Skip to main content

RSS and Atom Feeds

Last updated on by cernus76

The site publishes machine-readable feeds in both RSS and Atom so readers can follow new content from a feed reader instead of checking back by hand. There are two separate feed systems, and they are built in very different ways. This page explains both, and how the content you contribute ends up in them.

Two Feed Systems​

ContentFeedsHow it is produced
News (the blog)/news/rss.xml, /news/atom.xmlBuilt into Docusaurus
Reference architectures/docs/ref-arch/rss.xml, /docs/ref-arch/atom.xmlCustom plugin in this repository

News Feeds (Built In)​

The news section is a standard Docusaurus blog, and Docusaurus generates blog feeds out of the box. There is nothing bespoke here: the feeds are enabled through the blog preset in docusaurus.config.ts.

blog: {
    path: 'news',
    routeBasePath: 'news',
    // ...
    feedOptions: {
        type: ['rss', 'atom'],
        title: 'SAP Architecture Center News',
    },
},

With type: ['rss', 'atom'] set, Docusaurus writes rss.xml and atom.xml under the news route at build time and lists every published article. As a contributor you do not have to do anything: publish a news article in the normal way and it appears in the feed automatically.

Reference Architecture Feeds (Custom Plugin)​

Docusaurus only generates feeds for the blog. It has no feed support for ordinary documentation pages, and the reference architectures are ordinary docs. To serve the readers who rely on feeds to keep up with the catalogue, we built this ourselves.

Prefer to browse?

If you would rather read the recent updates in your browser than in a feed reader, the Recently Updated Reference Architectures page shows the latest reference architectures as a web page, with the same information the feeds carry.

The feeds are produced by a custom plugin at src/plugins/docs-feed-generator, registered in the plugins array of docusaurus.config.ts. It uses the feed package to build valid RSS and Atom documents.

How It Works​

The plugin runs in the postBuild phase, after Docusaurus has resolved every permalink and written the per-doc metadata. For each reference architecture it then:

  1. Reads the resolved doc metadata and keeps only pages under a docs/ref-arch/RA####/ directory.
  2. Sorts the surviving pages by last_update.date, newest first.
  3. Takes the 50 most recently updated entries.
  4. Builds each feed item from the page title, description, a plain-text excerpt (the first 500 characters of the body), the permalink, and the update date.
  5. Writes rss.xml and atom.xml into the docs/ref-arch output directory.

Because this happens during npm run build, the feeds are only regenerated when the site is rebuilt and redeployed, not live on every content change.

What Gets Excluded​

A reference architecture page is left out of the feeds when any of the following is true:

  • It has no last_update.date in its front matter. A date is required to place the entry on the timeline.
  • Its front matter sets draft: true.
  • Its front matter sets unlisted: true.
  • Its tags list contains archive or demo.
  • It is the RA0000 demo page.

So if a reference architecture you expect is missing from the feed, check its front matter first. The most common cause is a missing last_update.date, an archive tag, or a page left as draft.

# This page would be excluded from the feeds
tags:
    - archive
draft: false
unlisted: false
last_update:
  author: your-github-handle
  date: 2026-10-28

RSS or Atom?​

Nearly every feed reader understands both formats, so the choice rarely matters to a reader. We publish both for compatibility: RSS is the older and more widely recognised format, while Atom is stricter about dates, authorship, and how content is encoded.

References​