Feeds

Four of them, and they answer different questions. There is no email list you have to join and nothing here tracks you — a feed reader fetches a file and that is the whole arrangement.

• /feed.xml RSS

  The blog. Mostly incidents and what they turned out to mean. The newest ten posts carry their full text, so a reader gets the article rather than a teaser.

• /feed.json JSON Feed

  The same posts, for readers that prefer JSON. RSS stays the canonical one because every reader understands it; this costs nothing extra and some people would rather have it.

• /queue.xml RSS

  What the team is working on. Open Discussions from the engine repository, newest first. That repository is private, so titles and summaries are redacted before they leave it and items link to the queue page rather than to a thread you could not open.

• /ship.xml RSS

  What shipped. Every merged pull request, with its diff size, its merge-gate labels, and whether review sent it back. Of the four this is the one where every item is a finished, checkable thing.

Machine-readable, too

/metrics.json is the derived metrics behind the metrics page as a documented payload, so the numbers can be checked by script rather than by eye. /llms.txt is a plain-text map of the site. /sitemap.xml lists every page with the date its content last actually changed.

Everything above is derived from something public. If a figure in a feed disagrees with the repository, the commands to settle it are on the verify page, and the disagreement is a bug we want.