Oddbean

Can’t help but wonder where development here would be if the community here viewed cousin protocols like Bluesky as inspiration, not opposition. Bluesky rolling out a great new feature should be motivation to find a way to nostrfy it, but it instead seems to be a guarantee that nostr will never have it. Feeds, gifs, self-moderation, and app passwords all come to mind.

@Dan yes but not in a way that’s easy to use and makes sense to the users. And we don’t have easy ways to discover them either. Our DVM feeds are slow and only work in a couple of apps. We don’t have good follower suggestions that go beyond showing the most popular people on Nostr. It’s not about making sure the feature exists, but making sure it’s usable.

One of the internet's greatest centralizing pressures has always been good UX. It's true they aren't mutually exclusive. But we shouldn't lose sight of the fact that being able to have both is A) very new B) still very hard

Of course. To say one isn’t to concede the other. I think the major trade offs mostly happen at the protocol layer, which much of those issues are being solved very well, it’s simply about extending them up to those who can use them properly. Again nsec bunker like tools are a great example. Literally all of the tools are there, it’s just never been made intuitive and the edges haven’t been cleaned up. And if that doesn’t happen then there’s no demand from clients to support it. Circular problem. I think it’s less of a problem of trade offs for a lot of these, and more a problem of people excited to build a tool that they can use, but not interested in building it for someone who doesn’t understand anything about it. Devs I think get lost in chasing the next awesome thing, rather than doing the slow, difficult process of making the last thing built really streamlined, intuitive, and reliable. Not saying I am not guilty of the same. It’s like building a house but never painting it or putting up trim. It’s just the annoying part of building because it’s slower and it feels like progress isn’t moving at all. But it makes all the difference in the world from the user’s perspective. I also don’t think we should depend on the devs to do this. I think we need “another layer” of developers, who work with this stuff and try to clean it up for the regular user. Part of the direction I’m trying to take our team is related to that. Hoping I can help, but I also know how much harder it is to get it done vs *wanting* to make it happen.

you build your own house, and it takes longer than buying a house, and you don't paint it for a good long while and maybe people don't understand the type of person that would live in such a house. but you're the user of the house. it serves your needs. you can see a lot of people from your windows, and that is fun and exciting. and then when you have time you can paint it and filigree the edges., or not. even better, the house that Nostr built has blueprints, that anyone can look at and copy and use, without restriction. i honestly don't give many fucks about usability beyond this sort of use, as a tool for people that need a Nostr. if you don't like the state of the tool, or you so very much want it to be better for your needs, great. build your house. this isn't particularly aimed at you, but i've recognized, or just recently discovered, a trend where some Nostr folks look to a desired future when we'll have more users, more content, more zaps, better users, better content, better UX, better steaks, more steaks, etc. today i realized exactly how much the nak cli does. it's not hard to use, it's just typing into a program i already have on my computer.

If you build a cool feature but ignore its usability, you might have well not wasted time on the feature. This is a very seriously under appreciated reality in both #Bitcoin and #Nostr generally. (Not saying there aren’t tons of great tools, but it does seem to be a broadly present issue)

Because there is no such thing as a good follower suggestion. Good follower for who? Based on what, my other follows? Base on ai reading their content? The whole premise is silly. Twitter/x recommends are absolute trash. There is no need for them on nostr.

As someone who has tried very hard on multiple occasions to have secondary/sub keys and using nsec bunker to let others post with limited permissions to my key… I can confidently say it’s all trash. And I don’t mean that like nobody worked hard to build something cool, because I know people did, but it’s practically all unusable.

I was just thinking to myself this morning about nsecs and what to do if yours gets compromised? Is there a way one could have multiples linked together as you, so that you could deactivate one if it got compromised (or lost or something) and still proceed with your same identity and all your previous notes and social graph? It's not like you can just change your password as with other apps. Maybe an nsec bunker could help with this? Maybe it already does? Honestly, I've found them too confusing to use myself, and it sounds like nostr:nprofile1qqstnem9g6aqv3tw6vqaneftcj06frns56lj9q470gdww228vysz8hqpz4mhxue69uhk2er9dchxummnw3ezumrpdejqzrthwden5te0dehhxtnvdakqz9rhwden5te0wfjkccte9ejxzmt4wvhxjmcjgxv3n hasn't fared much better: nostr:nevent1qqsxtw0z7rjce222sygzl9r9h244k26tusy9yyqjvz6npdajuqcgayqpzamhxue69uhhyetvv9ujuurjd9kkzmpwdejhgtczyzu7we2xhgry2mknq8v7227yn7jguu9xhu3g90n6rtnjj3mpyq3acqcyqqqqqqgld2k0g #asknostr #askdevs

I think this is a supremely underrated necessity for a really great and secure Nostr experience. Basically two main options to fix this, imo: 1. generate a new key for every single client you use, and then sign with a master key that “this is mine” and so every user sees them as all one account. 2. generate sub keys *from* a master key that you can give specific permissions to and make a new one for each client, and it simply is allowed to reach out to a main device that has your master, and ask it to sign. (Basically the nsec bunker design) Both open up a ton of possibilities. The first is easier, but requires more complication and data gathering from the client side it seems. The second has the problem of needing an always online device that manages the master key, or having some sort of “good for 100 posts” key or something somehow. Both are not easy, but both seem very possible. And I think this is far more important to have a system for than people recognize, imo.

As a comparison, ‘app passwords’ on Bluesky have been in the app longer than feeds, gifs, videos, and any moderation tools other than mute. They are easy to create, revokable, single use passwords for use on third party apps and tools. The interface is stupidly simple, and it dramatically increases security of your main password for those that choose to use it.