Nostr Web Client

Nice! I don't think it will be hard to implement if it passes, however:

> A kind 32123 event by the original pubkey overrides the labeling in the original event.

is a MASSIVE change to the way Nostr works (non-replaceable events don't get overridden). I hope this doesn't start a trend with overriding events or we will be back in the loosey-goosey web.

I think a Reaction-like idea where events can be reacted to is a more appropriate way of adding (and deleting) labels to an existing event.

Matt Lorentz 2y ago

I'm confused. Kind 32123 is definited as a parameterized replaceable event, isn't this how they are supposed to work?

"A parameterized replaceable event is defined as an event with a kind 30000 <= n < 40000. Upon a parameterized replaceable event with a newer timestamp than the currently known latest replaceable event with the same kind, author and first d tag value being received, the old event SHOULD be discarded, effectively replacing what gets returned when querying for author:kind:d-tag tuples."

Reply to this note

Please Login to reply.

Discussion

Vitor Pamplona 2y ago

Yep, the event can replace itself and that part is fine. The hard is that it "overrides" the labeling of the original event.

So, if this becomes a thing, one can design a new event kind that overrides Kind 1 texts, for instance.

s3x_jay 2y ago

Do you feel parameterized replaceable events are inappropriate in the context of labeling?

s3x_jay 2y ago

To be more clear… It only replaces the labeling of the person doing the labeling. It does not override the labeling put on it by other users or the original author.

s3x_jay 2y ago

Where I'm going with this is the idea of "trust" - where a user can say they trust certain people and only the moderation actions of those people (and the actions of the relays they pull from) can filter something out of their feed.

I explained it in a lot more detail here: https://s3x.social/nostr-content-moderation (though that hasn't been updated with what I did this weekend). And it's in line with a research paper that was pointed out to be by nostr:npub16zsllwrkrwt5emz2805vhjewj6nsjrw0ge0latyrn2jv5gxf5k0q5l92l7 which calls it "TrustNet".

Vitor Pamplona 2y ago

Still. I don't think it's a good idea to create NIPs with event kinds that replace information in other event kinds. Replacing yourself is fine, replacing the original tags is not (IMO). Even if they come from the same author.

On the other question, the use of a replaceable event destroys the history of changes in labels (relays delete past versions of replaceable events). If we want to use this for serious businesses, it is likely that we will need to keep some historical context for auditing procedures. In that sense, a non-replaceable event structure (like reactions) makes more sense to me. You can save a reaction, you can save the deletion of a reaction, etc. All dates are kept.