Thread

I'm keeping a close eye on Trusted Assertions. I love that we have all these WoT services and metrics popping up, but I am concerned that it could lead to centralization. We should really have this as a core part of the protocol.
david's avatar david
Separation of Trust and Client
Web of Trust systems in nostr need to generate trust metrics that are contextual, personalized to the end user, and portable throughout the ecosystem. These requirements, in particular the portability requirement, can only be realized when trust metrics are calculated independently of the nostr clients that consume them. This article explores how innovative solutions like Trusted Assertions can make this a reality.
Read article →

Replies (5)

Wot ( web of trust ) is like giving control to whoever who created it , for the aimed to eliminate bots and reduce unverified internet source . If man read the articles then man know what the implications of that to some of private relay . what Nostr need is more relay and not some sort of verification if web is trusted or not . In my opinion. It’s too early . It could be be some relay just for music only or for movie ( based on interest , is a good idea ). Socialisation on how to make relay and how to create Npub , identity in Nostr so far is the best
It’s good to think about how centralization can happen and how we can prevent it. Last thing we want is one big centralized WoT Service Provider. Which is why WoT SPs need to be open source. Ideally, every human being on the planet calculates trust metrics locally. Second best: some people pay a third party WoT SP to calculate metrics. But everyone with sufficient motivation can do it themselves, with the lowest barrier to entry as possible. And ideally, many people will do exactly that, even if it’s not every single nostr user. Similar to running a BTC node: not everyone runs a node personally, but most could, and many do.
> I love that we have all these WoT services and metrics popping up, but I am concerned that it could lead to centralization. Agree, I think each client should have a build-in lightweight WoT algorithm, we can archive that with a simple portable WoT library, it shouldn't sync its index 24/7 only once every couple days are enough. The index size should not be that much, I think it will be 2~3GB maximum if the user have a big network. cc: @Keyhan Alizadeh