Old Revisions

These are the older revisons of the current document. To revert to an old revision, select it from below, click Edit this page and save it.

2026/08/27 23:21 design:dns – Generic (fable) review applied. Worst finding: 'What to report' claimed the extraction measures which resolver a paper used, which the provenance page says in bold it cannot — the 42.2% is a tool-version rate, and the page now publishes no resolver-identi karel.kubicek.claude +3.6 KB (current)
2026/08/27 23:15 design:dns – Figures re-check applied: drop the dnsdb.info A record (it rotated between two runs an hour apart) and correct the Rapid7 guard's tuple count. Authored by Claude. karel.kubicek.claude +92 B
2026/08/27 23:08 design:dns – DDR box: the lead sentence kept the unhedged 'deployment is broken' that the rest of the box contests. Rewritten around a fetchable primary source (the same authors' APNIC write-up: 321k DDR-enabled resolvers, Google 80.8%, top five 97.4%) rather than the karel.kubicek.claude +746 B
2026/08/27 23:03 design:dns – Apply figures-vs-script and external-currency reviews. Biggest: the published script's verdict() intersected all three resolvers at once, so two agreeing and one differing read as 'disjoint' — 17 of 28 names were mislabelled and the page's headline was wr karel.kubicek.claude +9.6 KB
2026/08/27 22:52 design:dns – Second self-review pass: the six lab-rig cases in the resolver-software family are all dnsmasq and the seventh sits in a different family (was wrongly attributed to seven resolver-software papers); and the LLM growth signal lives in the provisional 2025-2 karel.kubicek.claude +262 B
2026/08/27 22:51 design:dns – Citations review (sonnet) applied: restore Nisenoff et al.'s 73% as a verbatim quote — it is in paper.pdf and absent from paper.cols.txt, so the .cols-only check had wrongly called it unverifiable; verify_dns_figures.mjs now falls back to pypdf. Also corr karel.kubicek.claude +260 B
2026/08/27 22:49 design:dns – Self-review pass before the reviewers report: drop an unverifiable 73% (the extraction asserted it; the only '73' near 'heard of DNS' in the paper is the string 'Firefox 73'), replace an unsupported 'wrong 60%' with measured venue shares, replace 'the amp karel.kubicek.claude +1 KB
2026/08/27 22:39 design:dns – New page: DNS measurement as a design decision — population (244 papers, A=title 108 / B=instrument 202), instrument inventory with 2026-08-27 currency, five failure modes, encrypted-DNS timeline against RFC status, the web slice (82), reporting gaps, and karel.kubicek.claude +47.3 KB