For the next #zapathon I would like to improve the deliverability of zap events.
Our zapper is rate limited and blocked by some relays and we still often see timeout errors.
The zap event is the main feedback for the user so this is a problem.
What are your thoughts on how we deal with this in Nostr to scale?
alby zap only relay?
I think the main problem is that many zap events are delivered without reusage of the WS connection. One of the best ways that you could potentially go, it to maintain persistent web socket connection, and batch events using threshold (5 seconds or more) and number (under 10-50) per batch per relays. Hope it makes sense. 🐶🐾🫡
This is problematic.
