Many text notes are just too long. If we have long-form notes, can’t we introduce a char limit for text notes? Is there a NIP for that?
Why are (Apple) podcast recommendations so bad?
(How can I switch off that iPhone always puts the cursor to beginning of word, not in the middle where I touched?)
Same experience, but switching from Android to iPhone. Actually makes me hopeful that I’m just stuck in some vendor specific lock-in ways that can be overcome.
That sounds quite concerning to me. What if you do not like that forever thing? What if it turns out bad? I thought, the goal is good life.
In my favorite German news paper (F.A.Z., subject to change) the extreme right gets a lot more attention than I would expect.
Because the follower graph is used to identify spammers. If non spammers follow spammers, this gets harder to do, hence the penalty.
I think as Bitcoin is seen as a counterweight to more (re-)active poltical systems, especially their moneys, there is a decent chance it will lead to better wealth distribution. Should ideally be one aspect of decentralization.
There should be a conference aiming for serious exchange between Bitcoin and central banking system representatives.
Ist das nicht eher positiv? Man kann für Heroin immer zu TikTok, Twitter, Facebook, etc., und ein Nostr mit besserem Inhalt, dafür kleiner, wäre nicht verkehrt. Es müssen ja auch keine Aktionäre glücklich gemacht werden.
True, but the idea was to warn those that follow the hacked account, before you unfollow them. So they have time to react and unfollow the spammer.
Especially in the beginning, to keep momentum, I do also not want people to be too careful with who they are following.
Contact list (kind 3) events that contain at least 5 other pubkeys and are not older than 90 days were selected, colored with Louvain community detection and then embedded in 4 dimensions via Force2Vec. One nice result is, that both Louvain and Force2Vec lead to similar clusters. My hopes are, that closeness will lead to good follower suggestions and I will continue in that direction. (So far there has been mostly methodology, not much interpretation.)
Started experimenting with visualizing the Nostr contact list graph. This is a first rendering of almost 60 thousand nodes (users) with 14 million edges. Feels promising if not useful (yet). 
Do you mean in response to your notes?
Will probably try myself with https://github.com/HipGraph/Force2Vec later.
ci tool minute, except you already preselected the downloading file for opening.
To make sure that other users kind0 event are handled correctly is quite hard, since you cannot even be sure, they sent them to your relay, even if they sent other events to it. But some form of ongoing analysis might be possible. E.g. monitor some users via your relay and some other relays (wellorder or so) and compare published events.
I would probably write a test program and run it against the server. E.g.:
- Open Socket 1
- Send filter for pubkey on s1
- Open Socket 2
- Send kind0 event on s2
- Verify that it arrived on s1.
(Extend with other test cases, e.g. open another socket to ensure, event is sent with "historical" events before EOS.)
In this case, kind0s are not different from other relays though.
Do you want to get all kind0s from beginning of time, or only for certain pubkeys (your followings?)? Your use case is not fully clear to me yet.