I've been watching the conversations around whether or not instances set-up should include a (easily-overridden) default of importing a blocklist for known awful instances.
Those against it pattern match for me with a certain libertarian streak that doesn't want humans making decisions.
In economics, it's the gold standard and free markets. In politics, government is inefficient and should be minimised.
It's the idea that a system with human decision-makers could be undermined...
1/n
and thus humans should be out of the system. (Not unlike crypto, I guess.)
Thing is... systems with humans work better. They're more resilient and flexible if they actually adjust in a meaningful way. Government, done competently, is pretty amazing and achieves things we wouldn't do otherwise.
Yes, human-based systems can go wrong, but that's a reason to build them well, not give up and make something worse. And if they go wrong, try to fix them rather than destroy them.
2/n
When it comes to default blocklists, the concerns are often pretty hypothetical, while the blocklists and hideous blockable instances are very concrete. Maybe the concerns are forward-looking - what if it goes wrong? - but again rarely in a way that either allows for mitigations or acknowledges that the core devs pretty much have this power anyway, Better to discuss how power is managed than to pretend we're safe from abuse of power...
3/n
And at the end of it all, the same regular lesson:
Deciding to do nothing is not a neutral decision, whatever it feels like. It's often both a surprisingly opinionated decision and/or a bad decision.
4/4
(Oh, and deciding to not do something as a project management technique is a whole different kettle of fish.)
@skyeye I'm not quite sure exactly what you're saying, but I'm very much in favour of giving power to users - Fedi should allow a huge selection of server types and customisation to exactly how people want.
What I also want is a system that by default is friendly and easy to use and could support a migration of hundreds of millions of users (if only).
I don't see why, in theory, we can't aim for both. I'm not trying to be depressing! :)
@skyeye I'm actually pretty positive, because I've been reading how someone who's been working on blocking for ages sees things moving in the right direction (https://ubiqueros.com/notes/9ftp613p5n).
Just because I'm talking about the people who resist... I don't want to make them seem more numerous than they are.
My personal Fedi echo chamber is pretty positive about the idea. :)
Ha, look, I love the convo and the seeming support for blocklists and more nuanced blocklist management tools that I see. It’s pretty cool. But personally, it's so strange because I was demonized and harassed for these same ideas when I ran my old instance a few years ago. I remember people from Masto’s dev team telling me my thoughts around safety would ‘stifle growth’ and ‘take away free speech.’ It's just weird seeing people talk about it in a way that pushes it forward rather than demeaning the idea of a safer fedi. I'm gonna reflect on this.
Do you know if it's easy to enact a default-but-overrideable blocklist or mutelist in vanilla Mastodon?
@sgf There are no systems in which humans engage that are without humans.
Libertarians engage regularly in this rhetorical slight of hand where they pretend choosing to do nothing is not a choice. "Ah well, it was a disaster, but we did nothing, so we can not be blamed."
What they're describing aren't "systems without humans," they're "systems without empathy" or, more accurately, "systems where the humans are the biggest assholes imaginable."
@UP8 I think it's useful to distinguish server blocks from user blocks. Servers don't get blocked because someone gets angry. They tend to be blocked for deliberate, cool-headed decisions to set up servers that enable trolling, racism, coordinated attacks etc.
Maybe the future there is deep learning, but it's a problem right now, and we have pretty effective human-curated, explainable block lists right now.
@jztusk @sgf from the viewpoint of content-based filtering there is a spectrum of habitual and normalized microagressions such as the incorrect use of an "N-word". A simple filter can improve the ratio of productive to provocative.
I imagine these behaviors are "selfish memes" and don't necessarily reflect on people's deep character.
@jztusk @sgf you would condemn these people?
https://tootfinder.ch/index.php?query=nazi&submitquery=Search
who (exclusively of Ukrainians) mainly use the word are incorrectly. Only a handful of very old men survive the NASDP unless you count a handful of cosplayers like George Lincoln Rockwell.
You can pick better insults to sling at authoritarians.
if i am wrong and they rock a swastika I apologize in advance.
When it is family and friends you have to love people no matter what rabbit hole they go down
- warm regards
@UP8 @jztusk When people talk about the "n-word", they don't normally mean "Nazi", they mean a strong racist slur used against black people.
Ironically, it's also sometimes used by black people talking to other black people (trying to reclaim the word, perhaps), so it's a really bad candidate for a content-based filter that doesn't deeply understand context!
These people look fun:
@UP8 @jztusk I dislike this as a solution, in that simply ignoring bad behaviour and carrying on engaging on other content provides no feedback loop to discourage bad behaviour.
While individual bad behaviours might not reflect deep character, a failure to reflect on those behaviours and how they hurt people and try to improve with time does make me suspicious of that deep character!
@sgf Hrm ..... i think, such a functionality should ultimately be a opt-in ... not a opt-out.
Who will actually be responsible for said list? Which political interests are covered?
It's .... complex..... I like the idea of a bad-list. On the other hand, i don't like calling out ppl in public especially if we can be sure to get a fair amount of streisand effect from that...
Mh. ;)
@hackbyte I think there should be a default blocklist because anyone who sets up an instance who doesn't know about moderation and blocklists probably doesn't want to peer with poast.
You can see actual blocklists and who's responsible for them at https://writer.oliphant.social/oliphant/the-oliphant-social-blocklist .
Mastodon already has opinions on what the default acceptable behaviour should be - https://joinmastodon.org/covenant . Extending the recommendation from individuals to instance peering seems pretty reasonable.
@sgf Well ... sadly the standards for distributed lists will always succumb to the basic interests of the people who declare their own to be morally better than whatever they want to block.
And yes, i specially refer to mastodon.social and chaos.social ..... where you actually get a frickin instance block just because you might have a differing opinion to one of their users..
And because these instances have a relatively huge amount of users .... this actually creates problems and enforces recentralisation.
But that's just another completely bonkers aspect of the ongoing mastodontitis.
----
Before you ask. i'm a left leaning german, i dislike what goes on under the "free speech" umbrella, especially if it promotes hateful or even violent behaviour against anyone.
But, you should for example do not _need_ to change your avatar image to a pride flag, just to avoid getting scrutinized.
That's why i'm not for a default opt-out block list.
Make it opt-in for easy access .... ;)
@sgf Just another reason why i despise mastodon..... They'e acting as if _they_ are the fediverse......
But they're actually late to the party.
Sorry but my opinion stays. Public bad lists may exist but they can prose problems on their own.
Simon Frankau
skyeye
robryk
Allan Chow
Tom Bellin 
Paul Houle
jz.tusk
Gordon Oliver
Marsha
hackbyte Antifa (friendica) 13HB1