Oddbean

I agree with @fiatjaf about edits. There have been half a dozen edit proposals in the NIPs repo and I didn't think any of them was compelling enough to overcome the problems mentioned here. Annotations are really just replies with a tag requesting that it be treated specially by the client somehow (like rendering it within the original note). Clients can also just put all replies by the parent pubkey at the top of all replies so that people see the author's "corrections" first... and not even bother implementating the annotation spec (I plan to do that anyways). nostr:naddr1qqyxzepcx3jnxc3nqy88wumn8ghj7mn0wvhxcmmv9upsgqqqw4rsygpm7rrrljungc6q0tuh5hj7ue863q73qlheu4vywtzwhx42a7j9n5hfyhwm

I dunno exactly what fork is doing. If it just makes a new note as a copy of an existing note, with a link in the new note, that is fine. If it is using kind 1010 event versioning, that won't be recognized by other clients meaning it creates broken experiences.

There is a subtle difference : In #Amethyst , when I fork a note , it is presented to me as a copy , ready for the edit .. which is convenience ..but the crucial thing is there is a (system generated) link to the original note , which as a user I am abstained from edititing because the link shows up after I already posted .. I need to check if the link is visible in #gossip .. In a way , I am forced to link new note to to the original . I guess that is the subtle difference between copy/paste and fork !

Here is a forked note ... I dunno exactly what fork is doing. If it just makes a new note as a copy of an existing note, with a link in the new note, that is fine. If it is using kind 1010 event versioning, that won't be recognized by other clients meaning it creates broken experiences.