Feeds already checked

Addresses we fetched and parsed ourselves, not copied off a list. All 266 sites.

What the report says

Each check is one line: what was tested, what came back, and what to do about it. The checker looks at the connection (status code, redirects, HTTPS, content type, encoding, response time, conditional requests), the XML itself (well-formed markup, with the line number when it is not, and the unescaped ampersands that usually cause it), the structure required by RSS 2.0, Atom 1.0, RSS 1.0 and JSON Feed, and the items (date formats, unique identifiers, working links, relative addresses, publishing order, freshness). Feeds with audio are also checked against what Apple Podcasts and Spotify require.

Level	Meaning	What to do
Error	Feed readers are likely to reject the feed or skip items.	Fix these first.
Warning	The feed works, but some readers or directories will have trouble.	Worth fixing when you can.
Note	A recommendation or an observation.	Optional.

Every check is described on the list of checks, with the reason it matters and the usual fix. Give the box a home page, a YouTube channel, a subreddit or a GitHub repository instead of a feed and it finds the feed first.

Questions

What is the difference between a feed checker and a feed validator?

None that matters. A validator tests the feed against the format rules. CheckRSS does that and also tests the things the rules do not cover but feed readers care about: whether the address loads, how it redirects, the encoding, and whether the newest items look right.

My feed is valid but my reader does not update. Why?

Usually caching in front of the feed, a redirect the reader does not follow, a firewall that blocks the reader's servers, or items whose identifiers change on every request. See why an RSS feed stops working.

Do you store the feeds I check?

No. The feed is fetched, passed to your browser and analysed there. Nothing is saved on our side, and feeds that need a password cannot be checked at all: only public addresses on the standard web ports. Details are on the privacy page.

Is there a size limit?

Feeds up to 25 MB are analysed in full. Larger feeds are cut at that size and the complete items are still checked. The JSON API analyses the first 512 KB.