Oddbean

So you are a creator excited about nostr. You start using app A and like it, and you want your fans to join. You recommend app A and they start using it. Then app A changes - bans, censors, prioritises features you don't like etc. You've spent your clout to promote an app that you no longer like. You can switch, but it's hard to persuade your followers. That may or may not be a big or common problem, we don't know yet. But outbox model doesn't solve it. "Apps are interoperable" lowers the bar for switching, but it remains to be seen for how much. Does this make no sense?

I know you’ve got something in the works so I’m interested to see more, but I would have said this is why you use open source as a selection criteria for the apps you use. If they go in a bad direction, *you* or *someone else* in the community of users that also likes the app will fork it if the developers go in a direction that is bad. And it’s why we should select apps/clients that do things in an interoperable way so the switching cost stays low. But I don’t think other recourse exists.

No, it doesn't make sense because the outbox model should solve bans, censorship and all that. And I assume no one ever switches apps. Why do you think the outbox model doesn't solve it? I can see two problems there, but they feel unrelated: one is that as a creator you can recommend the wrong app -- an app that is not a real Nostr client, but something with the power to censor people -- and later regret; the other is that for a private community we need something else other than just outbox model, but I think we can stick to the outbox model for public posts in this discussion?

Every app has power to censor people, no matter how 'real nostr' it is. Someone's there to blacklist a pubkey and push the 'commit' button and the app gets distributed to it's users one way or another and someone gets censored. The motives for blacklisting don't matter, but I've seen enough in recent years to know that's inevitable. And I'm not even mentioning the lighter things like changes in design or features. App starts prioritizing Repost instead of Zap button and my audience's behavior changes without my control. Outbox doesn't fix that either. I'm not saying outbox is failing/wrong, it does the job at the relay level and makes bans in 'real nostr apps' much less likely. But it doesn't solve everything.

Technically an app was not supposed to change under me after it was installed on my device and we live in a dystopic world if that happens, but sure, yes, I get your point. I suppose "an app per creator" would be just some creator adding some branding and distribution to an open-source codebase such that they have control over auto-updates for their community? Update lovers everywhere will complain that this creator will either have to comply with upstream updates that may damage the experience he initially wanted for the app or they will have to fork the base open-source codebase, which he will be unable to, or he will have to never update again, which is a sin. I am happy with any of the three options, though.

> I suppose "an app per creator" would be just some creator adding some branding and distribution to an open-source codebase such that they have control over auto-updates for their community? Basically yes, but also being able to selectively extend/change the app with plugins. And even if creator is forced to do something (update, switch plugin etc), if that is happening on their own domain which fans are used to using then it shouldn't be an issue. Bigger question then is - would "true fans" use best-art-by-megan.com or not. Need to figure that out.