šŸ’¾ Archived View for idiomdrottning.org ā€ŗ defederate-meta captured on 2024-05-12 at 15:29:01. Gemini links have been rewritten to link to archived content

View Raw

More Information

ā¬…ļø Previous capture (2023-12-28)

āž”ļø Next capture (2024-05-26)

-=-=-=-=-=-=-

Defederate Meta?

I havenā€™t made my mind up and Iā€™ll keep editing this page until I have.

Iā€™m leaning towards ā€œnoā€ on the block, but itā€™s not an easy call and I havenā€™t made up my mind yet. The block side has two big arguments in their favor:

1. Google federated to XMPP (a.k.a. Jabber) and we were like ā€œOK coolā€ and I and many others federated with them. Then they shut off federation and moved all their users to their own closed network.

2. Meta had a big part to play in the Cambridge Analytica debacle. Not only is Meta the enemy (although I generally prefer to follow the Kumbaya Doctrine ā€” if I truly believed they were reformed), theyā€™re scary to deal with in case theyā€™ll start data mining Fedi stuff.

Thatā€™s in addition to the argument that others brought up: that the toxic culture on Facebook might place undue burden on moderators.

Wim wrote in, asking:

What are your arguments in favour of not blocking a Facebook/Meta-owned instance?

My arguments against blocking are weaker than the arguments for blocking so maybe I should block after all. Iā€™m on the fence over here šŸ¤·šŸ»ā€ā™€ļø

But for clarityā€™s sake, here are my arguments against blocking:

Iā€™m lazy and not super eager to keep up with all the fediblocklists. I do block a couple of dozen instances, and more as I run into bad ones (Iā€™m not in the free speech extremism camp), but Iā€™m not being super diligent with keeping up with the larger convo around this.

I value a heterogenous block landscape (i.e. I donā€™t like a ā€œthese are the good sites, and we all allow each other and we have the exact same block listā€ world) for three reasons:

1. If some ā€œgoodā€ sites can manage to keep up the convo with the asshole sites, like how Torvalds clearly and plainly spoke out against ā€œantiwokeā€ the other day, thatā€™s an overall good for society and can help deradicalize ā€˜em. Insert Paul Baranā€™s classic ā€œdistributed networkā€ image here. Itā€™s not all good because it also somewhat legitimizes the asshole sites, but there are good aspects to it.

2. Some instances are extra vulnerable and need an extra harsh instance block list (many even blocking mastodon.social). Since thatā€™s legit, then the converse is also true: some instances can handle a less harsh instance block list.

3. Even the ā€œgood sideā€ has values it hasnā€™t sorted out yet, and never will because discourse is never-ending and the arc of the moral universe is a wild and unruly rock-tumbler of chaos. As Pratchett put it: ā€œPulling together is the aim of despotism and tyranny. Free men pull in all kinds of directions.ā€ Iā€™ve blocked people other ā€œgoodā€ sites have allowed (for example Assange supporters) and Iā€™m sure vice versa too (such as mastodon.social).

I especially think that a site is responsible for its own users conduct only. social.kernel.org federates with many sites I have blocked. Thatā€™s up to s.k.o. and their mods to the extent their own users want or need that protection. I donā€™t block ā€œby associationā€, I only block sites that are directly bad.

I lost a lot of friends as the pandemic hit, and disproportionally (but not exclusively) those lost friends were other women. People I was hanging out with IRL and whose online interface is Facebook and other silo sites. Since I donā€™t have, and never have had, a Facebook account, we drifted apart pretty much instantly when we couldnā€™t meet IRL. If I could have some of them back by the silo sites going ActivityPub, thatā€™s a huge boon. When Google Talk went XMPP, I reconnected with friends I hadnā€™t talked to for a few years. (And lost them abruptly when they defederated.)

The following argument is a super-weak argument. Itā€™s stupid. But itā€™s still a factor: Itā€™s flattering on an emotional level that the silo sites want to become part of Fediverse, part of the social network that we built on rock and roll. Itā€™s tempting to go along with that.

And, as per my aforementioned Kumbaya Doctrine, federation going to be the ultimate outcome for them once they are truly reformed. (But I donā€™t think they ever will be or can be reformed given how much damage they have done, and probablyl itā€™s better that Metaā€™s sites just go away and wither.)

While I do think servers should be FOSS and I do agree with the Franklin Street Statement, I was still reading Twitter and Tumblr and Blogspot and Angelfire and Geocities over RSS as if they were any other server. I think there is a mile of difference between those kinds of proprietary-backend, open-protocolā€“frontend sites vs a ā€œlog-in onlyā€ site like Instagram or Facebook or Google Plus.

