I don't see that as viable. Either you block all npubs your contacts don't already trust and close yourself off from the public, or a determined spammer will just freely cycle npubs, potentially going as far as a new npubs for every message, and invalidating the web of trust.
Discussion
Read through it. Web of trust is a score. You would get 0 at iniciation of the npub. With every follow, that follows you back, you get a plus. And with every follow of a person you get blocked you will get minus. When I would say I only see posts of People above 3 every bot would at least to lounch three new npubs to have one personalized attack.
I would not want to spend significant batterylive to send notes, likes and posts. When it would be somewhere around 0.2 seconds of smartphone power, yes. But you have to think of all the majority of people that have old smartphones to build this up without rejecting poorer users.
That score could be gamed quite easily though. Just make an army of bots to follow each other, to build up a good score.
But the score could be different from the perspective of every user. As if everyone has no score. But only users, that have a relation to me or to other users I follow. And as further away a user is from my social graph, ghe smaller its score.
But could be very difficult to find new legit user when I set the score too high.
That “but” is the main pitfall with #WoT, I think. If you’re filtering out #spammers by filtering out people with low WoT scores, then you’re also filtering out new people that you just haven’t connected with yet.
Yes this point is correct. I thnk it would be worth it, that clients try to implement thos metrics in a opt in way. This way it is possible to try them without harm. When it does not work, it will go to have no big effect then.
If setting it to 3 becomes standard, how do new identities gain sufficient trust to be visible to the people they need to trust them? If any positive number is chosen, new users are isolated and struggle to build trust. If zero or only slightly below, only a few blocks could sink them, and that could be trivial to achieve on simple polite difference of opinion. If set significantly below zero, we're effectively setting a maximum number of people you're allowed to piss off before we drop you and you just... Make a new npubs. None of these scenarios can simultaneously welcome new users AND provide a meaningful obstacle for spammers.
PoW isn't my favorite option, but I think there would be a large difference between enough difficulty to discourage wide spread spamming and enough to be of issue to actual users. 5-10 seconds of PoW in the background wouldn't really impact your battery life unless you were sending dozens, maybe hundreds, of events a day, but it means a spammer on a typical user device is limited to 6-12 spam messages a minute, which may add up to 8,640-17,280 messages a day, but spammers often rely on orders of magnitude more messages due to the terrible conversion rates for spam. It also means the spammer has to dedicate an average device to that full time, all day, to achieve that, increasing the hardware and electricity cost to do it. As an added bonus, bots that don't provide a meaningful service the community is willing to support now have a financial cost attached, discouraging them from polluting the network with well meaning but unappreciated bulk messages.
Yes I think this has to be invented. I defnitly support implementing it. When it works, there will be big support for this. If not it will die and nothing is lost. 👍
