> (<code>0xF9 ‖ SHA-256(serial)</code>) with its identity record, and the <b>portal</b> entered from this machine's root that leads to it, which the userspace decoder reads back off the device. The gate skips with a reason when the stick is not in the machine</td></tr>
 </table>
-<p class="muted">The board names <b>28 gates</b> — <code>verify-no-store</code> (the walk defect's own gate), <code>verify-dongle</code> (a store on removable media) and <code>verify-store-readers</code> (a device refused before it is read) are all new today. The whole wall was run for this build — <b>26 of the 26 it named then: 26 pass, 0 red</b> — and <code>verify-dongle</code> was added and run on its own immediately afterwards, with the device attached: pass. A wall run from here on names 27. Every red the day produced is in §8 and none of them survives as a defect: two were gates accusing the kernel of what the live box was doing, one was the walk's own, and one was a store shape whose writes were acknowledged without a single flush on the path. The workspace suite — the crate tests that underpin all of the above — is <b>218 passed, 0 failed</b>.</p>
+<p class="muted">The board names <b>28 gates</b> — <code>verify-no-store</code> (the walk defect's own gate), <code>verify-dongle</code> (a store on removable media) and <code>verify-store-readers</code> (a device refused before it is read) are all new today. The whole wall was run for this build — <b>26 of the 26 it named then: 26 pass, 0 red</b> — and the two added afterwards — <code>verify-dongle</code>, run on its own with the device attached, and <code>verify-store-readers</code>, run on its own on this machine — both <span class="ok">pass</span>. A wall run from here on names 28. Every red the day produced is in §8 and none of them survives as a defect: two were gates accusing the kernel of what the live box was doing, one was the walk's own, and one was a store shape whose writes were acknowledged without a single flush on the path. The workspace suite — the crate tests that underpin all of the above — is <b>218 passed, 0 failed</b>.</p>
 
 <h2 id="s8">8. Defects the gates caught</h2>
 <p>Named, because each was found only by running the real thing, and a record that shows only the green is a ledger rather than a record:</p>