RSS Feed

An RSS feed is a file that keeps a website constantly up to date. It contains information about which posts were most recently published there. For each post, you’ll usually find a title, date, a short summary, and a link to the full version. This file isn’t made for humans, but for programs: it’s built strictly according to fixed rules so that software can read it reliably. Such a program fetches the file anew every few minutes and compares it with the last known state. If an entry shows up that it doesn’t recognize yet, it flags it as a new article. The abbreviation stands for “Really Simple Syndication.”

Fetching news instead of searching for news

Without feeds, you’d have to visit every site of interest individually and check whether there’s anything new. With twenty sources, that’s half an hour of work every morning. With feeds, a single program collects all the news in one place. You get a list, sorted by time, spanning all your sources.

The second advantage is control. You decide yourself which sources to read by subscribing to them. There’s no algorithm reshuffling the order based on click probability, the way social networks do. You see everything that appears, and only what you’ve signed up for. That’s precisely why RSS feeds have been experiencing a small comeback in recent years among people who distrust social networks.

For AI system providers, feeds are especially practical. Anyone wanting to feed a language model with up-to-date news needs the articles in a reliable, uniform structure. A feed delivers exactly that, whereas scraping ordinary web pages remains laborious and error-prone.

The structure of the feed file

Technically, a feed is a text file in XML format. XML means that every piece of information is enclosed by tags, like this: <title>Headline</title>. This lets a program clearly identify which part of the text is the title and which is the date. Humans can read the file too, but it looks messy.

The file resides at a fixed address on the website’s server, often ending in “/feed” or “/rss.xml”. You enter this address into a reading program, known as a feed reader. The reader queries the address at fixed intervals, for example every fifteen minutes. Each entry has a unique identifier so the same article doesn’t get counted as new twice.

Usually a feed contains only the last ten to fifty posts, not the entire archive. If you don’t check for a week, you’ll miss older updates on very active sites. Alongside RSS there’s the newer format Atom, which is technically more cleanly structured. In everyday use, though, most people call both “RSS”, and reading programs understand both formats anyway.

Podcasts, newsreaders, and news sites

You most often encounter RSS without noticing it: with podcasts. At its core, every podcast is a feed in which audio files are linked instead of articles. The app on your phone checks the feed and downloads new episodes. That’s why you can listen to a podcast in any app you like, without being tied to a single provider.

Anyone who wants to use feeds deliberately needs a feed reader. Well-known services include Feedly, Inoreader, or the open-source Miniflux, which you can run yourself. Many news sites, blogs, and even government agencies offer feeds, sometimes split by section. Financial sites often have a dedicated feed just for market news.

A common misconception is that RSS is dead. Google discontinued its popular Reader in 2013, which many people took as a sign of this. But the technology continues to run unchanged and carries the entire podcast world. It has simply become less visible because it works in the background.