Technically there is no need to publish a follow list, if a nostr client wants to show a feed of follows it just needs a list of contacts, this list does not need to exist in public or on relays, but the fact that it does is what makes the Web of Trust work.

Accidental

The only other Web of Trust that I know of is PGP’s Web of Trust, it is only used by IT and security people, which is also the reason why it is unusable for most people. It leaves no room for error because security has the highest priority in those circles.

But on nostr the Web of Trust is a side effect of just following people. You press Follow a few times and you are accidentally building a Web of Trust without even knowing it. It might not be perfect, but that’s also why it works, because it is so simple, like nostr itself.

Attempts to “fix” or improve the Web of Trust should be made with that simplicity in mind. Creating new lists with granularity of people you really trust, partially trust, don’t trust at all etc are all great, but this will make it possible for it to end up like PGP. Nostr’s Web of Trust works precisely because people don’t have to think about it.

There is also this idea that likes, reposts or any signal that can be made by users are useless because they can be gamed, but they can only be gamed if you don’t use a Web of Trust.

Web of Trust fixes almost everything

nostr:note1a8nq7kdqgr20dnvzmhv2a6quy8vx3l56pfzjlly3jv2u2l55exhs84cev8

Web of Trust fixes spam, bots, abuse of trending/hot algorithms

Zaps were supposed to fix the problem of gaming/abusing likes or reposts, but Web of Trust fixes it better. Zaps are useful for sending value, so let’s use it to send actual value. 1 sat zaps are not much different from Like-spam and considering the technical overhead required for zaps and possibility for timeouts and failures it’s actually worse.

So, Web of Trust Makes Likes Great Again, and with that you can make algorithms that are useful to the individual instead of the advertisers or platforms.

Example of a most basic Web of Trust filter (follows + follows of follows)

Nostur uses follows + follows of follows to filter spam and bots, and it uses likes and reposts from follows uniquely ranked to create Hot and Gallery feeds.

Another example is Doug Hoyte‘s https://oddbean.com, a community website where the content of the community is based on the administrator’s follows, the algorithm is incredibly simple and transparent, and it wouldn’t be possible if nostr did not have an accidental Web of Trust:

Challenges

Now there are still some challenges to be solved, new users being onboarded won’t have a Web of Trust yet, maybe we could give them training wheels, a preloaded Web of Trust until they followed enough people to use their own.

Also, new users don’t exist in anyone else’s Web of Trust yet, so they won’t be seen until someone follows them. This is not a good new user experience.

You also might miss someone’s post because they are not in your Web of Trust yet.

These are all challenges and I will keep exploring solutions for these.

Reach should be the result of how many people want or don’t want to listen to your message, there is no reason for a third party or algorithm to be required in this equation.

You should be in control of what you can or cannot see. Who you block should be up to you.

Ideally Nostur tries to reduce the need to require or trust any third party in order to communicate, unless there is no other way.

BUT

Nostur is built for iPhone, iPad and Mac users. In one way this is a limitation, (e.g., zap restrictions by Apple) but if we are using the Apple ecosystem already, we might as well use it to our advantage.

Nostur's target audience already trusts Apple with their data, their messages, photos, calendar, reminders, security cameras, etc.

We don’t need to store our data on random relays, and hope our contact list doesn’t get lost, or leak metadata to the public. We can store everything on our device, and sync using iCloud, it is reliable, and we already use it for far more important data anyway. Converting Apple users to Android or GrapheneOS is an uphill battle, I applaud people who take that route, but it’s not for Nostur users, we have people saying don't trust the media, don't use twitter, don't use iPhone, if we keep going, we end up sending smoke signals in a circle. As for interoperability, we can still publish or export using nostr native formats so you can use your data in another nostr app.

Nostur will publish to relays, but our data is on our device first, iCloud second, and third-party relays third. And you can still use a personal private relay for yourself if you want.

With this in mind, Nostur tries to let relays just be relays, not personal storage. Using relays for storage might end up centralising nostr, in the same way email, which used to be decentralised, is now mostly hosted by google and microsoft.

If you disagree with this mission or direction, Nostur's source is released under GPL-3.0 license, so feel free to fork and create a different version.

Ultimately, the goal of Nostur is not to get more people to use Nostur, but to get more people to use nostr.