That said, the recent spam attacks show the need for more advanced tools for server admins and mods to share information. The call for such tools became a point of contention during the first twitter migration last year and the fediverse and mastodon missed out on the support of black users, trans users, and members of other groups targeted by bad actors due to was was perceived to be unresponsiveness and "free speech absolutism" by some of the big players.
There was a sense that we were being told "this place is a paradise we don't need any tools to fight spam and hate"
I wonder what could have happened if there was simply... better listening. A better response to those concerns. It did not help that some "influencers" have always been invested in staying on twitter/X and even as they complain about what a terrible place it is they have no desire to build anything better. Because they have social power there and change might mean losing it. 2/
So maybe it was an uphill battle and all the listening in the world couldn't have changed the outcome. Which is that the fediverse has become a substantial community for academics, and some arts people but NOT the new home of Black Twitter or left commentators. (among other groups)
I would like to see tools for sharing blocks. Something very "opt in" and wiki-like where information could accumulate.
For example when viewing a flagged post could we see how many participating servers have blocked it too? Could we have a way for servers to be "friends" and blocks from friends would be highlighted?
A lot of work is being duplicated If the attacks get really bad people will get frustrated.
"This user's server is blocked by 13 servers you follow. Join them?" 4/
a) There is no defense for remaining on fascist twitter independently of how Mastodon behaves.
b) Moderation is fundamentally difficult but not impossible. The feature you suggest could amplify attacks against legit instances and cause users to flee - chaos.
I'm not black, I feel the absence of important groups here, I get that we need to own that problem.
The Mastodon dev community is pretty thoughtful, but the problem and it's solution are complex.
I want to be careful I sympathize with the frustratioin @futurebird and others are expressing. I am an admin of a small instance, I am not influential with the dev team who maintains Mastodon.
Moderation at the instance level is tough because it will cause collateral damage. Suppose a bad player on a "good" instance causes some people to block entire instance and that starts to be advertised?
A one way system of absolute punishment doesn't have much space for either justice or forgiveness. There's going to be plenty of people who are on a journey to civility who will need walking back from the wilderness.
If reports can be forwarded to the original server, then why not forwarded to an actual arbiter? A system of collecting evidence and providing more nuanced and reversible blocking for edge cases.
I mean, you will never get the whole Fediverse agree on the same arbiter (nor should it).
Which means that each instance has to choose which arbiter they agree upon, and if the instance that wants to use the blocklist and the instance that wants to be removed from the blocklist do not agree on who should be the arbiter, you keep the current fragmentation.
Still better than nothing, I guess.
@futurebird Sure.
I was not reacting directly to your proposal, but to the reaction to it that mentioned the risk of overblocking and the need to have a process for unblocks.
@futurebird This pretty quickly starts to look like algorithmic moderation. This isn't inherently a bad thing, "algorithm" after all essentially just means "repeatable rules". I think we have to be mindful to intentionally increase friction and human involvement in the process.
How about a service of sorts, that the user opts into, and it shares their blocks and mutes, in an anonymized, but possibly weighted manner (ie, server owners could get a heavier weight, people with a ton of followers get heavier weights, and users on certain pre-determined instances designated as marginalized get higher weights), which then can be fed into an instance's block/silence lists?
Once a certain number of "votes" (Blocks/mutes) happen, the account/domain is added to the list to be fed out to instances and users, via an auth token.
@ubergeek@futurebird This is pretty much exactly what thebad.space is trying to do. Unfortunately bad actors and software bugs led to temporary unjustified additions to the list about a year ago. The event caused huge drama here on the Fedi and eroded faith and unity in the project. mastodon.art is still defederated from tech.lgbt as a result of that fallout.
It's a real shame because I think it can still be really transformative, positive, and impactful. It just shows how complicated and sensitive this social problem is. Tech can give us tools to address this, but it is ultimately a social problem.
