From 9a928a79848a8e7effcbf034d471e5226b6329d8 Mon Sep 17 00:00:00 2001
From: "Jeffrey C. Ollie" <jeff@ocjtech.us>
Date: Fri, 11 Sep 2026 08:51:18 -0500
Subject: [PATCH] Note that the cited blog post still wants archiving

The convention is to record both the original URL and a Wayback snapshot, so
that a citation outlives the site it points at. Only the original is there:
archive.org answered 429 and then served its "Temporarily Offline" page while
the citation was being written, so no snapshot could be made or looked up.

The note sits next to the citation rather than in a list somewhere else,
since that is where whoever picks it up will be looking. A personal blog is
exactly the kind of source this convention exists for.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01NGYDXszLdpBxhkfqqHp17B
---
 README.md | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/README.md b/README.md
index 21e4e5c..30b0e8f 100644
--- a/README.md
+++ b/README.md
@@ -325,6 +325,13 @@ Some notes on throughput and how to generate test data are in
    4 March 2026.
    <https://ghostinthedata.info/posts/2026/2026-03-04-csv-test-suite/>
 
+<!-- TODO: submit the Hillman post to the Internet Archive and record the
+     Wayback snapshot alongside the original URL above, so the citation
+     outlives the site. It was not done when the citation was added because
+     archive.org was returning 429 and then serving its "Temporarily Offline"
+     page; nothing is wrong with the URL itself, which answered 200. A
+     personal blog is exactly the kind of source worth archiving. -->
+
    Turns the RFC and a catalogue of real-world failures into concrete tests.
    Two changes here came out of reading it: a byte order mark is consumed
    rather than handed back as part of the first field, and an unclosed quote
-- 
knotmirror 0.0.1