support for deletions

[khimaros]

is this already planned?

reference: https://github.com/nostr-protocol/nips/blob/master/09.md

[dpc]

It's implemented already under the hood, just lacking a button in the UI.

rostra/crates/rostra-core/src/event.rs

Line 105 in 10620fe

/// Bit `0` - delete previous event - All well-behaved nodes

[dpc]

Once something gets published on the Internet it is (at least potentially) there forever. Even in a centralized system like Twitter, people just screenshot stuff to post in case it gets deleted.

There's just no way around it, especially in a self-generated-identity system like Rostra/Nostr. Everything needs to be signed, so it's relatively trivial to store copies proving their authenticity. Deleting can only ever be a UX convenience to allow people edit a post, etc.

[dpc]

In Rostra followers store copies (currently complete, but in the future possibly with a decaying probabilistic reclamation of old content) for availability. Nodes are not expected to stay online all the time, so they sync between each other.

[khimaros]

i understand true deletion is impossible, especially with clients that disobey spec, but defaults still matter and they define the experience of most users. i think the UX convenience in this case is still very important for a social network.

[dpc]

I do not understand what your point is. Do you mean that you'd like the deletion / editing to be exposed in the UI? Yeah, I might get to it eventually.

At the moment I don't see a lot of activity or users on Rostra, so I'm distracted working on some other ideas. Generally I do not have high hopes or ambitions of Rostra gaining a lot of popularity no matter what it does under the hood, as simply people are already happy using Twitter/BlueSky/Mastodon and p2p is technically cool, but mass user couldn't care less. :D

AFAIK the existing state is perfectly usable as a minimal proof that under the hood things are working, and if I see interested users/developers I might invest more time improving things.

[trosel]

AFAIK the existing state is perfectly usable as a minimal proof that under the hood things are working, and if I see interested users/developers I might invest more time improving things.

Your users are essentially developers and users of other decentralized protocols. So having a clear POV on topics like this will be important. Because your users will have already been familiar with how other solutions handle them.

Regarding the competition though, how much of the value proposition of Rostra is really just "it's a better nostr relay" rather than "it's a better decentralized protocol"?

No relays. Actually p2p.

This should give users more control over their own data, not less.

Scales. (I hope, LOL)
The lower layers of the implementation are more advanced tech-wise, for a good reason. Can still expose simple "sign me a json" API-s too.
Abandons JS, embraces Rust and Unix tech. Optimizing for JS-integration is non-goal, especially if it was to sacrifice being P2P.
Optimizes for performance and resource usage.

There are multiple Nostr relay implementations. These points could all just be critiquing a specific relay.

Events are chained together in a DAG.

This is cool, but is this why deletes are extra problematic? I recall manyverse was focusing on adding deletes to SSB https://www.manyver.se/blog/2022-10-permanence-and-safety

[dpc]

Deletion is now exposed in the UI.