What should we do about DMs on nostr?
Login to reply
Replies (16)
Ser, what is the difference between the first and last options? 😂
1) just remove DMs from clients
4) keep them there, but still broken
What does @White Noise do? Love their product and philosophy!
Marmot
Well, get it on. iOS support would be killer too.
Marmot is White noise right?
Is it crazy that I want my open decentralized protocol to focus on what the client is for?
Twitter clone: Just microblogging
Youtube clone: video hosting
Vine/insta clone: short video/photo hosting
Substack clone: long form articles
Signal clone: direct and group messaging
These can all be their own apps with their own focus. Not every client needs to be a swiss army knife.
We are not sure lol and so far it looks like no
Both insta and Twitter and pretty sure all socials have some kind of dm functionality, it is useful, if it works, without being full blown chat imo
I don't know enough about Marmot or NIP-17, but whatever the "good enough" option is seems to make the most sense to me. Thats subjective, but I don't need Nostr for ultra secure messaging. There are other things for that.
I think Nostr changed the game in that respect. The reason they have DMs is because there's no reliable way to tie a username to a direct message outside of their client. With Nostr, you just message the same NPUB.
#Vector
Marmot is a protocol 
same as nip 17. Personally a fan of marmot.
GitHub
GitHub - marmot-protocol/marmot: The Marmot Protocol is a messaging protocol that specifies how to do efficient end-to-end encrypted group messaging using Nostr's decentralized identity & relay network combined with the MLS Protocol.
The Marmot Protocol is a messaging protocol that specifies how to do efficient end-to-end encrypted group messaging using Nostr's decentralized...
Marmot is tied to a device, makes sense for some. Clients
What's wrong with NIP-17? In my experience it works well. Use it daily across multiple clients. Nospeak is my go-to lately. Marmot, on the other hand is just too early. For all the hype, even the reference implementation (White Noise) is still horribly unreliable at the most basic functions & feature incomplete. At Marmot's current pace, I think it will be at least another year before they catch up to the reliability & features of NIP-17 clients.
Wait, what?? Marmot has some type of device ID built into the protocol? Thats not good.