Usability is much more important than decentralization this early on. If your protocol is super decentralized but noone wants to use it, then who cares. Switching to more decentralized ways of pulling notes is something can be easily changed later.
Discussion
👍👍👍👍👍👍👍👍❤️❤️
Right decision
Primal is a prime example of this. Absolutely killing it in UX while not focusing too much on decentralization early on. It shouldn’t be too hard for them to make things more decentralized over time.
If a company is ever faced with the choice: stay centralized and make more money vs make things more decentralized but make less money, they’re probably going to choose the money
Yeah, or maybe just stay centralized and capture the network and sell it for a profit.
Yup, history is littered with examples of that, with ZERO examples of the reverse.
Its not too big of a deal there will be a wide range of clients to choose from with varying levels of centralization. Market will decide what the best approach is at any given time. Early on better ux will win, but they are not mutually exclusive and both styles will likely converge overtime on both sides.
Yes different topologies will always be possible but may become irrelevant if while nostr is still small and a big player takes the embrace, extend, extinguish approach.
How are they killing it on ui if the android app still requires your nsec
what is your solution?
Maybe Primal, when an user join it, ask just for an email and encrypt the nsec in their services, so if the user needs the nsec, they could request their nsec by email?
Like the services where when they join, the service send an email, the user click the unique link and the user is logged in, without need of passwords.
this wouldn’t change anything. The app still needs an nsec to create posts
Ok, I think when user join Primal, should have two options.
1- The user create an account with Primal and add email and password.
2- User create an account without email and password, just like today, with Nsec.
I see the biggest issue as a noob is that the distro don’t request both nsec and npub for start a login
Integrate amber?
nostr:note1v4vlp9m8h4203f0vws0crtp59g3g3e29ctdx5v4la0eqg7xst5ysheq60p
Isn't this the Bluesky ethos?
Yeah and they already failed at decentralization
Another important aspect is that the path to decentralization should be easy and obvious. Its not clear if bluesky has that
https://github.com/did-method-plc/did-method-plc
Cryptographic identity w/ key rotation.
The centralized PLC server can evolve to a decentralized relay-type solution.
I'm drooling.
nostr:npub1w4uswmv6lu9yel005l3qgheysmr7tk9uvwluddznju3nuxalevvs2d0jr5 can Amber evolve to something that integrates different cryptographic identities on different platforms.
Like "this key on nostr is the same person as this other key on BlueSky".
Also: https://github.com/simplex-chat/simplex-chat/issues/2859
Content ID format from IPFS btw. (CIDv1)
Totally agree, the end user will always give priority to what is easy for them to use. Value your experience above everything else.
By my definition Nostr is decentralized
It can be to different degrees
1. Connect to 1 relay (centralized)
2. Connect to 7 mega relays (a bit less centralized but arguably still centralized)
3. Outbox model, 1000s of community relays: decentralized
3. Outbox model, lots of people running their own relay: a cypherpunks-wet-dream decentralized
The flexibility is the most interesting thing to me
I can't think of a single example of something that started centralized and "decentralized eventually".
💯
This is exactly the problem with SecureScuttlebutt
Cool idea, can't get into it
I hit a similar wall with SSB, nothing seemed to work
Decentralized things grow slowly and organically. Slow adoption is not lack of adoption. 🐰🐢
The superier experience will bring the masses. The principles of the protocol will keep them around
zaps on Damus w/ WOS was good usability
indeed, will try to recreate that with something a bit less centralized
I am enjoying Damus! 🇨🇦😎❤️