SMF·REEDER
Can a prompt replace Reeder?
Read it later & RSS — read-later, bookmarks and RSS
Exhibit tracking slip
Verdict
Reeder's redesign threw out the thing every other reader is built around: the unread badge. What is left is a single timeline you dip into, mixing RSS, YouTube and podcast feeds, which you scroll like a social feed and never feel behind on. That is a small, opinionated app and it builds in one sitting, because the hard problems in feed reading are all in the parts Reeder deliberately dropped. What you cannot rebuild is that it is a native app that feels right on an iPhone.
Exhibit A — The prompt
Received on31.07.2026Build a local unified timeline reader. One chronological river of feeds, videos and podcasts, with no unread counts and no inbox metaphor anywhere in the interface.
Subscriptions: accept a feed URL and detect its kind. Plain RSS or Atom is an article feed. A YouTube channel or playlist URL resolves to that channel's feed and produces video items. A podcast feed produces audio items with enclosures. One subscription list holds all three; the type only changes how an item renders.
Fetching: poll on an interval derived from each feed's own publishing rate, using conditional requests. Deduplicate by GUID, then canonical URL. Store the full item: title, author, published date, content or summary, and the enclosure URL for media.
The timeline is the product. One reverse-chronological list of everything, with no per-feed unread badge, no folder counts, and no "mark all as read" — because there is nothing to be read. Items are marked seen implicitly as they scroll past, purely to control what the app opens on next launch. A "catch up" gesture jumps to the newest item without any bookkeeping.
Rendering: articles show the feed's content inline, with a "fetch full text" action that runs Readability against the source URL on demand rather than automatically. Videos embed the player. Podcast items get an inline audio player with playback position stored per item and a 30-second skip.
Saving: a star marks an item for later, and starred items are the only filtered view in the app. Full-text search across everything fetched, using SQLite FTS.
Storage: SQLite. Import and export OPML, plus a full JSON dump.
Out of scope: native mobile apps, syncing with an external reader account, offline media caching, folders, and any unread count anywhere. The absence of those counts is the design, not an omission.
Opening prefills the prompt — press enter to run it.
Exhibit B — What you lose
- B.1 the native iOS and macOS apps and their gesture handling
- B.2 iCloud sync of read position across devices
- B.3 sync with an external account like Feedbin or Inoreader
- B.4 offline caching of articles and audio for a commute
Prior art
Exhibit C — Why people still pay: capture polish, sync, and content partnerships
Because a reading app lives or dies on how it feels under a thumb, and a one-dollar-a-month native app that gets that right is doing something a web build in a mobile browser cannot.
Questions
Can I bring my Reeder subscriptions over?
Yes. Reeder exports OPML including its YouTube and podcast subscriptions, and the importer reads all three kinds from the same file. Read position and starred items do not export, so you start with a clean timeline.
Does dropping unread counts actually change anything?
It is the whole point. A reader with unread counts becomes a queue you owe work to, which is why people abandon them. A timeline you scroll and leave has no debt attached, and after a week of using it the difference is obvious.
What does it cost to run?
Nothing. It runs locally against a SQLite file with no server and no account. Podcast audio streams from the publisher rather than being cached, so there is no storage growth beyond the text.
What is the one thing that does not survive the rebuild?
The phone. Reeder is a native app with sync, offline caching and gestures that make it pleasant to use standing on a train. A local web app on a laptop covers the desk and nothing else, which for this particular product is most of the value gone.
Related tools
Receipt