Files
cable-latency-survey/README.md
T
Toby c094eb88fa Open-source release: pipeline code, raw logs, derived data, licences, tests
Adds src/ (scripts, raw measurement logs, derived data, report source),
tests/, LICENSE (MIT, code) and LICENSE-DATA.md (CC BY 4.0 for the survey's
data; third-party terms). The report is updated to the current site; the
v1 and v2 data hand-off files are unchanged.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-10-27 20:09:54 -04:00

3.7 KiB
Raw Blame History

Cable Latency Survey

Measured round-trip times between cities, compared with the minimum propagation time along submarine-cable routes on four models of the Earth's surface: a sphere, a North-pole azimuthal equidistant (AE) map, an AE map centred on 0°N 0°E ("dual-pole"), and a flat Mercator map. A round-trip time below the fibre propagation floor along a model's route is inconsistent with that model.

  • Report: https://pingstats.aethercosmology.com (also cable_latency_survey_report.html, a single file that works offline)
  • Code: MIT (LICENSE). Measurements and derived data: CC BY 4.0, with third-party terms (LICENSE-DATA.md).

Files

Path Contents
cable_latency_survey_data_v2.json Current data hand-off. 29,341 measurements, 305 sites, every routed path and cable geometry ([lon, lat]). The revision block lists the changes from v1.
cable_latency_survey_data.json v1: the same measurements without coordinates. Kept for reference.
cable_latency_survey_report.html The report, identical to the website.
src/ Pipeline code, raw logs and derived data (below).
tests/ Headless-browser checks of the report (tests/run.sh).

Raw measurements

Source File
Globalping probes, runs 1 and 2 src/data/live_20261023T1812Z.json, src/data/live_run2.json (each result with its Globalping measurement ID)
Vultr servers (8, 2026-10-23 21:23–22:30 UTC) src/data/vps/<server>/ping.log, src/data/vps/<server>/mtr.log (traceroutes)
Residential connection, Michigan src/data/home/ping.log (location withheld)
RIPE Atlas sea crossings src/data/trace.json (hops, timings, measurement and probe IDs)
WonderNetwork, 2020-07-19/20 per-pair statistics src/data/wn_clean.json; archive downloaded by src/fetch_raw.sh

Pipeline

Run from src/. The measurement steps contact external services and are not needed to rebuild the report.

Step Script Output
Distance on each model, constants models.py —
Globalping measurements measure.py data/live_*.json
Vultr servers: create, measure, collect, delete vps.py data/vps/
Vultr logs to measurements vps_analyze.py data/vps_cells.json
WonderNetwork archive: statistics, mislocated servers analyze_wn.py data/wn_clean.json
Published cable lengths on each model cables.py data/cables.json
RIPE Atlas sea crossings trace.py data/trace.json
Cable-and-land routing on each model network.py data/network.json
Best attainable flat arrangements flatfit.py data/flatfit.json
Network attribution from public BGP data asmap.py data/asmap.json, data/rpki.json
Shape fit without positions pingmap.py data/pingmap.json
North-pole AE longitude solver (hand-off only) solver.py data/solver.json
Report build.py site/index.html, web/index.html
Data hand-off export.py handoff/cable_latency_survey_data_v2.json

Reproduce

git clone https://git.aethercosmology.com/toby/cable-latency-survey.git
cd cable-latency-survey/src
pip install -r requirements.txt
./fetch_raw.sh          # TeleGeography, WonderNetwork and BGP inputs (not redistributed)
python3 build.py        # the report, from the files in data/
python3 flatfit.py      # flat arrangements
python3 export.py       # the data hand-off

build.py needs only data/ and the TeleGeography files; it does not contact any measurement service. fetch_raw.sh reports whether the TeleGeography data it downloads matches the snapshot used (src/raw/tg.sha256).

Browser checks: cd tests && npm install && CH=/path/to/chromium ./run.sh.