Cette page n'est disponible qu'en anglais.
Crawler and feeds
Mera is not a crawler. Its news comes only from the RSS and Atom feeds that publishers choose to publish, which usually hold a little text and a link back. It doesn't crawl sites for news or copy articles.
Mera stands on the publisher's side against full-content crawlers. It exists to send readers to publishers.
What the fetcher does
| Property | Value |
|---|---|
| User-Agent | MeraNewsBot/1.0 (+https://mera.news/bot) |
| What it requests | The one feed URL you published |
| How often | Once an hour |
| Conditional requests | Sends If-None-Match and If-Modified-Since |
If your feed hasn't changed, your server can answer 304 Not Modified and send nothing.
The fetcher doesn't render pages, fetch images or follow links. It also doesn't fetch robots.txt, since it only requests the feed URL you published. To stop it, use the steps below.
Mera has also run a separate tool, started by hand, that looks for publishers' subscription pages so the app can link readers to them. It requests a publication's homepage and a few likely pages, at most 6 requests per site in a run, with the same User-Agent. It collects no articles.
The bot page is the short version for people reading their access logs.
Which feeds Mera reads
The full list is public in the feed registry on GitHub. You can search every publication in the directory, or fetch it as data from the public API.
What goes in and what comes out
Mera documents the boundary of its news service, not how it works inside.
| Direction | What |
|---|---|
| In | The title, link, description and details in a published feed |
| Out, to the app | Headline, link, image address, publication, country, language and the feed's description, which is used to judge relevance and never shown |
Inside, Mera's servers use feed text to translate, describe and match articles to topics. Some of this runs through the service providers named in section 8 of the privacy policy. The results are the same for every reader and carry no user identifier.
Article records are deleted after 48 hours. Article addresses are kept for 90 days to spot repeats.
What readers see
- The headline, verbatim
- A short note from Mera on why it may matter to them
- A link to the publisher
- The image your feed points to, where it has one. The phone loads it directly from that address
Readers don't see the article body or the feed's description in Mera. The content policy sets out the full position, including its legal basis.
Getting into Mera, or out
To add a feed, use the contact form on the publishers page or open a pull request on the feed registry.
To leave, do either of these:
- Filter or rate-limit the User-Agent
MeraNewsBot. Mera sends it plainly, so matching on it is reliable. - Write to
publishersupport@mera.news, or open an issue on the feed registry.
If you run bot protection, check that it doesn't block your feed by accident, unless that's what you want.