osmosfeed VS Invidious

You may be interested in Osmosfeed: https://github.com/osmoscraft/osmosfeed

It is a static site feed aggregator primarily designed to go with GitHub Pages. I host one to aggregate my own writing on different sites. I think it may fit your use case because your Osmosfeed site itself outputs a single Atom feed. So, for example, if I have an Osmosfeed site that aggregates feeds 1, 2, and 3, the Osmosfeed site has a single feed which will include the three individual feeds. Mine has about 10-12 feeds and it has worked perfectly thus far with no issues. Not sure if it would have problems at higher numbers.

I have been using osmo feed [0]. Rather than self hosting, it uses GitHub actions. I have my own usename.github.io/to-read linked to it, to access it from anywhere. So far I have liked this approach.

[0] https://github.com/osmoscraft/osmosfeed

3. I digest the knowledge and connect them into notes with osmos::note.

All of them are done with plaintext (some sprinkle of markdown), remote hosted on GitHub, so they are easy to run NLP and ML against.

In the long term, I was hoping to create a "positive feedback loop". Use ML to extract patterns from my notes, make connections for me, and recommend interesting reading in the osmos::feed. On the other end, osmos::feed can use NLP to detect how each article in the feed might connect to ideas from osmos::note and make note-taking even easier.

The parent project (https://osmoscraft.org) is still in super early stage. Would love to let the community give it a spin while I keep iterating.

Thanks again for the ♥

GitHub

Youtube seems to be doing some A/B testing with the comment system which has made proxies like Invidious and yt-dlp/Newpipe unable to load comments. There is a patch for Invidious [1] which solves this problem but it is not in master yet. I tested it on my own instance and it does solve the problem.

[1] https://github.com/iv-org/invidious/pull/4576

BTW, I don't understand the workaround: https://github.com/iv-org/invidious/pull/4552/files

Which was taken from here: https://github.com/LuanRT/YouTube.js/pull/624

Could anybody explain it to me?

Use a Youtube proxy like Invidious [1], problem solved and you get to subscribe to channels without telling the Beast about your interests. Add Sponsorblock (which supports Invidious) to get rid of any in-stream advertising which remains and you'll be transported back to those hallowed times of yore when men were men, women were women and advertising was something you found in newspapers. Youtube will try to make this harder just like Xitter is trying to make it harder to use proxies like Nitter [2].

[1] https://github.com/iv-org/invidious

[2] https://github.com/zedeus/nitter

Going to drop this here for others who haven't heard of it https://invidious.io/

Now, how do we fix this? YouTube's ad model sucks. Their algorithm sucks. Their front page sucks. They've captured a bunch of creators though so often YouTube is the only place you can find someone.

I want those creators to benefit from me viewing their videos. I want the fact that I view a video and like it to help other people find that video in their recommendations. I want an algorithm that shows me things that are interesting and relevant not one that promotes the spammiest and most ad heavy videos that barely have anything to do with my watch history.

Having an alternative front end is nice but I don't want to rob YouTube of the money they spend on hosting the videos.

So, how do we do this?

Peer to peer fails when there is little interest in something or when most people leech and it sucks for archiving old content.

Hosting it all in one place is super expensive and hard for a small group to manage without turning into YouTube.

Maybe we could find a way for the creators to host their own content and get paid when people view it while being part of a large federated network for easy discoverability?

Please list any projects you know of, I'm sure there are a lot of people here who would be willing to contribute or donate.

Sometimes a half-solution will do, like Invidious or Piped.

[0] https://invidious.io/

[1] https://github.com/TeamPiped/Piped