Hello world
#introductions
monstr is cool.
There was a time when I was probably seeing most of my follow’s content.
But I’ve been out of the loop for a while and paid relays have z̶a̶p̶p̶e̶d̶ taken off.
So I’m probably missing everything now.
I can’t autocorrect.
https://nostr.watch/ has statistics now, and data, historical and present, that comes from the server instead of from your own browser.
I fixed gross perf issues in my relay fast last nights just in time for this.
There wasn’t a single “nostr” in the reply thread!
And trust users more and give normie users more credit.
People can back up data to any and all relays that let them.
They can run their own relays if they need to.
Nostr is amazing.
We need clients and user experiences that expose users to relays so that users can gravity to and “select” relays of quality.
We need this kind of market economy with relays or nostr doesn’t work long term.
We need to stop hiding relays from users.
It is working: https://nostr-relay.untethr.me/.well-known/nostr.json - I think snort is not validating them correctly.
You’re not giving normies enough credit
We had AOL - how did we ever get normies off of that?
Damus recommends relay sets, will soon effectively default/auto-pilot relays.
Where this devolves to is most of user base will have no concept of relays, “I’ve got my client and there’s some backend something out there cool I’m on nostr f the system” - but which devolves into every other standard social platform out there.
Yes we have “portable” identities but I’m not sure portable keys model saves you here … for all intents and purposes.
The nostr vision as I initially read it was relay-centric. This seems to go the opposite way entirely.
Maybe I’m overstating all this.
But we’re in prime early phase here and this seems the time to really embrace the core nostr model not hide it all.
Interesting yes query params are more flexible but the urls less likely to compare “equal” since you can reorder params - that might matter to some places in spec or clients don’t know
To users they would behave identical to regular relays so most use cases that apply to regular relays. Probably other use cases like ephemeral private-ish channel-like experiences. There are pay-for-your-own relay services out there, those are probably doing something similar or at least sharing machine resources across relays.
How many active connections? What’s your cpu load / how many cores? I serve about 200 concurrent and I’m about 20-30% of one core/vcpu. Majority of load is reads; sqlite shouldn’t balk at concurrency—using separate db just means you’re shifting some compute over to your database cores.
Probably good practice to change your mind even when the facts don’t change.