I should probably finish the demo for the Vouch tool first but I REALLY want to begin looking into this Akismet for Abuse (that’s the operative name) tool. My idea for it is a bit of what I’ve seen so far in solutions from other people.
Client A) receives a post from someone else (
Actor A). Before even considering to parse the information request, the software/tooling that
Client A has aims to run the incoming post from
Actor A. In the land of HTTP, we can do a few things (check the IP it’s coming in from, host name) and go deeper (run nodeinfo checks to see if it matches a criteria like faulty software version, software known to be run by malicious folk like Gab). That could be a first flag but not enough to stop it. A few more tests can be run to see if should be filtered out (content checking - seeing if it doesn’t match a specific content filter, etc). Of course, there can be cases where you do want to let people in.
Client A can explicitly mark some people/agents/links as “safe” which would allow it to skip pass everything above here. One could go as far as ‘subscribing’ to another list of “friendly” domains.
Some parts of this is already done in the space of federation but having something like this exist outside of it will allow the IndieWeb to leverage it as well as any other tool that can communicate into the world of HTTP (maybe Dat folks can collab on a bridge?). I know I’ll be tinkering on something like this in the near(-ish) future because dozens of people will need something like this to counteract the vile nature of some people on the Internet. And the applications for this can grow. The most important thing here to not is that the range of the tool is up to the viewer. They can use this to set automatic blocks or use it to move those messages into a separate inbox (I’ll end up doing the separate inbox in my webmention server).
This is my solution/idea/draft, obviously using prior art from things like Blockbot on Twitter and how blocks can federate on the ActivityPub/ActivityStreams world.