FWIW, I think deletions on Nostr are a "meh" idea and edits are a bad idea. Consider the following: - It's an additional layer of complexity. Nostr should arguably be simpler and it's already more complex in some ways centralized platforms are not. - Different relays will disagree on whether to accept the edit or the delete event, so the original note may stay available. It may also become apparently unavailable and then available again in the future. Someone could send the old version to relays you don't use. - Unlike some specific features like articles, zaps and whatever one can think of, it only makes sense if consistently implemented, but that's something we can only assume to be true of NIP-01. - It makes retrieving a note less efficient. You now need to wait a response from all relays, just in case one of them has a deletion or a modification the first one to reply doesn't have. - There is value in having modifications, but there is also value in having platforms that don't allow modifications, IMO, where an ID truly identifies a specific piece of text and nothing else. Now that Twitter allows modifications, Nostr can use the fact that it doesn't to diversify itself so as to convince users to use both (otherwise the network effect plays against Nostr). On centralized platforms which allow modifications, users sometimes write a post with the plan of modifying it later. On Nostr that would be a bad pattern, however and I think the modification features can lead to bad usage of Nostr and, therefore, a bad user experience. Ultimately it will prevail if clients and relays implement it and fail otherwise, but I'm not in favor of it.
In this era of complete propaganda, lies, and scammers; engraining post history in stone is awesome. It becomes impossible to claim "I didn't say that", or scrub history, or scrub accounts. It's much harder to manipulate for nefarious reasons. More truthful.
To be fair, edits don't necessarily prevent that. You can have an edit history, like some social media platforms do.
Edits would kill nostr. The performance hit would eventually have hobby relays revert back to no edits. Might as well not have them. We already cannot guarantee that all the notes are read by all your followers. If you add edits to the mix, it becomes a shit show and a half.
Edits can be seen as just another note. So the performance argument does not apply anymore than saying a sudden increase im users would kill the network. Relays will do what thay can what they must, which includes everything from charging to curating. And users do what they must, from paying to running their own relay.
> Edits can be seen as just another note. So the performance argument does not apply anymore than saying a sudden increase im users would kill the network. This is not true either. The client needs to actually fetch the edit before displaying it. Currently to display a note you only need to fetch that note and you can cancel the request immediately after receiving the response from one relay. If you are waiting for edits, you need to wait until all relays have answered (within some timeout), to make sure that you are receiving the most recent edit. Maybe the performance cost of this isn't very significant at the relay level, but it may be more significant at the client level. > Relays will do what thay can what they must, which includes everything from charging to curating. And users do what they must, from paying to running their own relay. Personally I'm in favor of relays neither charging nor curating (beyond having some way to remove spam and illegal content) and users to use them interchangeably without thinking about it. For this to work, we'll need relays run by nonprofit organizations.
This convo highlights structural issues with Nostr. Paying subscriptions for relays will never work. Subscription for social monetization does not work en masse. Same for clients. Ad-revenue does. Stick to what works and flow the ad-revenue from clients to relays. Removing "spam and illegal content" defeats the entire point of Nostr. If gov bans free speech, and Nostr enforces it, then Nostr is rendered useless. Nostr (or any decentralized network) needs a way of obscuring and anonymizing relays. So they can host whatever they want without gov sniffing them out. Illegal or not. Filters can apply at the client level. But remember, any filter risks censorship. Even good ones.
