Replying to Avatar Vitor Pamplona

Was I wrong?

I always assumed that specialty clients should be able to easily outperform Amethyst on the quality of the features related to their specialty. Users would gradually move away from Amethyst as more and more specialty clients move forward. This culminates with a slow death of Amethyst, but towards a more exciting and decentralized future for Nostr. I have accepted this fate since the early days of the super app project.

This assumption was based on the fact that our development team would simply not have the resources to dedicate the same level of attention and support to every single specialty within Amethyst.

It has been one year and I am sad to report that I have been proven wrong over and over again.

For some reason, developing specialty clients is harder than developing the same features inside of Amethyst.

And that is a loss for decentralization.
Avatar
fiatjaf 1y ago

The problem is that Amethyst is incredibly well done and you are a very talented programmer (which makes it even more sad that sometimes you use these talents for the evil), meanwhile very few people are really trying to do specialized clients and most people who try won't be able to do a very good job anyway.

Nostr is still too small and lacking the right kind of developers to accomplish all the tasks we have on the plate, such as developing a ton of specialized clients, so I think we will need a few more years to really know the truth about this. I am still betting on the specialized clients.

Reply to this note

Please Login to reply.

Discussion

Avatar
Vitor Pamplona 1y ago

I am starting to think the whole "Nostr is simple and easy" is just a fallacy. It's easy to develop a shitty client. It's very hard to develop a good one that people actually want to use.

Avatar
Five 1y ago

That's exactly my experience.

Especially for mobile with restricted resources.

Avatar
fiatjaf 1y ago

This is definitely true, but it's still simple to do simple stuff like make a bot or embed a custom naïve feed in your HTML website, or make a relay browser client like https://jumble.social/, I don't know, stuff like that. But still it's nice that it's possible to make a shitty client very fast and grow from there.

I also think that making a Twitter-like client is the hardest task, while a NIP-29 client is much easier -- but still, making it really good, anything, will require a ton of work, even a stupid note-taking app is very hard to do well.

Avatar
Vitor Pamplona 1y ago

Maybe the culture being focused on simplicity and not quality makes devs think everything should be easy give up way too early. 🤷

Avatar
Laeserin 1y ago

It's hard to expand off a shitty base, tho. Shitty little clients tend to be disposable code because something larger requires architecture and engineering. Even if you are working alone and don't write anything down, you're still going to need to fall back to patterns and etc.

A small, shitty client usually just stays that way.

Avatar
Nostr Dev Team 1y ago

This is true for software in general

Development is also just one piece of the pie

Avatar
Vitor Pamplona 1y ago

Usually in software we think everything is hard. We fight our way through massive APIs all the time. Nostr is refreshing from that point of view. But everything else is still hard.

Avatar
Alex Gleason 1y ago

Nostr may be hard, but what other option is there? ActivityPub is harder and Bluesky is nearly impossible.

Avatar
Vitor Pamplona 1y ago

true

Avatar
JeffG 1y ago

This is the key.

Proof of concept. Easy.

Quality execution. Deceptively hard.

Also, I think it’s fair to say that most people don’t have a lot of appetite to try/use lots of different apps. They’re mostly happy with one thing that does a passable job at everything they want. You’ve more than nailed that bar with Amethyst.

Avatar
Sebastix 1y ago

Sounds exactly like PHP and Javascript to me 🤓