There is a recommendation to silence the largest Brazilian server (ursal.zone), which is a left-wing, anti-fascist Brazilian instance that welcomes several black and LGBTQIA+ people.
The administrator of Mastodon Art without ascertaining the facts and without any evidence chose to block this instance after a single person defamed it.
I remember that at the time the Mastodon Art admin was even based on a poorly translated message in Portuguese. Which was bizarre, not to say prejudiced due to the lack of importance and negligence that he dealt with this Brazilian audience.
That is, a blocking list where an important instance may appear, for the sole reason of personal feud between administrators, is completely useless and a disservice to Mastodon.
This team should have a serious commitment to evaluating evidence, rather than following rumors and unsubstantiated claims.
Pete @Pete: "You won’t be able to stay ahead of the garbage if you try to block domain by domain. You want an automated solution that leverages the collective moderation efforts of the Fediverse. I use FediBlockHole (https://pypi.org/project/fediblockhole/) and sync with the oliphant unified tier 2 blocklist (https://codeberg.org/oliphant/blocklists). Set it up to run daily in cron and you’ll see an immediate improvement. Follow @fediblock who does a great job curating, de-duping and boosting #fediblock"
@mastodonmigration I am encouraged by efforts like oliphant lists and fediblockhole but I don't fully automate moderation because community standards are not uniform, and fully automated systems can be gamed.
As such I would recommend using such automation to notify and the action to only be taken after human review. The vast majority of time such a review would literally take a couple seconds (Nope...click lol) but a handful of times things are more nuanced.
Cooperative efforts are fantastic though. I hope they are refined and such a concept "fragments" into offerings to several like-minded groups of communities.
@msh@mastodonmigration@Teri_Kanefield@fediblock Whatever works for you. I prefer to block first and whitelist later. I’ve found it’s more effective in shielding users before they get dog piled, rather than reacting after they suffer through it.
@msh Have a look at oliphant's 100% list first, then see if literally any of the sites on them would be allowed under any reasonable "community standards."
I guarantee you'll end up with the same core blocklist. Other people have flagged the minefield before you, you don't have to let your users get blown up getting there!
@mastodonmigration@Teri_Kanefield If it's really bad and you're behind Cloudflare use their firewall to geo block big chunks of the planet until it slows down. Then get your fedi block lists in place and start un geo blocking.
@mastodonmigration@Teri_Kanefield if you are an instance admin, are marginalized, very visible online or otherwise at an elevated risk of harassment I also highly recommend locking your account (ie. require manual approval of all follow requests)
Malevolent trolls sometimes even have lurker accounts or friends that follow without ever interacting for the purposes of adding your content to a public timeline. Manual follow helps mitigate that by giving you an opportunity to reject follows and "leakage" to toxic servers.
Even if a requestor looks fine check the instance they are on. If their server doesn't pass the smell test do not interact even if they appear OK themselves. Best deny the request and block the server.
@mastodonmigration@Teri_Kanefield I would also like the names of the bad servers so I can forward it to the admin of my instance immediately and I'm sorry that she was a victim of harassment at least here we can block the instances where on centralized platforms you could only block individuals not groups
@mastodonmigration@Teri_Kanefield Everyone with a server should use that blacklist. Automate collecting it. That's your friend. If the jerky jerks are known jerky jerks, blacklisting works very well. Yes it does.
Add comment