What this is

Most “on this day” sites give you a list. Lists are useful, and that ground is already well covered. This site does the other thing: each of the 366 dates gets a single continuous narrative that reads the day end to end — what happened, who was born, who died, and how those entries sit against one another across the centuries.

The argument is simple. A date is an arbitrary container. But reading one in sequence, rather than scanning it as bullets, is how the accumulation becomes visible.

Every event, name, year and place on this site comes from a cited source. Nothing is written from memory or inferred. The generator pulls from two places:

• The Wikimedia On This Day feed — the curated selected set plus the wider events, births, deaths and holidays lists for each date. Every item used links back to its Wikipedia article.
• The Library of Congress “Today in History” collection — linked on every date page for readers who want a second, editorially written account.

The narrative quotes the source text for each item and only adds connective phrasing — transitions, era framing, and intervals arithmetically derived from the cited years. It never introduces a fact that is not in the source.

The feed is global, so items are scored for American relevance before selection. Named institutions and founding-era terms score highest, state names lower, and subjects belonging to other histories — Roman, Byzantine, papal, dynastic — score negative. Anything before roughly 1500 is discounted regardless of keywords.

Births and deaths are only included when the nationality recorded immediately after the person’s name is American, which avoids false positives from phrases like “user of American Sign Language.” People are ordered oldest first, so historical figures surface ahead of living celebrities.

Some dates are genuinely thin in the American record. Those are labelled thin record rather than padded out with unrelated world trivia.

The site is a Next.js static export: 366 date pages, twelve month pages, a calendar index and this page, all rendered to plain HTML at build time. There is no CMS, no database, no accounts, no comments and no API call at request time. No environment variables or secrets are needed to build or deploy it.

Content is generated once by a script and committed to the repository as markdown, so the archive is reproducible and reviewable in version control. Regenerating is idempotent: existing files are left alone unless the run is forced.

Because the export is static, the home page is baked with the build date. A small script checks the real Eastern-time date on load and forwards you to the correct story if the build has gone stale, and every date remains reachable at its own permanent URL.

America On This Day is an independent project. It is not affiliated with, endorsed by, or connected to the Wikimedia Foundation, Wikipedia, or the Library of Congress. Their feeds are used under their own terms, and every page links back to them.

There are no generative images of real people, no synthesised audio and no AI video anywhere on the site. The decorative plate on each date page is drawn in CSS rather than fetched, so it cannot break a layout by failing to load.

Read today’s story, browse the full calendar, or press / anywhere to jump to a date.