Luxas on Nostr: Welp, I was successful in not only refactoring Nosflare into typscript, but also ...
Welp, I was successful in not only refactoring Nosflare into typscript, but also linking it through Durable Object to create websocket state cooridination (in order to broadcast new events to all clients in real-time), but this creates such a bottleneck and extra hop that the trade-off in speed is not worth it. Basically, with the current iteration of Nosflare, due to Workers being stateless, the websocket state doesn't entirely work so the Workers runtime eventually just force kills the hung worker after 2min. This is pretty much fine as it doesn't impact the actual relay functionality (and I'd be worried if it really took more than 2min to do anything anyway), but it just means the error rate in logs is a false-positive as it looks like 99% lmao
Published at
2025-06-07 04:05:39Event JSON
{
"id": "1620d5ba3ff8d5d1ea2685c4acb6bfe725e7e8174f40d7ed83e5f29836c34590",
"pubkey": "d49a9023a21dba1b3c8306ca369bf3243d8b44b8f0b6d1196607f7b0990fa8df",
"created_at": 1749269139,
"kind": 1,
"tags": [],
"content": "Welp, I was successful in not only refactoring Nosflare into typscript, but also linking it through Durable Object to create websocket state cooridination (in order to broadcast new events to all clients in real-time), but this creates such a bottleneck and extra hop that the trade-off in speed is not worth it. Basically, with the current iteration of Nosflare, due to Workers being stateless, the websocket state doesn't entirely work so the Workers runtime eventually just force kills the hung worker after 2min. This is pretty much fine as it doesn't impact the actual relay functionality (and I'd be worried if it really took more than 2min to do anything anyway), but it just means the error rate in logs is a false-positive as it looks like 99% lmao",
"sig": "ce6b869900737c6b07fd14ee2b3c8d06a0ab0d41a8cf446ab17f792b2a8a9b7e5fd476451c8aa0b6a49a183ab66269ce4506635ecc2a870ada959b62e4935248"
}
