What was being written.
Every headline this site's collector has seen, kept. The Feeds page shows the current window; this is the whole record behind it, searchable by term, source and date. Headlines are the publisher's own words, linked to the publisher's own page. No article text is stored, nothing is summarized, and nothing is ordered by anything except date.
Read the record
Filter by category, source and date. Type a term to search titles and sources; a term prefixed with a minus excludes it. Your filters live in the address bar, so any view here is a link you can keep.
What this is, and is not
The same disclosure the Method page makes, restated where it applies.
Public RSS and Atom feeds are fetched on this site's own server every six hours. What is kept is the title the publisher syndicated, the link, the publisher's name and the date. Nothing else. Following a link leaves this site and goes to the publisher, who is the only party that can give you the article.
A feed is a window, usually a few days deep, and what falls out of it is not offered again. So this archive can only ever hold what was captured while it was there. It starts on the day the collector started keeping, and it has no opinion about what came before.
Some publishers send no date with an item. Those are filed under the day this site first saw them and are marked seen, so a fetch date is never mistaken for a publication date.
Headlines get rewritten after publication. When the wording behind a link changes, the first wording stays as the record and the new one is kept alongside it. That is the same rule the change record follows: publish the sequence, never a verdict about it.
There is no ranking here. No score, no relevance, no editor's pick, no trending. The only orders available are newest first and oldest first, which is the whole of what this page knows how to do.