Didn't want to further derail the exploding heads vote thread, so:
What are the criteria that should be applied when determining whether to defederate from an instance? And should there be a specific process to be followed, and what level of communication if any with the instance admins?
For context it may be useful to look at the history of the Fediblock tag in Mastodon, to see what sorts of stuff folks are dealing with historically in terms of both obvious and unremarkable bad actors (e.g., spam) and conflict over acceptability of types of speech and moderation standards.
(Not saying that folks need to embrace similar standards or practices, but it's useful to know what's been going on all this time, especially for folks who are new to the fediverse.)
For example:
- Presence of posts that violate this instance's "no bigotry" rule (Does it matter how prolific this type of content is on the target instance?)
- Instance rules that conflict with this instance's rules directly - if this instance blocks hate speech and the other instance explicitly allows it, for example.
- Admin non-response or unsatisfactory response to reported posts which violate community rules
- Not sure if there's a way in lemmy to track incoming/outgoing reports, but it would be useful for the community to have some idea here. NOT saying to expose the content of all reports, just an idea of volume.
- High volume of bad faith reports from the target instance on users here (e.g., if someone talks about racism here and a hostile instance reports it for "white genocide" or some other bs). This may seem obscure, but it's a real issue on Mastodon.
- Edited to add: Hosting communities whose stated purpose is to share content bigoted content
- Coordinating trolling, harassment, etc.
For reference, local rules:
Be respectful. Everyone should feel welcome here.
No bigotry - including racism, sexism, ableism, homophobia, transphobia, or xenophobia.
No Ads / Spamming.
No pornography.
There is no way we should defederate an instance because of this. Particularly, as we know reports will grow as the number of users does over time anyway.
Breaking the users' experience because your tooling is insufficient is a bad look.
The Beehaw method
Indeed.