Thread

I like edits or at least the theory behind them. I make typos. I want to fix the typos. However, @fiatjaf has convinced me that they're bad. His editing and the gamification of his edits are fun, but if hundreds of people in my feeds are doing this, then it would immediately degrade my experience. I would hate it. It would be annoying. I now see edits as an attack vector and a performance degrader. Maybe we need a maximum number of edits? 3? I don't know. I'm just thinking out loud. Maybe we just do away with them altogether? πŸ‘€ Thoughts?

Replies (42)

He convinced me without his antics the first time the debate was brought up, but I absolutely loved the way he showcases it πŸ˜‚ I like edits too, I am also fine with annotations, I also like min number of edits, but I am convinced that we should not scream this feature into existence. It needs to be very well thought out. Maybe a client will come up which will show us the right way to do this.
NO EDITS, NO DELETES
Derek Ross's avatar Derek Ross
I like edits or at least the theory behind them. I make typos. I want to fix the typos. However, @fiatjaf has convinced me that they're bad. His editing and the gamification of his edits are fun, but if hundreds of people in my feeds are doing this, then it would immediately degrade my experience. I would hate it. It would be annoying. I now see edits as an attack vector and a performance degrader. Maybe we need a maximum number of edits? 3? I don't know. I'm just thinking out loud. Maybe we just do away with them altogether? πŸ‘€ Thoughts?
View quoted note →
there is nothing to think about - we need edit history like wikipedia - you can edit as many times as you want - but every version is saved permanently. i don't care what @fiatjaf thinks about this. he disqualified himself from the position of being the supreme leader of NOSTR when he called for censorship. there was ever only one person who understood what NOSTR is about and that is me.
I don't think they are more of an attack vector than regular notes, just unfollow npubs abusing them. One could also publish annoying gibberish as regular note, where is the difference? Edits are often useful, you just really can rely on everyone seeing your root post also seeing the edits which makes them a bit confusing for users new to the Nostr concept.
I think that is the edit answer. Allows the edit with chain of changes, or delete replace, depending on how the relay wants to handle it. Backward-compatible, accommodate simple fixes, or in a system that wants to be a mini GitHub for every post they can, use nip 37 to do that on their client relays. This is how I understand it as put forward. Just a matter of getting the clients to onboard it. See my other mock ups for Primal to adopt.
I'd ABSOLUTELY say no edits. I think it's part of the ethos of Nostr--and a reminder that Nostr is different--it's about privacy, and freedom of speech. No edits are a part of that-- Just like in real life--once you say something, you don't get to pull it back. Yes, you can apologize but your words are forever. So it is with Nostr. Think before you speak--both IRL and on Nostr.
I say, Yes to editing on the client layer, because typo. But, I say NO to edits on the protocol layer. There is a greater value to the protocol for the permanance of the notes than to a flexible note for the one user. Here's my thought process and where I arrived: I want to know that the post I'm looking at is the post that was placed, and the comments are about this version of it, and the repost are about this version of it. That's important to me as a community observer. As a contributor, I want to fix my typos. So, on those times that I look back and see a typo, it's usually right after I post it and before anyone has engaged with it. I just pop over to Nos.social, open the post. Copy the content. Delete the post. Go back to Primal, make sure it deleted, make the post anew with the typo fixed, and resubmit it for all to see without the typo. That workflow could be provided by Primal. Primal could let me "edit" whereby my original post would be deleted and the new version would be submitted as a new post. After my edit is ready, Primal tells me it's about to request relay deletion of the original post and submit this new version of my post, to which I confirm. That ability could expire after 5 minutes or some short term useful time limit, at the client level. But, I as the user am not inclined to do that anyway after someone has engaged with it. I'm more likely to just post a comment with the whole "I mean *typo." if it's even worth it. This is the case because the note's consistency after engagement is more valuable to the community than my ability to fix the typo is to me. This suggested functionality on Primal or any client is my ideal workflow on the protocol, while preserving the important features yet accommodating the necessary one for user experience.
We had this debate years ago on the Steemit blockchain... There's what I believe to be a very simple solution. Initially, Steemit allowed editing for typos and second thoughts for a brief span (~10 minutes?) after posting. As an author, I wanted complete control in perpetuity over my work, and fought for that outcome tirelessly. The final solution?πŸ€” Versioning. Each new edit is saved as a serialized version of the original post. Anyone seeking the "truth" about the ugly, typo-ridden original is welcome to dig back to ground zero, as nothing is deleted, while serious authors can fix typos they may have missed years later.πŸ˜ƒ
Great insight. I like both, but for different purposes. At a protocol level it's oft best to accommodate client desires openly and flexibly and simply. Do you see a fiatjaf describes Nip37 as offering the best of both worlds to the clients/relays on how they choose to handle it, with the end goal keeping a simple flexible protocol? Clients need flex to accommodate infinite use cases. That's the tough job of a protocol.
I hate reading a note, to which I wrote a word wrong or for some reason, I β€œate” a letter. So, editing would be valuable - and I wish it was possible to see this tool acting on @primal; my favorite client. I don’t know the size of the impact that this would cause, but for sure, a test in practice could be done; and if it is found that it is bad, excluded later.