Skip to content
Show Almanac

Sources

Where every fact comes from.

13 sources enabled, 1 switched off. Every feed source was fetched live before it was written down, and its published terms were read wherever the operator makes them readable — a working endpoint is not a permission. A manual source is never fetched: a person reads the venue’s public page and types the facts in, and the freshness shown is when a person last looked.

Note

Sources that were examined and rejected

Seven publishers were fetched and turned down, each with the publisher’s own words as the reason: three library systems that disallow every unnamed crawler site-wide, a public radio station that names AI crawlers individually and refuses them, a tourism site whose terms sit behind a bot challenge, a university whose only data path is excluded, and one large open map database that refuses automated fetching on all three of its access paths.

That ledger is kept in the repository at docs/sources/VETTING_LEDGER.md, so a rejection can be re-checked when a publisher changes their terms — and so nobody re-surveys ground already covered.