nostr:npub1l2vyh47mk2p0qlsku7hg0vn29faehy9hy34ygaclpn66ukqp3afqutajft I love what you're doing with relay sets. I think everyone should be able to run and manage a relay with just a few clicks. This will open the gates for š« nostring in the future. Happy to see that relays are on your radar! šÆš¤
making relays first-class citizens is absolutely on my radar because that will open up a bunch of new uses of nostr
we're using nostr as if it were a client-server architecture, but it's not. We're missing out on a bunch of innovation because we're not thinking in the right terms/allowing for relay balkanization
Yes, make relays first class on nostr. Creates a world computer.
We need smarter relays in the mix.
That is, it executes simple code on notes.
why?
It opens a whole realm of possibilities like an internet computer. Those smart relays could be paid depending on compute timeā¦so both win. Of course we got to be wary of bad actors.
Why a smart relay? Why are they smart?
š¤·āāļøI just came up with it on the go. I assumed it being able to execute simple code makes it better than ādumb relayā that only stores data. Another use case I imagine is- say someone wants their note to go viral. They can attach 5000 sats to it with a code that pays 100 sats to everyone who retweets it with a portion going to the relay that executes the code! This allows advertisers to pay users directly.
Dumb relays is what make nostr tick š š
What would making relays a first class citizen look like? Iām hugely in favor of a swarm of small, dumb relays. Really helps decentralise and add censorship resistance
Surfacing interactions with them to the user level and make them top of the knobs a user plays with.
But for that to make sense they need to differentiate. Chicken and egg problem.
But in this case, relays canāt do much to differentiate/Balkanize if clients donāt provide more levers to the user. 
Yeah Iāve been thinking about that too.
I started building a relay explorer so that I could inspect whatās in a relay (and eventually query for things like communities or relay lists or other items that help with discovering more Nostr folks)
It makes me think Iāll wanna add the ability to not just inspect a relay but discover relays.
Would that aid in the effort of balkanisation?
nostr:note1yaqlvj8zr4nwk0jf77qn55jpg3efdymcz87ryfk0lzvavhz4zyus4zaf5g
In general tools to understand relays are severely lacking, but I think thatās a byproduct from the fact that all relays are used indistinctively by clients, so all people that might query stuff about a relay solely use speed/uptime metrics, which I think is boring and barely useful.
A tool like the one you mention is downstream from clients changing their behavior.
But hey, you want to skate where the puck is going, not where it is.
The question every mildly curious person asks is āwhat relays should I use?ā
Is a good question (currently) without an answer. Itās highly subjective. And, again, speed/latency/uptime is maybe 2% of the answer.
Agreed, Iāve definitely had that same question.
I definitely wanna hack on this.
I know you can query relay metadata. And people can publish relay lists. Iām wondering if relays can currently state in their metadata some concept of ācommunitiesā that the relay is āintended forā
Something optional that just indicates what the relayās host wants to attract. They wouldnāt boot people based on this, theyād just be raising their hand so that people can find the relay based on interests/identity/locale/etc.
If that already exists then Iāll just build a relay discovery tool that aggregates this info in a way thatās searchable by Nostr users.
If not I may try to add a NIP for it / set a pattern and see if people are interested even before a NIP.
What do you think?
nostr:npub1l2vyh47mk2p0qlsku7hg0vn29faehy9hy34ygaclpn66ukqp3afqutajft do you have any additional thoughts? Gonna work on this in some spare time but would love feedback
š this idea
why doesn't the code execute in client side?
šÆ
I don't run a relay but I know it is the way.
1st class relays would give relays a chance to differentiate themselves, and open up better ways for relay operators to keep the lights on through value-add.
Would the concept of client side relays make sense for some use cases?
I'm mostly thinking P2P DM's, but maybe there'd be others (but keeping in mind that most clients won't want the bandwidth costs of running a 100% public relay).
I guess nostr:npub1xtscya34g58tk0z605fvr788k263gsu6cy9x0mhnm87echrgufzsevkk5s's new LMDB-backed database for events would be a nice building block for this...
To clarify, I mean that each Amethyst or Damus installation would set up a local personal relay for the user.
I think it would be a great idea, the only thing is you'd still need a server somewhere if you wanted anyone else to be able to connect because of being behind a network firewall. There is server software that exists that could make this doable, but you're still stuck with *someone running an extra server to negotiate the p2p connection for you.
Without this, only you could connect so it would only be useful as a backup server for your own notes.