They could be a boon for Fediā€™s problem with scaling and resource use. Thatā€™s not super likely since theyā€™re so evil, but they have contributed to server stack development so they might. šŸ¤·šŸ»ā€ā™€ļø

Ultimately a network on thousands of servers eating up the Earthā€™s precious resources is not more efficient because itā€™s labeled ā€œthe peopleā€™s stickā€.

The Email Sitch

I came up with another argument, or rather, an analogous situation.

I run an email server. I have not defederated with Apple, Google, and Hotmail. And most other services, like Proton Mail and Posteo, donā€™t either.

Even though we know for sure that Google and Hotmail on a massive, site-wide, systematic level actually literally reads, scans, indexes the contents of peopleā€™s personal email and has used that data for ads (Google says they stopped doing that in 2017).

That is the sick and twisted reality we live in.

On the email side, most instances (probably better known as ā€œemail service providersā€) current practice is that we can ask and beg and plead our users to ask their friends & fam to sober the heck up and switch away from these untrustworthy giants, or to turn on e2ee (I know e2ee is not a solution for the Fediverse since it's world-readable. Even though we're learning over and over and over again that world-readable, casual, off-the-cuff conversation has problems and I don't think we can solve 'em. šŸ¤·šŸ»ā€ā™€ļø) , but we donā€™t block email coming to and from them. Thatā€™s been my mentality too, as an admin.

As a user, I love that most people do have email, even if itā€™s ā€œevilā€ email. I donā€™t have iMessage, WhatsApp, Facebook etc and I rarely text since I donā€™t have a smartphone. I use email to keep in touch with my friends. I can low-key promote non-ā€œevilā€ email (with mediocre success) but in the end I still reply even when itā€™s from a gmail or hotmail address. (And a lot of my peeps on even those services are on e2ee, PGP specifically.)

Thatā€™s not meant as a line-in-the-sand, glove-tossing war cry running up that hill to die on. Far from it.

Iā€™m on the fence about blocking meta. I think some of the pro-federating-with-meta people have used super weak arguments, bad-faith reasoning, false info etc.

I donā€™t know yet which of these three are right:

I have a slight lean towards the first of those three but thatā€™s just preliminary. I wanna think more about it. And hear more good arguments about it (Iā€™ve heard enough bad ones).

Follow-ups

Computers and Blues wrote in, suggesting:

we could this time just listen to the people raising these concerns, and maybe future reality will be a bit less sick and twisted.

Thirty-One Privacy and Civil Liberties Organizations Urge Google to Suspend Gmail, from 2004

The time for listening back then wouldā€™ve been before gmail was widely adopted and before Facebook & Instagram were widely adopted. Iā€™ve been refusing & protesting Facebook for fifteen years or more but now they are widely adopted and there are already millions of people on there.

We want people off of Meta (and off of Gmail). Thatā€™s a given. Question is: does defederating help or hurt with that?

In addition, I šŸ’Æ donā€™t fault, blame, scold specific instances from defederating. That has been one of the most sickening arguments from the pro-federating crowd. I think the blocking instances are doing the best for their users locally šŸ‘šŸ»

I guess my main concern is thatā€¦ fedi is like inherently problematic. By design. Not sure how humanity can recover from it or adapt to it.

The timeline management issueā€¦ I wanna break it down to three parts:

1. Toxic content. That is a huge issue. Itā€™s pure poison over there.

2. Overwhelm on a protocol / s2s level. Yes. Already with Lemmy weā€™re seeing dropped posts and API misfires.

3. Overwhelm on a human level. IDK, I donā€™t have a problem with that. I get like three or four posts per day on here. Itā€™s crickets. Although you follow ten times as many people as I do so maybe you have more experience with the potential issues. Yeah, client side list tools are needed. Thatā€™s not a protocol issue. (I mean, above it being an inherent flaw of the mental ice age caused by world-readable casual conversations that I beg every waking moment to wake up from and it was just a dream and weā€™re a handful of friends with a pot of tea, a deck of cards, and rhubarb pie and we are not just a life flickering quickly past on a punch card in mother computerā€™s offloading central šŸ˜­)

Iā€™m not onboard with the unified policy thing (the "pact" part of blocking) since I value the heterogenous blocking landscape so much. I donā€™t support scabs so I hope the ā€œstrikeā€ analogy doesnā€™t hold up (i.e. if I think about it more, and come to the conclusion that "no, this is exactly analogous to a strike", then I'd have to change my position and join the "pact" šŸ˜°).

I think the whole idea of ā€œall of fedi needs to do the same thing and follow the will of the council and be on the secret admin Discordā€ is bad. Different instances has different needs is a good thing; we can and should help each other, of course, but thereā€™s a difference between helping each other and becoming each other.

