RSS Anyway is a preview of an RSS reader that works for every site on the web, including the sites that don’t publish a feed.
The pitch is in the name. RSS is the best way to follow the web — one reader, all your sources, no algorithm deciding what you see — and the problem everyone has hit is that most sites don’t offer feeds anymore. Your favorite blog, that company’s news page, the forum you check every morning: no feed, no way to subscribe. RSS Anyway builds one for you.
That is the whole product. Paste a URL into RSS Anyway and it returns a feed — either the one the site already publishes, or a synthetic one assembled from whatever structured signals the site does publish. The feed looks and behaves like a normal RSS feed because it is a normal RSS feed. Your existing reader doesn’t need to know where the data came from.
The classifier runs in order and short-circuits on the first hit. First, known
patterns: domain-specific shortcuts for Discourse, arxiv, YouTube, GitHub, and
a handful of other platforms that publish feeds at predictable URLs. Second,
link-scan: walk the URL from most-specific to origin in parallel and look
for <link rel="alternate"> tags or feed-shaped <a href> anchors. Third,
common-path probes: fire parallel requests at /feed, /rss.xml,
/atom.xml, and friends. Fourth, sitemap extraction: fetch /sitemap.xml,
group by first path segment, filter to blog-shaped sections like blog,
posts, news, writing, articles, essays. Fifth, index-scrape: the
last resort — find <a> tags with chrono-parseable dates on the site’s own
index pages, group by path segment, require at least three items with distinct
titles before we call it a feed. No LLMs in the pipeline. Pure heuristics and
content extraction with Readability and linkedom.
The homepage at / is a Reddit-style hot ranking of whatever’s trending across
every feed in the index, recomputed every five minutes. The /new view is
reverse-chronological. Every individual feed exports as RSS, Atom, or JSON.
There is no personalization, no login, no tracking, and no “for you.” The point
of RSS Anyway is to give you back the ability to follow the web on your terms,
and that goal is incompatible with an algorithm quietly deciding which posts you
see. You pick the sources. The reader shows them to you.
Every URL that has broken ten points on Hacker News is already a feed on RSS Anyway. The site doesn’t know. The author doesn’t know. We just made one. That is how the index bootstraps without a submission form. Hacker News is a surprisingly rich discovery seed — every story that gets traction is a URL someone thought was worth sharing, and most of those URLs live on sites that publish more than one post. When a story crosses the threshold, we classify the source, add it to the index, and poll it on an adaptive schedule so frequent publishers get checked more often. Three hundred items stay warm on the hot page at any moment, all of them discovered this way.
One thing is missing, and it’s the obvious one: you can’t submit your own links
yet. Today, discovery is seeded entirely from Hacker News. That is a
surprisingly large chunk of the useful web, but it is not the whole web, and the
submission form is the next feature. When it ships, anyone will be able to paste
a URL and get a feed back — and at that point, ryanxcharles.com/news is
deprecated and will redirect to RSS Anyway. The /news aggregator was the
prototype for this. RSS Anyway is the version that works for everyone, not just
me.
Until then, browse rssanyway.com to see what’s already been discovered, watch the status page if you want to see the crawler work in real time, and export any feed you find into the reader you already use.
Subscribe to anything.
