Took a couple hours to figure out how to get local uploads working on blossom, but damn if that first 200 doesn’t evoke pure joy. 🌸🎉 
Can’t help but wonder if their relay and cache are under a spam attack
No, sorry…. This one
I’m not convinced websockets are the only game in town. TCP/IP doesn’t care if you send smtp or http
Anyone else been thinking about notes being broadcast over a proper message broker like Rabbit?
I've been wondering what the outbox design pattern is. Appreciate you taking the time to describe, but I'm struggling to see how this improves things as you’re talking about a potential n^2 problem with clients.
He thinks nostr is Damus and Primal. Also, someone needs to tell the nostr CMO to start calling it something other than a protocol. People have no idea what that means.
Don’t be fooled by the digital monarchs trying to dazzle you with infinite options.
Convenience is a negotiation with autonomy.
Also if a piece of music is meaningful to you, it’s advantageous to own it vs renting it.
The subsonic API and nostr have a lot in common.
Post really strong opinions on the value of likes, then throw in a theory on how to make deletes work at scale (it doesn’t need to be viable, just say it)
I think the publisher subscriber relationship is due to invert at some point - you will whitelist trusted npubs and ignore everything else. People can scream filter bubble all they want, but it’s the only pragmatic approach to what is coming.
More and more it seems like there is a rush to recreate the thing we are trying to escape.
The idea of a global town square is a flawed premise. It is what’s gotten us into this mess in the first place.
Not being obsessed with the mass market is a form of freedom. That’s why I’m bullish on the philosophy behind Ditto. Find your people.
We already have music statuses https://github.com/nostr-protocol/nips/blob/master/38.md#live-statuses
I have looked at that before, my aim is more towards discovery, recommendations, personal playlist algorithms, something beyond ephemeral status. I’m starting by building something that will be useful for me, that way if I remain the only user, it will still be worth the effort.
Something else which would be awesome is relay sets for a column. Then a user could interact with different groups from the same interface. Damus, Ditto 1, Ditto 2, Local etc….
Yeah absolutely, this gives me incentive to get it pushed to a public repo. I’ll add my thoughts to the nostrability thread as well.
Been reading the docs and it looks like you specify them in the server config.
https://docs.soapbox.pub/ditto/install#add-relays
Kudos to Alex on the documentation as well. It's very clear and easy to follow.
Ditto appears to be following a pattern in which a user's relays are simply the relays of the Ditto communities they are in. This makes sense, it's just a change in mindset because currently there are 3-4 major relays which we all connect to in hopes that they have everything (every note, like, zap, profile, etc....)
This model embraces the fact that relay consensus at a network level is not only difficult, it’s not necessary.
Ah ok, so “membership” is handled by the background process on the local instance to only include approved npubs etc etc?
If it’s a moderated community it does seem counterintuitive to blast the note to every relay in my profile as there will be many out of context notes floating around. It’s an interesting problem.
Ditto looks amazing and the idea of smaller communities loosely coupled is much more appealing than a social monolith.
The data architecture is confusing though.
Notes created by an npub still propagate across all relays in a user’s profile and they are just filtered by the Ditto relay? nostr:npub1q3sle0kvfsehgsuexttt3ugjd8xdklxfwwkh559wxckmzddywnws6cd26p