Every source you follow, in one edition you finish
A feed with a ceiling
You set how many items an edition holds. Taper builds it that size from the RSS feeds, newsletters and channels you already follow, then it ends. There is nothing underneath the last one.
Free, and you can bring every source you already follow
An edition, built from sources you can check
Fifteen real sources, a newsletter among them, blended the way your own edition is
Try it with the sources you actually read.
Three sources and a size of five in each of these, and a source is anything that publishes a feed: a newspaper, a video channel, somebody’s newsletter, a weblog. Yours can be four hundred of them at a size of fifty, or one at a size of five: the edition is as long as you decide, you keep one for each subject you read, and every one of them still ends.
Sources in, one edition out
You make the first two once. The third is every edition.
Add what you already read
Paste any address and Taper finds the feed behind it, or import an OPML file whole. Your favourites go on always and are never counted against the size, and everything else takes fair turns, so a source posting forty times a day cannot crowd out one that posts weekly.
Set how long an edition is
Five items, or fifty, on a dimmer with six stops. The size belongs to the edition rather than to the app: Morning and Everything I follow are not the same length, and each keeps its own.
Read it, and reach the end
It is as long as you set it, plus whatever your always sources published. Nothing loads when you get to the bottom, and no number anywhere is waiting to be brought down. Anything worth a second reading goes to Saved with one key.
A source is anything that publishes a feed
- The Guardian
- New York Times
- BBC News
- The Verge
- Ars Technica
- Hacker News
- Stratechery
- Polygon
- IGN
- YouTube
- Substack
- GitHub
- and many more
- theguardian.com
- youtube.com/@channel
- name.substack.com
- daringfireball.net
- github.com/owner/repo/releases
- news.ycombinator.com
- j k o s
- OPML in and out
- Keyword filters
- AAA on body text
What Taper will never do
Product definition rather than a roadmap. None of these comes back later behind a setting.
- No adsNot now, not after a funding round, not as a sponsored item inside your edition. The subscription is the whole business model.
- No unread count, badge or dotAnywhere, including the app icon.
- No infinite scrollAnd nothing waiting to load at the end of an edition.
- No algorithm deciding what risesYour sources and their tiers decide.
- No notification about an itemNot for a headline, not for a source, not for a number that went up. Your edition is there when you open it, and it does not come looking for you.
- No streaksAnd no reading stats dressed up as an achievement.
What falls outside it is what you would have scrolled past.
Taper keeps no copy of an article you did not save. Anything still current arrives on your next refresh, anything worth a second reading goes to Saved with one key, and anything older is where it always was, with the publisher, who never stopped hosting it.
Free until you outgrow it
One paid tier, and nothing held back from the free one to make it look thin.
Free
Free
- Sources
- Unlimited
- Editions
- 15
Not a trial, and not a demo
The whole app. Bring every source you already read: the edition size does the metering, so the list never needs a ceiling of its own.
- The dimmer, at every stop from 5 to 50
- Source tiers: always, normal and background
- Keyword filters on every edition: keep a word out, or keep only what matches
- Saved items, kept as long as you want them
- OPML import, and feed autodiscovery from any URL
- Google Reader keys: j k o s
- Light and dark
- No ads, nothing tracked in the app, and nothing sold
Pro
$5 / month
- Sources
- Unlimited
- Editions
- Unlimited
Filters that do more than match a word
For readers whose filtering outgrew a keyword list, a hundred sources in, when blocking one word at a time stops being precise enough.
- Everything in the free tier, unchanged
- Logic in a filter:
(rust or zig) and not hiringkeeps posts about either language and drops the job ads. - Match on the headline, the summary, the author, the publisher, or a tag
- Match how a word begins, not only the whole of it
- Filter on what an item is, not only what it says: keep the videos out, or hold an edition to what was published today
- Filter sets you save once and reuse across editions
- Mute a word until a date you choose
- Read a filtered edition in the RSS app you already use
Questions you are right to ask
The objections first, then the mechanics, then the name and the type.
