I don't know. I was trying out the inbox/outbox effect on Nostrudel, Primal, and Coracle and then it was like metadata froze and wouldn't let me save changes, and I noticed that npub IDs had wandered into the dataset. I didn't add them in, tho, at least not intentionally (copy-paste error, maybe?) and I can't figure out how to get rid of them. 🤷♀️
Discussion
Were you using metdata.nostr.com to update your relays, or other clients too?
Both. Made a big mess, testing stuff.
I'll recreate your relay list tomorrow and try and make the tool resilant to that particular flavour of brokenness
I replicated your profile but couldn't recreate your issue. I could still manage relays without a problem. There wasn't a raw hex pubkey in the profile, but from the screenshot, it looks like you had already deleted it before reporting a problem.
It's like this now.
Why don't you restore a backup?
Tried. It gets stuck on "Sending...".
This happened because it was trying to send the event to the old set of relays (the pubkey) plus some fallback relays that are no longer active. I have made a 2 changes to address this:
1) It backs up the new event first (with the relays you want to restore) and send updates these relays instead of the ones you wanted to replace
2) I have replaced the fallback relays with relay.nostr.band, nos.lol and relay.damus.io which I think are accepting events from any pubkeys. events also get sent to blaster.mutinywallet.com as they are profile related.
nostr:npub180cvv07tjdrrgpa0j7j7tmnyl2yr6yr7l8j4s3evf6u64th6gkwsyjh6w6 I'm not convinced the changes on the master branch are being deployed to metadata.nostr.com. In your capacity of CEO of nostr can you check please?
Of course not, it's a manual process.
Also I've forked your repo because you never gave me write access to yours, so there are a bunch of changes there now: https://github.com/fiatjaf/nostr-profile-manager
How can we merge both?
You had access but branch protection was on, forcing everything to go through a PR. I turned that off now. There seems to be a few linting issues and test failures on your master branch.
We should probably start from your tip, as that's whats in production.
I have proposed some changes to resync.
I merged it using ngit and deployed it. It worked amazingly.
Okay, sounds good. Tell me when I can test.