@AcornSquashbuckler@ubergeek@futurebird This is kind of why a transparent built into mastodon service is superior to and adjunct outside service or worse passing around “block lists” that have no notes, posts or reasons for the block. Is @ coolGuy35 @ tiny serve .io on the list because he’s a known bad actor and spammer or did he just say mean about ants and I put him on the list out of spite? If you can see related posts and history it’s transparent—
@futurebird@AcornSquashbuckler@ubergeek I agree completely and there’s some difficult levels in that. Eg in the drama around the bad space, an entire fake cultural rift was fabricated between Black and trans fedi, with accusations of transphobia or anti-Blackness whenever one argued with the other. Reactivity, hurt feelings, labels… it’s tricky.
@futurebird@AcornSquashbuckler@ubergeek Following what you said above about servers being "friends", what I think we'll see long term is consolidation towards a smaller number of larger instances that have each other on an allowlist. Maybe a web of trust type model to make it easier to include new/smaller instances like what exists for public key crypto (there must be trust & safety papers on all this). It's too easy for bad actors to spin up new instances for just a blocklist to be enough.
@futurebird@AcornSquashbuckler@ubergeek
I know a lot of folks were really against quote toots, but it might be good to have a block directory that recorded:
User and/or server blocked
Explanation for block
Example toot highlighting reason for block. Preferably a static capture of the offending toot so it:
a) Is preserved in amber as an example and can't be deleted.
b) Doesn't give traffic to the offender.
Notice if the block has been recinded, and if so by who & for what reason.
@futurebird The advantage I find more in Mastodon is the possibility to change of server. Different of main stream social media like if I have a Twitter account and wanted to migrate to other site, I need to start from zero where many people feel a very unpleasant to lose all contact.
At least in Mastodon have a "framework" in common, so if some server has rules I doesn't accept, I can move to another server, so I doesn't stay locked by one server. At least in harassment and similar need a lot of effort between servers admins and their users to block, but it's possible.
Now on spam side, this I don't know how to, put a limit to been approved before can send a DMs or PMs, or need a vote to decide to blacklist some server.
For now, the best is deal what we have in our hands now and see if someone with more experience or interest in create a new API like Mastodon where it's deal with spam and harassment. This is the fun part of a FOSS, right?
@futurebird as a non tech admin of a small server, I also would like to see a much easier and well known way to share block lists. For anyone reading this, the folks at Alaskan Social would be happy to talk about linking up with like minded servers. Most of all I just want it to be possible for regular people to operate a safe server without having to spend ten years learning tech jargon etc
If anyone is curious about why so many are skeptical about this in relation to past deeds by the Bad Space and Nivenly Foundation, here's a doc you can read:
@futurebird I've long had this idea for like...reputational moderation. Something where if someone has a good rep and is trusted, their decision on moderation activities like blocks is more trusted. I wonder if you could build this as a kind of social graph, where folks could publish their blocks and the reasons, and you could indicate you agreed, and then if enough people you trusted had blocked someone, you would autoblock them...
@futurebird hmmm now I'm considering making something like this. I'll see if there are admin panels with plugin capability, otherwise a separate tool
And taking @doctormo's criticism into account as well, because I think that's solvable while still making/using the tool
I think that the fediverse is the top community for leftists, although not "left commentators" if that means influencers: the natural home for influencers of all kinds is Bluesky.
People suggesting block list sharing do have to take the history of the Bad Space project into account: a bad-faith project that attempted to block kolektiva.social and pretty much every other large leftist server, as well as a long list of servers favored by trans people. Bad Space supporters said that any pushback to the idea that one's server was going to be listed on a global block list must be due to anti-Blackness.
@RichPuchalsky@futurebird I have not encountered any block list projects without the “just trust me bro you gotta block all these” flaw. I’m proposing that these list grow organically out of each server choosing to share or not share who they block (along with what was blocked and any notes but anonymized) this information could be made not annon for servers that choose to share lists.
This is very different from the giant spreadsheet someone once said we had to block.
I understand how your proposal would be different: I'm just writing that there is lingering bad feeling about that other project that people would have to deal with upfront.
Some kind of informational list about which servers are choosing to block with other servers would also be helpful to people choosing a server in the first place.
I don't remember this being the discourse at the time.
More "learn how to use the tools that exist, then you'll be able to make constructive criticism/proposals".
I remember people getting extremely butt-hurt when they were amicably told "get your own instance, that's how you can tailor moderation to fit your needs", though.