For example, I wrote a feed generator that targets a specific software package release page. It scrapes the page and converts it to an Atom feed [1]. I then wrote a matching systemd service, which triggers the scraper once a day [2] and writes it to a file, no hosting required. My RSS reader is pointed to that file.
[1]: https://github.com/claui/release-feed-mediola/blob/fca39e897...
[2]: https://github.com/claui/release-feed-mediola/blob/fca39e897...