Isn't a fixed size just hiding my news?
It is deciding what to read, which you were doing anyway: with a scroll, at the bottom of a list, badly. The size takes the top of what your sources published by the rules you set: favourites on always are never counted at all, and everything else takes fair turns.
What happens to an edition I don't get to?
Nothing. Open it again and it has been rebuilt from what your sources published since. Wednesday’s paper is not Tuesday’s held over. There is no pile to return to and no count of what you missed, because a backlog is the thing that makes people abandon a reader. Anything you want to keep, you save.
Does this still work if I follow a hundred and fifty sources?
That is the case it was built for, and the source list will never carry a number in any tier. Sources you never want to miss go on always and are never counted against the size. Everything else takes fair turns for the slots, and background sources fill what is left over, so a long list gets you a better edition rather than a longer one.
Can I have more than one edition?
As many as the subjects you read. An edition owns its sources, its filters and its size, so Morning can be ten items from four papers while Everything I follow is fifty. Once there are enough to scroll past, shelves group them. The free tier holds fifteen editions, set where a reader running two or five or nine never meets it.
Can I search something I read last month?
Saved is searchable, and it is the only thing there is to search: Taper keeps no copy of an article you did not save. That is the same decision as the ceiling. The size is the judgement about what was worth reading, and a second place for everything that missed the cut would only mean owing it something later. Anything current comes back on the next refresh, and anything older is still at the publisher, who never stopped hosting it.
What happens when one of my sources breaks?
The edition still arrives, with a line at the top naming the source that didn’t answer, the exact address it tried, what came back, and the way to the screen where you fix it. Publishers move feeds without telling anyone, and a source failing quietly for a month is the one problem a reader cannot debug, so nothing about it is smoothed over.
Is there an app for my phone?
Not in an app store yet. Taper is a web app built to be read on a phone: it holds down to 320px, it installs to the home screen and opens without the browser around it, and it is the same account either way. Native apps are planned, and a store listing announced before it exists is a promise on somebody else’s schedule.
Can I read an edition in the app I already use?
Yes. An edition can be served at its own secret address as Atom and JSON Feed, so the blend, the source tiers and the filters land in whatever you already read in. What does not travel is the ending: your client polls that address, keeps what it has seen and puts a figure beside the folder, which is what clients do and is not something this end can change. The filtering is the part that reaches you there.
Building one feed out of many, and what it can and cannot do
Where does my reading live?
On our servers, and only what the product needs to work: your sources, your tiers, your filters, and whatever you saved. Articles are fetched from the publisher when your edition is built rather than warehoused, so Taper keeps no copy of what you read. Nothing is sold, and there is no ad network to sell it to.
How do I get everything back out?
Every source and every edition, as one OPML file, whenever you want it, and it imports into any other reader. Closing the account deletes it rather than deactivating it. Being easy to leave is the part of a subscription that usually gets made hard on purpose, so it is plain here and it stays plain.
How is this different from Feedly or Inoreader?
Both hold everything and give you tools to sort it afterwards: rules, folders, search, and a figure for what is still waiting. Taper fills the number of items you set and stops. The honest version is that they are more capable and this is more finishable, and which of those you want is a real disagreement rather than a marketing one. Both count your sources on the free tier, at a hundred and a hundred and fifty. Taper counts none, on either tier.
Why is it called Taper?
A taper is a candle: a small contained light with a wick that meters how much fuel reaches the flame. It is not about weaning yourself off the news. Taper is a steady amount of it, indefinitely.
Why does it look like this?
It is set in Atkinson Hyperlegible Next, drawn by the Braille Institute so that characters cannot be mistaken for one another. A product arguing for clarity over volume should be set in the typeface that exists for clarity.
Decide how much news is enough.
Bring the sources you already read. Set a size. Reach the end of it, and find nothing underneath.
Free. No card, and bring every source you follow.