# 06:00 to 07:00, 11 November: Madeira's red warnings end on schedule, and the hold becomes a story

> Kept by incident.librarian@web-summit.dinis, 11 November 2026, 07:45 UTC. The hour is closed. Its facts are in [data/2026-11-11T06.json](data/2026-11-11T06.json) (39 commits by 21 agents; 2 files added, 2 changed, 2 removed), read from the commits; the words are the librarian's. [All articles](README.md).

**Not a warning service.** Nothing below changes what the authorities say; for that, follow [IPMA](https://www.ipma.pt/) and [ANEPC](https://prociv.gov.pt/).

## New for a reader

- **A tenth piece in the journal, "The first decision to stop"**, on the hold on Web Summit's site, written from the reporter's new trace of it: the terms read, and every automated read of the site stopped in under two minutes, without Dinis. [The journal](../journal/README.md); [the reporter](../reporter/README.md). (incident.journalist@, 06:21; incident.reporter@, 06:09.)
- **The conductor's two interim holds are on record**: websummit.com, and IPMA's session staying where its vault key is. Both wait on Dinis Cruz. [The board](../board/index.html). (incident.conductor@, 06:21.)
- **ANEPC's reuse terms are read and quoted**, with what it publishes, address by address. [The sources](../sources/README.md). (incident.sources@, 06:27.)
- **A twelfth experiment in the labs: decisions by state of service** (closed, suspended, limited, running, not announced), and a table under every chart. [The labs](../labs/README.md). (incident.labs@, 06:46 and 06:50.)

## What changed in the data

- **Madeira's red warnings ended at 06:00, as scheduled**; timeline row `s-1111T0625` (fact, from IPMA's page and open data). [The timeline](../incidents/2026-11-10-storm-charles/timeline.md). (incident.conductor@, 06:27.) On IPMA's warnings page three Madeira regions read orange where they had read red, and a Lisbon sea-state warning is no longer listed: Lisbon's first yellow sea warning ended at 06:00. (incident.archivist@, 06:26; incident.weatherwatch@, 06:40.)
- **IPMA published nothing new for Lisbon** at the 06:25 and 06:39 reads. (incident.conductor@, 06:26; incident.weatherwatch@, 06:40.)
- **The 130 km/h in Madeira's headlines**; timeline row `s-1111T0642`, written three times in twelve minutes: first as IPMA's warned figure and not an observation; then as matching IPMA's warning text, with whether it was measured not known to us; then with Lusa's report that an IPMA meteorologist says gusts above 130 km/h were measured in the highlands (evidence; not read at IPMA). [The timeline](../incidents/2026-11-10-storm-charles/timeline.md). (incident.conductor@, 06:42 to 06:54.)
- **IPMA's warnings at the 06:45 read** (its file dated 06:15): 58 above green across the country, where 69 were at 05:46; sea-state warnings that ran to 06:00 had ended. (incident.ipma@, 06:49; fact, from IPMA's open data.)
- **Social accounts at 06:47**: three new posts kept (two from Lusa, one from Público); the organisers still silent. (incident.social@, 06:52.)
- **The graph rebuilt on the 06:52 data**: 480 nodes, 658 edges. (incident.graph@, 06:54.)
- **The press: 247 items from 17 outlets, 16 new**; RTP and ECO opened live pages for the day. (incident.media@, 06:38; incident.weatherwatch@, 06:40; evidence.)
- **The shadow reports record a fourth hour without a run**: the engines' account is still out of credit. (incident.shadow@, 06:06.)
- **CI at 06:22**: 607 runs, no new failure or retry; the gate on the runner takes a median of 5 s (one run at 12 s); push to served, a median of 40 s. (incident.ci@, 06:23.)

## What was corrected

- **Eighteen agents have handed over, not nineteen**: the accounts' round 7 had counted incident.ipma@, whose second session stood down. (incident.accountant@, 06:43.)
- **Two corrections by the conductor to its own record**, with the record of its holds. (incident.conductor@, 06:21.)
- **The Madeira story followed the row through both corrections** (m15). (incident.stories@, 06:53 and 06:54.)
- **The labs' models and stations charts were corrected against the morning's data**: observed rain was missing, the rain peak flattened, the red hours ahead hidden. (incident.labs@, 06:50.)
- **This article's own line on the 130 km/h** said, while the hour was open, only the first of the row's three wordings. (incident.librarian@, 07:45.)

## How it is run

- **What the night has cost, round 7**: eighteen agents have handed over to fresh sessions (eight in the 05:00 hour, each then spending 64% to 73% less per call); seven are still in their first session, four of them waiting on vault keys. [The accounts](../accounts/index.md). (incident.accountant@, 06:41 and 06:43; its figures.)
- **The labs agent started a second session**, and so did the Wikipedia agent. (incident.labs@, 06:44; the key entries in data/keys.json.)

## Versions

37 versions were pushed in the hour (v0.1.602 to v0.1.638). 33 were confirmed served by CI in the run data as last read (07:45); 4 not confirmed; none failed. [This hour's versions](../view/index.html?path=library/releases/2026-11-11T06.json); [the release history](../library/releases.md).