Avatar
Thorwegian 🇳🇴
c708943ea349519dcf56b2a5c138fd9ed064ad65ddecae6394eabd87a62f1770
1983 vintage. Lives in Oslo, Norway. Interested in tech, cooking, history, music, art and travel.

speaking of vending

if you like any of the music i have out on https://thorwegian.bandcamp.com you can zap this post with €1 worth of satoshis and tell me what song you want, and i'll send you the WAV file for it.

for paid content, say if i share one of my Bandcamp songs on here, a "zap to unlock" feature (a micropayment) would be more effective. maybe if i had a music bot that you could zap with a message to order a copy of a song...

first reason i can think of: it is exceedingly rare that i look at a post and think "this person should receive money for posting that!"

i have my doubts about how useful the zap button really is.

it looks way easier to make a barebones Nostr client from scratch with no library than to make a barebones Mastodon client from scratch with no library.

for being such a popular choice for email spam filtering, naive Bayesian classifiers are surprisingly rare outside of that domain.

if nobody made a "hacker's client" for Nostr yet, where debug/trace info is available with a click of a button and API calls can be manually issued if needed, maybe one should exist. doesn't need to be good for daily usage, but more for checking things if you're a bit tech savvy.

is @elon the real Elon Musk?

having paid a minuscule one-time admission fee for posting access to a relay isn't exactly the same thing as saying you're not a bot.

i found this thing, which should at least let me pipe the output of an individual relay to /dev/stdout to check what's on there...

https://github.com/blakejakopovic/nostcat

problem with using nostrgram.co: it automatically refuses to read from free relays since it assumes they're all full of spam. and one of the things i want to assess is whether this is always true, to which extent it is true, etc.

did anybody implement a Nostr relay that applies Bayesian anti-spam filtering to posts?

i wish clients would let you check which relays a post came from. i'm trying to assess relays here and it's difficult when you can't even see that information. there is also a NIP that recommends a relay along with a post and i'd like to see that information also. i need to look at some dev tools i guess. there must be a "debugger" type tool out there that lets you view all the data coming from the relays.

man, i'm hearing more noise from the upstairs neighbours than i used to when i first moved in here. they have a washing machine and right now some woman is coughing like she's got pneumonia and needs an antibiotic. since it's been so quiet up to this point, i assume the place must've been empty until now and that someone moved in there recently. the floors are much thinner than i thought they were. i can hear them so clearly.

don't know what's going on with the filter on the Universe timeline in Damus. each time i toggle a relay there, it reverts back and i have to do this 5 times over before it finally sticks. is something else constantly updating that list?

the way iris.to filters out bot spam by hiding users unless they're in your network and have 5 or more followers is effective, but also effective at hiding new users who aren't networked with anyone yet. it works as long as other clients don't do the same, but if every client did like iris.to, no new users would ever get any followers. so it's a "selfish" kind of filter that way.

valuing the contributions of developers more than the contributions of users is a mistake, i think. i ran a web community once where i made the grave mistake of assuming that i was the leader of the community and could change anything i wanted. as it turns out, no, you can't do that, because users can walk out on you, and without users, software is worthless.