The comparison table scored Signal wrong, in our favour (relay#11)

`/comparison` scored Signal **No** on "post-quantum in the ongoing
ratchet". Signal shipped exactly that: the Sparse Post-Quantum Ratchet,
announced 2025-10-02, combined with their Double Ratchet as the "Triple
Ratchet" — continuous post-quantum key agreement, not just PQXDH's initial
handshake.

🔴 **Wrong in the direction that flatters us**, on a page whose own closing
note calls that a bug and asks to be told. Verified from signal.org/blog/spqr
rather than from memory, because the whole problem with this row was
somebody being confident.

The knock-on sentence went with it. "Apple is ahead of everyone here on
post-quantum" was true only while Signal's cell said No; both carry it
through the ongoing conversation, and the paragraph now says we are not
ahead of anyone on that row and names the date this page was wrong.

⚠️ **Our own two cells were checked at the same time and are RIGHT** — I
went in expecting to correct those too. Android's ML-KEM-768 is
BouncyCastle's FIPS 203 implementation, not a stub, and iOS's is CryptoKit
`MLKEM768`, unconditional since the deployment target moved to 26.0. The
cross-engine interop proves it end to end: "Android picks v3 for an iPhone
peer", "the first Android envelope carries a PQ ratchet step", "a v3
session survives being archived and restored mid-conversation".

659 tests pass.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

Resources/Views/comparison.leaf

@@ -83,9 +83,9 @@
           <tr>
             <th scope="row">Post-quantum in the ongoing ratchet</th>
             <td class="mine"><strong class="part">Android now</strong><br><small>The row above protects the start of a conversation; this carries the same protection through every message, with fresh maths each time rather than once at the beginning. Out now on Android, on iPhone once Apple approves &mdash; and like everything here it counts only once you both have it, so today that means two Android phones.</small></td>
+            <td><strong class="yes">Yes</strong><br><small>Signal&rsquo;s Sparse Post-Quantum Ratchet, shipped in October 2025 and rolling out gradually. It runs alongside their Double Ratchet rather than replacing it.</small></td>
             <td><strong class="no">No</strong></td>
-            <td><strong class="no">No</strong></td>
-            <td><strong class="yes">Yes</strong><br><small>The most advanced of the four on this row.</small></td>
+            <td><strong class="yes">Yes</strong><br><small>Using PQ3.</small></td>
           </tr>
           <tr>
             <th scope="row">Keys verified automatically, not just by hand</th>
@@ -154,8 +154,12 @@
       rather you trust the table than the tone of it.
     </p>
     <p>
-      All four are serious pieces of engineering, and Apple is ahead of everyone
-      here on post-quantum. But every other messenger on this table has one
+      All four are serious pieces of engineering, and on post-quantum we are
+      not ahead of anyone: Apple and Signal both carry it through the ongoing
+      conversation, and Signal has been doing so since October 2025. This page
+      scored that row wrong &mdash; against Signal, in our own favour &mdash;
+      until 2026-09-04, which is exactly the kind of mistake the note below asks
+      you to report. But every other messenger on this table has one
       thing in common: somebody else runs the server, and you are trusting them
       to. Flamenet Messenger is built so that you do not have to — the whole
       thing runs on a machine you control, under a licence that lets you keep it