Instead of hardcoding starter accounts, we should work toward traversing lists/labels/topics based on WoT (maybe bootstrapped by a static list).
Hey Nostr, we are working on improving the onboarding experience within Primal, and could use your help. Please shill us the best accounts to follow in the following categories:
art
bitcoin
finance
fitness
food
fun & memes
gaming
human rights
music
news
nostr
photography
sports
tech
travel
Also shill us any categories we didnβt cover that already have good content.
Our criteria for suggested accounts to follow: the account needs to be actively posting high signal content on nostr. Our current working list is available below. Please tell us who we missed:
https://listr.lol/npub12vkcxr0luzwp8e673v29eqjhrr7p9vqq8asav85swaepclllj09sylpugg
Discussion
There is no WOT yet without starter auto-follows, or categorized suggested follows.
Primal is rightly incrementally improving by going the latter.
What Iβd like to see on nostr:npub18m76awca3y37hkvuneavuw6pjj4525fw90necxmadrvjg0sdy6qsngq955 is a powerful search on onboarding that asks βwhat are you must excited aboutβ. If the person types βknittingβ, the dozen knitting profiles are returned with option to follow.
cc nostr:npub149p5act9a5qm9p47elp8w8h3wpwn2d7s2xecw2ygnrxqp4wgsklq9g722q nostr:npub1uapy44zhu5f0markfftt7m2z3gr2zwssq6h3lw8qlce0d5pjvhrs3q9pmv
For sure, you always need to bootstrap. And Coracle currently uses static lists too, it's a fine first step. But it would be amazing to see these bootstrap lists turned into a "explore" tool which opens up the network based on dynamic trust networks.
This was my idea behind crowd funded classifications that are published publicly via labels.
The problem is the classification work youβre asking for is expensive (if you use a 3rd party) or time consuming (to build it yourself) and nobody is willing to pay for it.
CONTEXT is one of the essential ingredients for getting WoT to work.
Category-specific follows is one example of that.
this is something that invites would help resolve... they could see from their inviters' WoT immediately
THIS
Here is the user story. Shout out to nostr:npub1aeh2zw4elewy5682lxc6xnlqzjnxksq303gwu2npfaxd49vmde6qcq4nwx for inspiring discussion at nostrville on my favorite user story thus far
Coracle has something very similar to this (pubkeys, relays, groups, rather than topics).
https://coracle.us-southeast-1.linodeobjects.com/invites.mov
Love the personalization. Nostr needs more personalization
it's a client side thing though, just a matter of an invite link with onboarding flow and and immediate view of that feed while they spin up their identity
need an easy flow for nos2x type tool - it doesn't have a keygenerator in it last i recall using it... alby has a master key seed generation scheme but that's invite only, so, yeah
i think if you want to promote onboarding you need to fork nos2x and add a better visual and put a generator and backup word key thing
yes just should add... WoT was the killer feature that made it really hard for me to switch to nostrudel because i needed outbox model, and i miss the almost working NIP-42 as well
make outbox model for coracle and tighten up NIP-42 and i'll be back
You keep saying coracle doesn't do outbox, but... it does. What in particular is missing?
when i use it with my relay's chat bot, which often doesn't have even a profile published to the relay, i get all kinds of funny things happening with DMs, and the NIP-42 often takes a minute to actually go through
i wrote a chat flow that asks the user to reply with a challenge text and it lets me auth on chat in 10 seconds
i've written a tool that knows how to respond to a NIP-42 challenge auth envelope, and it does it instantly, it usually takes a minute or more, way way longer than i have the patience for to do testing
the lack of NIP-42 support is almost my biggest pet peeve of all existing nostr clients out there
it's SO easy to do it, but none work
Is that why my filter.nostr.wine relay keeps getting dropped on Coracle? I keep adding it, but it doesn't seem to work and eventually just disappears from the list.
ah yeah i get that one too... very often being dropped, or publishing disabled
it's because the NIP-42 support is not very solid yet
it's better! actually, come to think of it, it seems to work perfectly in nostrudel but nostrudel doesn't respond to my relay's auth challenges
i can only speculate that nostr.wine accepts posts signed by paid users by default no matter where they came from, and that they are not already existing in the database
nostr:npub1ye5ptcxfyyxl5vjvdjar2ua3f0hynkjzpx552mu5snj3qmx5pzjscpknpr nostr:npub18kzz4lkdtc5n729kvfunxuz287uvu9f64ywhjz43ra482t2y5sks0mx5sz question: nostrudel doesn't support my relay in development when it has auth-required in the limitation field of NIP-11 but it does work with nostr.land and nostr.wine - either of you two have any tips on how i get my relay to cooperate properly there?
Not me π€·ββοΈ
what relay are you running nostr.wine on? so i can try and figure out why it makes nostrudel do auth... i'm guessing maybe it's a NIP-11 thing, and i thought i'd implemented it but maybe i haven't i mean, if it's vanilla i suppose the relay info page will show
it only gives info about it being 0.3.1 what? no software field filled in in the NIP-11
doesn't say it's strfry, is it the rust one?
The backend is strfry but we run our own custom proxy socket in front for AUTH and account management (not open source).
ah, and here we see more reasons why this hasn't become widely supported
thanks
i won't be renewing my nostr.wine subscription or recommending it anymore until i hear you open sourced this sprocket
i can't believe what i'm reading rn
I don't understand what I'm reading, TBH. What are they doing and why is it not ideal?
you know about the direct message metadata leak problem right?
we are so far away from having this solved while people are keeping necessary NIP-42 implementations under wraps
it's short term thinking about their own gain and zero thought about the bigger picture for the protocol and its future, completely unnecessary, there really isn't that many competitors out there and this kind of short cut to keeping them out is contrary to everything this stuff is supposed to be about
NIP-42 shouldn't only be a gate to paid access it should also gate access to DMs and application specific data
So, you're saying that the implementations are too narrow in scope?
i'm saying that it's an elephant in the room - free speech is fine but a reasonable layer of protection against access to privileged data - we here see a prisoner's dilemma that nobody wants to take the cost (a few days effort) to solve, which is abombinable
I think this won't change, until more people use SDKs. Everyone will just build the absolute minimum they need. The correct model is for them to only implement the minimum they need from the core library.
Yeah, Nostrudel seems to work.
I left the npub and everything off, this time. Seeing if that works. Just the relay address.
Ok, but none of that is the outbox model
doesn't change the fact i can't use it for testing
Yeah, but I can't fix outbox if I don't know what's wrong with it, you didn't answer my question
it isn't contacting my relay even though it's in the relay list - the only entry even
it looks like the relay list in coracle still works by the old relay lists and not NIP-65 in/outbox lists, to me, that's why i said it
it only worked well when i ran an instance and locked it to my relay
the NIP-42 worked better in the custom instance situation too, like sometimes as quick as 10 seconds or less to auth
Exactly!