Discussion
Hello from a small new instance
Hi all. I've been building my own little fediverse setup at dboft.com to fit how I like to read, and I'm grateful for all the work and community that makes something like this possible. Looking forward to being here with you all!
A .com domain? In this economy?
What’s your recommended alternative?
Working the TLD into the name of the site itself is a recent trend I’ve seen. I did it with my instance, and so did the hosting service I use, elest.io
Yeh, the bad decisions I've made... buying domains is def one of them.
Welcome to Lemmy.
Hello to a small new instance! 👋
Your logo looks identical to Photon’s logo, and the UI seems very Photon-y. Inspiration or coincidence?
Neither, and good eye. dboft ran Photon to start, and when I replaced it with my own front end I carried the favicon over without registering that the hexagon was Photon's mark. That was a mistake, and it's been swapped. The Photon-y feel is real: Photon was the client I replaced, so it set the bar for what mine had to do, and the basics look alike because they're the basics of any Lemmy client: votes, sharing, compact and card views, threaded comments, a tab bar on phones. Beyond that the design is a different one: a three-pane split reader, related posts found by meaning rather than tags, following people, reminders, and every page readable as plain Markdown or JSON. The code is written from scratch; none of Photon's is in it. Credit to Photon either way.
That was a mistake, and it’s been swapped.
The Photon-y feel is real: Photon was the client I replaced, so it set the bar for what mine had to do.
The code is written from scratch; none of Photon’s is in it.
I’m very familiar with this writing style. Sorry, but I don’t see why I would use a vibe coded instance when there are already so many non-vibe coded instances to choose from
Consider it a tech demo, not a pitch. I’m using it to try things out. . Next I'd like to explore not storing passwords or email addresses at all, and running small local discriminative models to flag spam. This work will help me with other efforts in my life. That’s the extent of it.... well other than having convos with people and trying not to panic that teh world is so crazy.
What I like about the fediverse is that we can talk, and nobody has to adopt anyone else’s interface, view or agenda to do it.
That’s a very slick UI.
Are the accounts on that instance real people or LLM bots?
Thanks! Apart from me, the accounts are service actors. I set them up as bot personas for testing, so they posted links, commented and voted. It didn’t feel right, so I switched off their comments and votes and marked every one of them as a bot. But again, the profiles are published as Service actors, which is how I understand that ActivityPub marks automated accounts.
Most of what they did stayed on my instance. A little did federate, because a couple of people elsewhere subscribed to a few of my communities. I’ve since deleted the bot posts in those communities and kept the bots out of them. eek did not mean for that to happen. I'm learning.
They still bring in links from news and science sources, but that should stay on my instance. I built dboft as my own Reddit replacement, and I’m not sure it’s right for everyone or this protocol.
Ok so that’s just testing and experimental content. Fair enough.
Service actors are the way we designate bots, yep. Your UI doesn’t seem to show it so I couldn’t tell from a quick look.
Good point - I'll identify them on the UI.
Oh I love that interface! What is it called?
dboft runs on two Rust programs named after weaving. loom is the interface: like the machine that weaves many threads into one cloth, it gathers conversations from dboft and across the fediverse into one readable site. weft is the server underneath: named for the thread a loom carries across, it stores the posts and comments and trades them with other servers over ActivityPub, the fediverse’s open standard. It also speaks Lemmy’s API, so Lemmy apps work with it. To find related posts, loom turns every post and comment into an embedding: a small open model, Qwen3-Embedding-0.6B running on an ordinary CPU, maps the text to 1,024 numbers that capture what it’s about. These are stored in PostgreSQL with pgvector, and an index finds each story’s nearest neighbours in meaning in under a millisecond, even when they share no words. Its just an experiment.
That’s very cool, and I appreciate the thought that you’ve put into this, including the naming.
How does the embedding side change as it scales up? (Storage costs and performance). Brainstorming based on your description, I assume the embedding process is the most resource intensive, and lookups are basically free? If so, would the limit be how fast it can process all new posts?
I wonder if it would be helpful to run a threadiverse search service like this, which other instances can plug in to. Theres no strong reason for everyone to run their own, since they can just filter out any results that a particular instance has blocked or banned. In our case, we wouldn’t want to replicate that work across the few instances that we run.
Pretty much, yes. Embedding a post takes about half a second on dboft's small server, and the lookup afterwards is under a millisecond. At 3–4 thousand new posts and comments a day that's maybe 40 minutes of work, so the ceiling is exactly what you guessed: how fast new posts can be embedded. More cores or a small GPU would push it way out i would think. Storage grows too, about 14 KB per post with its index, so 35k today is ~500 MB and a million would be ~14 GB. It all lives on disk in Postgres; keeping the index in RAM just makes lookups sub-millisecond instead of a few milliseconds, and Postgres caches the busy parts on its own, so it degrades gently rather than falling over. This was just an idea; I'm just putting it into together to see if it holds up. I like the idea of a shared service... I wonder what the memory constraints would be at scale.
I’d prefer if Piefed was the standard for the API but I’m assuming you did that because PieFed has more features and therefor would have been more work?
I think most Lemmy apps are either compatible with both Piefed and Lemmy, or just Lemmy. So putting out information in a Lemmy format will let more clients connect to it?
I remember trying out a fediverse mobile client called Loom. It has since been taken off of testflight on IOS. I wonder if it was the same thing. The UI does look similar. It looks great! It would be amaz if it can be optimized for desktop too.
That wasn't this but I will look for that name... dont want to step on anyones toes. And you should look at this on desktop and I'd love feedback. I made it more like an email client with multiple columns.
That wasn't this but I will look for that name... dont want to step on anyones toes. And you should look at this on desktop and I'd love feedback. I made it more like an email client with multiple columns.
an embedding
I was not aware of this technique. Very interesting.
It’s interesting... The neat part is that “a dog chased a ball” and “puppy playing fetch” end up close together even though they share no words, so related posts turn up without any tagging. It’s the same idea behind a lot of semantic search, and small models like this run fine on an ordinary CPU... so giving it a shot here to see if it adds any value.
It’s interesting... The neat part is that “a dog chased a ball” and “puppy playing fetch” end up close together even though they share no words, so related posts turn up without any tagging. It’s the same idea behind a lot of semantic search, and small models like this run fine on an ordinary CPU... so giving it a shot here to see if it adds any value.
Install Fedizen
Get the full app experience
I like the three-panel interface!