At the end of the day fedi would ultimately be a super dumb idea with tons of overhead and climate-wrecking clunkiness compared to a centralized site so if we were to throw away our heterogenity which is the one supposed merit of decentralization, what are we even doing? Why does the council then donā€™t get together and start a centralized site that we all sign up for and obey? ā† The counter-point to what I just wrote is my own existentialist ethics, probably better known as ā€œya gotta case-by-case itā€, a.k.a. some issues are different than others; some can be handled heterogenously and others concern all of fedi and needs to be handled collectively. As an analogy, in union work itā€™s fine if one workplace fights for the right to go home from work early and another fights for the right to work nights, while they both need to get together to stop climate change since that is a looming disaster thatā€™s gonna ruin it for everybody.

So thatā€™s why the larger philosophical question of ā€œis blocking good or badā€ is vital; is it an unambiguous disaster like climate change or is it an opportunity for fedi to break down the wall and liberate our aunts and uncles from their silos? If you defeat the empire you become the empire, the say, and if the silo sites that I hate so much are becoming like us, isnā€™t that a good thing? The iron curtain is coming down. Are we the side that locks our people in or are we the side that the people are running to and longing for?

Going back to the email analogy. Hotmail launched in the summer of 1996 as one of the first webmail sites. It got millions of users quickly. It was many peopleā€™s first or only personal email account. It wasnā€™t compatible with pop3 or imap back then, but you could email with other people (and mailing lists) over SMTP.

Imagine if they hadnā€™t been compatible with email at first but still somehow had amassed millions of users (which to me sounds dumb and impossible but that's just what has happened with other sites like Discord, Instagram, Facebook, TikTokā€”people love to sign up šŸ¤¦šŸ»ā€ā™€ļø). That more people were on Hotmailā€™s own proprietary network than on normal email. (What Tutanota wishes could happen for them and their own fake email "network". šŸ¤¬)

And then letā€™s say that if after almost twenty years decided to open up so that other people could email their users and their users could email other people. By then, in this alternative universe, email is a tiny retro alt cult utopia and Hotmail juggernaut with ten times more users. Is them opening up a good thing or a bad thing? Itā€™s a difficult question, right?

Thatā€™s a hypothetical, but there are some real-world things like the death of Shockwave Flash and the death of ActiveX web stuff, and we rightly celebrated and welcomed those horror shows getting replaced by open standards.

Yā€™all who seem to have made up your minds: good for you! I donā€™t keep up with tech stuff and I didnā€™t see this coming. To me it sounds, at first, like a mind-blowing dream. Facebook are finally tearing down their own silo wall, the wall Iā€™ve been throwing rocks at for two decades, this is what I wanted and wished forā€”and then some very wise peeps on our side is saying ā€œhold your wooden horses for three seconds, maybe this is a poison in a pretty pill, theyā€™re trying to destroy usā€, and some foolish peeps on our side is trying to cling on to some sense of power or I-dunno-what, to keep the status quo as if it was an acceptable or good or OK status quo, and I am having a hard time telling wisdom from foolery right now.

If Twitter federated, Iā€™d rejoice! And then promptly instance block them because the current Twitter as it exists today is a messed up place when it comes to values and culture. And maybe itā€™s the same with Facebook.

I guess I donā€™t have the underlying warm love and hope for fedi as some of yā€™all do since I know there are so many bad actors and harassment sites out there. I donā€™t see fedi as this successful, wonderful thing. I see it as a smoldering mess already, a post-apocalyptic wasteland of small enclaves who are trying to band together in the face of wave after wave of radiationā€”and the radiation is using the very same tools themselves. Weā€™re already in a mess. Things are already bad. The order and harmony is illusory and flimsy.

The harasser-and-troll sites can easily see everything we post. Itā€™s just fuel to their fire. There is no normal, casual, kitchen-conversation here. Itā€™s all glass walls where weā€™re the human zoo animals.

Csepp wrote in, saying:

And if the rumors are true, they are not even going to federate with small instances, only a few big ones

That would be insanely messed up if so. That would cement my opinion šŸ’Æ against them and for the block instantly if that turns out to be true and I hope that plenty of the anti-block people would, too. I hadnā€™t seen that rumor.

Instagram specifically has had a history of ā€œembrace and extendā€, to come out the gate all campfire songs and RSS and then once they had critical mass they slammed the locks down tight. Thatā€™s a pattern that many of these proprietary services follow. When they are upstarts, they are all ā€œdonā€™t be evil, look at our awesome API, check out our protocols, here is a repoā€ and once the gorilla has reached 800 pounds itā€™s all EULA and handcuffs gand a boot stamping on a human faceā€”for ever.

