Thanks - that’s a useful feed, but it’s also not really as good, because it’s somewhat ‘unstructured’ content. I would have to parse the title element, and that relies on human consistency - as soon as someone is inconsistent with the version number (perhaps putting a ‘v’ or ‘version X.X.X’ or putting it not at the start of the title) it could break.
This is why the strict, structured data from Github was really useful. Is there any reason it was removed?
Anyway, I’ll maybe try and rely on /tags at Github and just try and detect a higher version or something. Thanks nonetheless