The Many Broken Feeds | Hacker News Reader

The Many Broken Feeds

notes.abhinavsarkar.net

37 pointszdw19 comments

> There is also an option to put your website behind a CDN front, such as Cloudflare or Github Pages, and let the CDN provider deal with renewals.

The biggest cause of broken feeds for me is actually Cloudflare challenges on the feed URL.

I've seen those in the past for RSS feeds from Substack - https://simonw.substack.com/feed for example - which is frustrating because the whole point of an RSS feed is to support polling by automated systems!

Yes, I came here to say that. My RSS reader runs as a daily GitHub action, which looks enough like a bot to get blocked.

https://github.com/robalexdev/Feeds

I collect a lot of feeds, I see each of these a lot. It'd be great if there was a service to politely notify the site owner, as I'm sure many don't realize their feed is broken / incorrect / missing.

That's a great idea! Maybe one of the feed aggregating search engines can do this. Feeds often come with publisher emails so it should be possible to automate.

How do people discover new feeds? Are there any nice collections of blogs or are people just accumulating them over the years based on HN posts?

Google News Alerts

If a website looks like a blog there's a high chance it has a feed and is probably using HTML metadata to advertise it, like this:

  <link rel="alternate" type="application/atom+xml" title="Atom" href="/atom/everything/">

Most feed readers know how to read that, so you can paste the URL of the blog itself into e.g. NetNewsWire and it will locate that rel="alternate" link and subscribe that way.

Yeah, but this doesn't solve the question of actually finding relevant blogs with feeds in the first place.

Follow links to articles on blogs from places like Hacker News. If you like what the author is saying, subscribe - then watch out for other blogs they reference in their future writing.

Try to find link blogs, since those are more likely to help you discover other things to read.

I view the webpage source and search for the strings "rss", "atom" and "feed" until I find one. I currently have 145 feeds. I follow blogs, news, I get updates about local events, I follow youtube channels, I get notified about new releases from various github repos, follow updates on various status pages, follow various subreddits etc.

Feed discovery is a challenge, but there are a number of current efforts to make it easier. Far too many to list here. Briefly:

There are curated lists:

- https://blogroll.org/

- https://minifeed.net/

Community lists (planets)

- https://planet.debian.org/

- https://www.planetpython.org/

Curated themes (like Bluesky starter packs):

- https://minifeed.net/lists

- https://www.youneedfeeds.com/starter-packs

Webrings are cool again:

- https://webring.xxiivv.com/

- https://brisray.com/web/webring-list.htm

This collection of HN blogs is pretty good too:

- https://github.com/outcoldman/hackernews-personal-blogs

- https://blogs.hn/about

I've been trying to aggregate many of these into a giant list, https://alexsci.com/rss-blogroll-network/blogrolls/, which is especially focused on OPML blogrolls: https://alexsci.com/blog/blogroll-network/

Wow, thank you for this, that is a really great collection of resources!

I ran an RSS aggregator ~15 years ago. All these things were still problems back then… also FeedBurner was a complete plague, at least that’s mostly gone away.

I’ve been trying to work out a new approach to newsreaders for a couple years now, and have all but given up on RSS. The standard needs to improve in significant ways — most notably with discoverability — if it ever wants to get back to the Google Reader “glory days”.

I’ve tried reporting bugs to rags like The New Yorker and The Atlantic, each one is a a black hole. Unless you can make a case for helping monetization, they don’t care at all, and I can’t say I blame them.

For me the biggest problem is sites that truncate their feeds. As an example the "the verge" feeds are basically useless now.

I don't mind truncated feeds too much, as long as there is an explicit notice at the end that the full content is available on the website.

Try a feed reader that can download the source article for you (similar to reader mode in web browsers)

The Verge is heavily paywalled now. They won't put their full articles in the feeds, since they want you to pay.

> There is also an option to put your website behind a CDN front, such as Cloudflare or Github Pages, and let the CDN provider deal with renewals.

Or they could just use a modern webserver setup? Manually updating certs on webservers in 2025 is just a massive waste of time.

Also of the opinion many a personal site being dynamic is less than ideal. How many outdated php sites I've seen hijacked, ruined when their hosting service updates things behind their back, run into resource related issues, caching bodges that go wrong, etc. Meanwhile the site maybe gets updates like once a week.

At the very least the RSS feed certainly shouldnt need to be generated on every request unless theres good reason.

Well aware my complaints yelling at the void as many just want a prebuilt solution which 'just works', even if it requires a LAMP stack.

Reply on news.ycombinator.com