2w ago

Tuvix - Self-Hosted RSS Aggregator

Take back control of your internet with Tuvix, the open-source RSS aggregator. No algorithms, no tracking—just your content, your way.

55 comments

I have a feeling this will be unceremoniously killed and forked into two different projects.
- Huh, the connection was actually intentional, albeit definitely not for that reason
  Why "Tuvix"?
  Tuvix is named after a character from Star Trek: Voyager who was created by merging two individuals into one. The name came to mind when thinking about one of Tuvix's core features: merging multiple RSS sources into a new public feed. And who doesn't love a good Star Trek reference?
  We believe the best reading experience doesn't require sophisticated algorithms or endless personalization. It just requires giving you the tools to find and organize content you care about.
  https://tuvix.app/about/
  
  Makes sense, but it still makes me want to grab a coffee and order its murder.
  
  Of course
- Janeway did nothing wrong
- I'm sad that it took me a second to get it.
- *three projects
  Don't forget the plant.
- And rightly so!

- MURDERER!

The devs aren't worried about Janeway?

I'd still like some algorithm in my RSS aggregator. One, that detects articles talking about the same thing and groups them.
- Yup and if you gave it a basic text to speech and an audio player then you got custom radio with a news break
- Yeah and maybe AI summarization and digest feature

What’s the difference between this and say FreshRSS?
- Same concept, different implementation. FreshRSS is a PHP app, in my opinion.. a little ugly, still super functional of course. I wanted to try to create something with a more modern UX, and try to appeal to not just the tech folks. FreshRSS still supports things I don't yet, like WebSub, but give me some time to catch up. I have the massive benefit of just starting much later when many awesome libraries and AI exist.
  I actually started this API in Go, and it was nearly complete before I started over entirely in Node. And I did that so that it could run in serverless environments. You can of course still run this in Docker Compose, but it's actually focused on Cloudflare deployments, where you can run this entirely for free.
  
  For free you say?
  
  When you say AI, do you mean you’re vibe coding this, or like you’re using AI as a learning tool?
- FreshRSS doens't give the impression that it's fishing for VC money for one
  
  I'm not interested in money and you can quote me on that. I just care what my shit looks like

Is the selfhosted version able to also take email newsletters? I hate them and I'm using https://kill-the-newsletter.com/ to turn them into RSS, but I wish I had an all in one solution.
Also, is it fully FOSS or is it open core?
- AGPL v3. Take that as you may. I'm *open *to changing it. I just wanted to give the project a chance to thrive without someone just hosting it as a paid app and calling it a day. That makes it so that if anyone does want to create a commercial product from this, they would need to offer up the source. My monetization plans only extend to possibly offering an additional plan on the hosted version to help cover hosting costs (which right now are $10 a month and I am covering that, it may become $30 a month if this gets popular).
  
  AGPL works for me. Good to know.
  I just avoid using "source available" and software that has artificially paywalled features, the most common paywalled feature is OIDC because most devs seem to think that it's a business only feature.
  I pay for Home Assist Cloud, because I want to support them, every feature is available if I wanted to self host it. I freaking love them.
  The only exception being Bitwarden, although they have paywalled features in their selfhosted builds I don't know of a better-for-me alternative. I could self-host Vaultwarden, but I pay for their subscription just because I want to support them.
  My point is, if it's justified, I'll pay. Otherwise, I'll keep using standalone RSS apps on my devices and just backup my OPML every once in a while.

My problem is finding the feeds lol
- Ya. I'm working on that too. And trying to keep in the spirit of not being biased or heavy on algorithms.
  My first step - A chrome/firefox extension. This is currently in review on the web store. This exposes RSS feeds on sites you visit to make it easier to subscribe to the places you already visit. This is especially great when you find a great blog on Reddit or Hacker News. https://github.com/TechSquidTV/Tuvix-Tricorder-Extension
- Just add sites (by top level domain) that you use to read. You would be surprised how many provide feeds and even more surprised how fast your feed reader gets overwhelmingly filled 😅
  
  But I haven’t found a good site. Most are just your average capitalistic site with flashy keywords and filled with adds and not really talking about anything.
- RSSHub and RSS-Bridge can handle that. My issue is adding too many websites to my reader.

The docs look so good. So does the app.
- Wait until you try the Win95 theme
  
  Is there a problem with the verification email sending on the public instance at the moment?

Nice - going to try this :)

While I support the idea of using RSS readers to break free from algorithmic and/or AI curated feeds, I've mostly stopped bothering, since all the content that gets into the feeds has become algorithmic, AI slop.
There's just no escaping it these days.
- While super fair and accurate, I take this as an opportunity to follow much smaller blogs. When I find a good post on Hacker News or stumble upon someone through my research, I now actively make a point to subscribe to their RSS.
  TBH my original motive was to find good sources of content to submit to Hacker News... but all the same.
  
  lol I started to reply, suggesting a recommendation feature to help find non-algorithmic tech feeds but then realized that's exactly how all this started.
- There are plenty of outlets that specifically do not use AI in their writing. Those are what I fill my feed reader with.
- I guess getting a flip phone and stick to an eBook reader would be a good solution

The project looks nice and RSS Aggregators are the way to go.
I switched to a file based (cloud storage) syncing app like News Explorer a while ago. Sometimes less infrastructure involved is a blessing.

Can we get a few screenshots or a demo acc?
- Demo account is a good idea. ill work on that. Here's a couple screenshots

FYI, a few typos in "3. Organize with Categories" first paragraph in the getting started tutorial/blog post.
- Thank you!

We killed it. Good Job Janeway! I'm working on a fix. I thought my last deployment would have fixed it, I'm still looking into it.
Right now you probably notice no new feeds coming in.
tldr; too many RSS feeds, and tbh, it wasn't that many. I'm working on improving it to deal with the serverless limits on Cloudflare. These issues would not exist on Docker Compose.
edit:
fixed for now! We should be able to support at least 100 users now, and after my future changes it will be essentially unlimited but I will have to think about server costs.

Any plans for OIDC and read-only/non-root/no-cap container running?
- OIDC, maybe? I haven't thought about that, but super open to input. I haven't published the containers yet because I want to get to a more secure/optimized spot there first. I'm not actually sure if a read only container is possible? There's SQLite DB writes at a minimum, though that could be externalized. I wouldn't mind getting to that point.
  
  I am loving OIDC giving a single login for all the things I’ve got going, I see it as a near-essential for adding new services!
  Read-only is easy! You just need to confine where the writes happen. You use volumes for stuff you want to remember were written and tmpfs for stuff you don’t want to remember. Tmpfs for /tmp if needed, volume for the DB, good to go. It is super useful for security since only what is included in the container can be executed greatly reducing the attack area. No way to introduce a new excutable to the container! (you set noexec for tmpfs/volumes)
  I’ve seen difficult setups like a “work directory” where key files, executables, and temp files go. That structure can’t be secured, avoid that. Basically the temp files go in somewhere that’s not a big pile of a “work directory” - like /tmp - and then that structure once again works!
  Of course I wouldn’t say no to an LCARS theme either…

55 comments

Why "Tuvix"?