Iā€™ve mentioned the pro-block sideā€™s moderation argument in passing five or six times and I think itā€™s a great argument and I donā€™t dispute it. Fediā€™s model of small instances work well.

But unlike the other arguments (embrace/extend; crash/DoS levels of scale issues; tracking/tracing/ads/datamining), the moderation argument doesnā€™t merit a pact.

Small instances that want to protect their users are ļæ½ļæ½ļæ½ allowed to block and itā€™s a great idea from them. Promising your users that you wonā€™t federate is a good idea for some instances, and I get that itā€™s also a kind of ā€œpactā€, but for the sake of clear semiotics Iā€™ll reserve the word ā€œpactā€ in this discussion for the fedipact, i.e. admins promising each other that theyā€™ll block.

So weā€™re kind of discussing two things here:

The moderation argument is great in the block question but a lot weaker in the pact question. Itā€™s not non-zeroā€”knowing that you have your fellow admins with you is valuableā€”but itā€™s not as much of a be-allā€“end-all in the pact question as it is in the block question.

In addition, the other arguments have a much stronger claim for pre-emptive caution with a block; moderation is a ā€œwould-be-niceā€ to block ahead of time but blocking after giving it a few months (or days or seconds or whatever it takes until the rotten eggs start overwhelming the good ones).

Repeating thoese two last points for clarity: the case for homogenous ā€œpactā€-style blocking and for pre-emptive ā€œday zeroā€ blocking is a lot stronger with the other arguments than with the moderation argument.

Thatā€™s not discounting the merit of the moderation argument in sich at all. Itā€™s huge and itā€™s many peopleā€™s biggest and most urgent concernā€”but in the block question, not in the pact question.

Thatā€™s why I want to spend much more of my own focus on the other arguments. The moderation argument can be summed up with ā€œif you have an instance with a low mod/user ratio, and you wanna protect those uses, of course blockā€.

I also want to mention in passing the whole ad hominem spiel from the anti-block side. The whole ā€œnot that kind of openā€, ā€œseems like these people donā€™t wanna be part of normal society after allā€ super tired and unā€“thought-through spiel. That line of thinking is misguided. Instead, Iā€™d say we on the fedi/foss side take our responsibility to not make society even worse than it already is, is all. We fight against our own complicity in the banality of evil with mediocre results but to the best of our ability and with, for the most part, good intentions.

Hereā€™s where I am right now.

There are cases historically when open standards have won over embrace / extend / extinguish. PNG won over gif. CC won over OGL. JavaScript won over ActiveX. Even in the face of (or sometimes even with the support of) some pretty dang evil corporations. Itā€™s not hopeless.

Meta does have a particularly bad track record in that regard, so Iā€™m gonna try to stay wide awake. Iā€™m usually not great at keeping in the loop (I'm less of a "finger on the pulse" writer and more of a "uh, hold on, let's think this through" writer) so letā€™s help each other out here.

And as Iā€™ve already said, if Meta is allowlisting and only federating with a couple of specific big instances like .social, (as opposed to blocklisting problematic instances) that is messed up and a crime against the World of Ends.

The Kumbaya Doctrine

Franklin Street Statement on Freedom and Network Services

World of Ends

A few months later, tinyrabbit wrote:

Instances often de-federate with other instances that are toxic because of bad moderation or CoC.

I think this is very good and should continue and is much more necessary on Fedi than on, say, email or XMPP, since (de facto, not de jure) most convos here on Fedi are world-readable which brings out the sealions in droves.

They often de-federate from those who federate with those. This second degree de-federation is a stop gap to truly be insulated from the crap

I think this is something to be much more careful about. I mean, those who federate with the toxic-est intances often themselves have bad moderation and bad-actors, but then thatā€™s in and of itself a good reason to deferate from them. It can suck when an instance about a topic I thought I was into turns out to be federating with adn schmoozing up to some bad apples but then the schmoozing is the problem, not the federation.

What does Threads moderation and CoC look like? Do they live up to fediverse standards (as you perceive them)? Do their incentives for running an instance matter?

Theyā€™re bad. Dot social also is bad but still there are plenty of really really good people on dot social among all the bad. So what to do vs a popular server like that? If youā€™re running a ā€œthis is very expressedly a safe spaceā€ like dot art then defederation is part of the mandate, but itā€™s less of a clear cut for other servers. I still havenā€™t decided.

As Iā€™ve also said before:

For those running a medium or big server, federating with a big and poorly moderated server like dot social or threads is a huge burden.

For those on single user instances, or instances with only a few users, itā€™s more possible because then you can get more milage out of blocking specific bad posters.