> Paying subscriptions for relays will never work. What I think we need is nonprofit organizations that run relays as part of their mission. > Same for clients. Clients don't need revenue because (unlike relays) they don't have expenses. We know that the FLOSS community often builds software without economic incentives (sometimes there are sources of revenue in FLOSS, but not always). This won't be an issue. > Ad-revenue does. Stick to what works and flow the ad-revenue from clients to relays. Nah, fuck ads. Hackers and nonprofits go brr… > Removing "spam and illegal content" defeats the entire point of Nostr. It really doesn't. "Illegal" by itself is meaningless because there is more than one jurisdiction. I don't follow Canadian laws and, I presume, you don't follow Italian laws either. Nostr doesn't guarantee nothing get censored, but it does guarantee nothing *that at least one relay will host* will disappear. If your content is so harmful for the relay itself that no relay would host it, or those that do cease to exist as a result, your post will disappear. If you want to prevent this, you'll have to look to other systems, such as blockchain-based social media. Some exist, but Nostr isn't one of them. > Nostr (or any decentralized network) needs a way of obscuring and anonymizing relays. The way exists outside of the scope of the Nostr protocol and it's the darknet. You can very well make a hidden Nostr relay. The main issue I see with it is performance. The reason you might want to remove spam is simply that it's of no value to the user, while still costing you space. There surely isn't any requirement that Nostr relays must remove spam. In fact, most probably don't, currently. But if it becomes necessary I don't see that as an issue. > If gov bans free speech, and Nostr enforces it, then Nostr is rendered useless. My stance when I learned about Nostr was, and continues to be today, that Nostr is not a solution against government censorship but, rather, against corporate censorship. Right now the public is often constraining themselves to what Google, Meta and Twitter want to allow. This isn't good and Nostr can solve it. The solution against government censorship isn't technical, it's political. We need to defend fundamental human rights and liberal values. That said, there is never going to be any way to make sure that Nostr relays host all content. This was never the intention of Nostr. The workaround against filters at the relay level is using multiple relays and trusting none of them.
This is why we shouldn't have edits. nostr:nevent1qy88wumn8ghj7mn0wvhxcmmv9uq3camnwvaz7tmwdaehgu3dxyhxucn09eskuempde5jucm09uq3uamnwvaz7tmwdaehgu3dwp6kytnhv4kxcmmjv3jhytnwv46z7qgewaehxw309ahx7um5wghxymmwva3x7mn89e3k7mf0qyd8wumn8ghj7mn0wd68ytnrv4exxct5wfhhvcfwd4jj7qguwaehxw309ahx7um5wghxx6r9v4ek2un0vfhhgtn0wfnj7qfqwaehxw309ahx7um5wghx26tww4hxg7nhv9h856t89eehqctrv5hszxrhwden5te0dehhxarj9enx6apwwa5h5tnzd9az7qgcwaehxw309ahx7um5wghx6ctnwdkh27pwvdhk6tcpremhxue69uhkummnw3ezumtfv3jxc6twvuhx67tydeeju6ns9uqzqw09w2jktj6mysdvl5s68cq2s3enf524qrvlqglmj0pnjktjd2vrfj9kt8
edits are dumb. agreed
The beauty of Nostr is that people are free to use the protocol in anyway they see fit. If people want to edit stuff, they will edit stuff. I just provide the tools for that. It's the user's choice to edit or not.
the edit could exists as a special selfreply to the original post; if some clients implement it, it seems like an edit (with the visibility of the original), or if some clients don't implement it, it could seem as a normal reply, that corrects the original one
This. I pretty much mentioned the same thing a while back nostr:note19kyue5j08wr8vuu6vgf8yhpu0zkwwdth4kwdjtx7c86ulkay4fdswc6e2x - Show a "show edit history" button and that solves some of the issues. - Have clients ping a few relays from the user's relay list in the background to see if there's edits. That solves some issues. - Have edits as a specially reply (there's kind 1 post id, and there'd be its Kind1PostID:0, then :1 , then :2, then :3, etc, and clients present the highest number as the primary post. I'm not a programmer so this prob a dumb way to go about it), that solves some issues. - A client doesn't have this feature? It shows it as a normal reply. Also, big whoop. People go to clients that does have this feature. The market moves around.
This sounds a like a good compromise solution. Also, I kinda like community notes on Twitter, wish that could be implemented in a similar way.
最近 Amethyst 引入了一个给 Nostr 笔记再 "编辑" 的功能(就像 Twitter 那样的). 但是社区很多人认为为短文本消息(kind:1)的再编辑是徒增复杂性, 本身 "删除" 功能就是一种 "看起来有用" 的实现, 再编辑只是这种做法的延续. 目前 Nostr 允许内容互相替换的只有 kind:30023 (参数化可替换事件), 如果再引入编辑功能可能会继续影响事件的传递和检索效率(服务端一侧). nostr:nevent1qqsrnetj54jukkeyrt8ayx37qz5ywv6dz4gqm8cz87uncvu4jun2nqcpzamhxue69uhkummnw3ezuendwsh8w6t69e3xj7szyz8wg2gvnzeqk2vezw9nyt6xd63qcatq3atg4dfu7y2f90um96zfcqcyqqqqqqg4rasxv
Well said.