Let me put a phrase into your mind: nonconsensual virality. It's why quote-posts on Twitter led to harassment. People's words stolen, taken out of context, used purely to incite a mob of griefers. The answer is to give #Mastodon users control over whether someone else can quote-post them, with a simple "quote or not" setting that can be set before or after the post goes up. We should be allowed to stop people from taking our posts viral without our consent.
TikTok already allows users to prevent others from turning their posts into a duet, which is roughly the same thing. If TikTok can do it, surely Mastodon can.
@annaleen ā€˜surely’ is doing a lot of work there. your posts can already be quote posted on akkoma, on fedibird (a mastodon fork), on misskey. QT is primarily a user interface design. one server can’t restrict how another displays content, unless you advocate for a nuclear option like defederation from those that display a post linking to your post in a way that the linked post doesn’t want…
@mikarv are you saying that a person on my server could set their post to not be quotable, but another server could allow me to quote it because their configuration doesn't care about the "don't quote-post" setting?

@annaleen @mikarv

in a federated world you cannot control what you post in public

not just qt

also deleting or expiring posts

and replies with visibility set to unlisted

and unsearchable without hashtags

@annaleen yep, it’s not really possible to use technical means to bind what people do with unencrypted information they receive other than find a means of sanctioning them through another channel. digital rights management of eg video has this problem, basically have to put chips in peoples hardware to stop them being able to compute with certain kinds of information (Apple Fairplay, Google Widevine…)
@mikarv so it sounds like what we'd need are 1) quote-post functionality with user opt-out; and 2) a quote-post policy that instances could agree to (which I am not pretending is EASY I'm just saying that's what we'd need).
@annaleen i think instances respecting requests are good, but in the longer run the fediverse is not just mastodon, but so many activitypub services. so if you’re just interested in regulating notable mastodon instances, might work, however I think we should be opening up a diversity of functionality across the fediverse, not locking it down. one lesson from do not track signals: it will not be clear to define what is or is not a ā€˜quote’.
@annaleen it already may be hard to ensure instances respect deletion syndication requests, and deletion is pretty easy to define. expressing preferences over how your post appears when federated to and used within another service which may look more like YouTube (PeerTube) or Tumblr (Misskey) or Reddit (Lemmy)…

@annaleen completely agree! There's been some discussion here about how to implement it:
https://github.com/mastodon/mastodon/issues/14762 https://github.com/mastodon/mastodon/issues/8565

Hopefully this wave of users also is bringing developers who can help make this happen. It seems like a really interesting problem, figuring out how to do this in a federated environment with non-compliant servers, without fundamentally changing how posts are distributed.

Enable Twitter-style Reply Controls on a Per-Toot Basis Ā· Issue #14762 Ā· mastodon/mastodon

Pitch Twitter's reply model has been extended with some LJ-like features. Replies to a tweet can now be restricted to: Replies only from accounts @-mentioned in the tweet Replies only from accounts...

GitHub
@annaleen Yes. I’ve brought up the TikTok analogy multiple times. You can make a post available for dueting and/or stitching or not. It seems like a solid model to copy.
@coty @annaleen This does nothing to stop copy/paste of a quote the old fashioned way. Even on tick tock it can still be done. It’s as easy as a screenshot. This won’t fix the problem.
@ercanbrack @annaleen What is the problem that we’re trying to fix? You’ve pointed out that people can quote to abuse without the affordance of official quoting support. You are correct that there is no stopping that. We can, however, make non-abusive quoting work better.
@coty @annaleen If the problem is ā€œquotingā€, then I’d say it has no solution in a social platform—pleople can and will quote. However, like theft deterrence reduces theft, making quoting harder to do, by not building in automatic easy to use quoting, can be just enough friction to reduce its likelihood. Friction as deterrence may be the only solution.
@ercanbrack @annaleen So I don’t think quoting is the problem, per se, and I don’t know why it should be hard to quote my posts if I don’t mind their being quoted. Unlike Twitter but like TikTok, I do think it should be hard to quote if I do mind. Opt-in quotability.

@annaleen Mastodon *can't* because of the nature of the decentralisation:

Not everyone runs Mastodon, and not everyone runs unpatched Mastodon.

Which is why getting a resolution on this that most people are ok with is fairly urgent. The longer it takes, the more likely people just decides to provide the worst possible totally unconstrained variant.

@annaleen

TikTok is a much different development environment than Mastodon.

TikTok has a large staff focused on delivering features for a single platform.

Mastodon is much smaller and so the delivery of features is slower.

It is also an Open Source platform so, as many have pointed out, anyone can fork the code and build out a griefer's paradise of features into a new instance.

@lolzac Yep. This is where we get into the difficult thicket of "we need a policy that many instances agree to." Why does it always turn out that tech is easy and policy is hard?

@annaleen

As always I default to Sartre

"Hell is other people".

Policy is very difficult in this sort of case because the entire system is decentralised and no matter how many people you get to agree with you there is always a way for some people to avoid any majority consensus and do their own thing.

Its a bit of a libertarian nightmare which makes it important to find a good "neighbourhood" to set up in and to be able to just block servers

@annaleen

For instance, Gab runs on a version of Mastodon and I don't know if they still try to distribute their "content" but I don't see it on any timeline because my instance blocks it.

If people from the Washington Post *need* QTs then they can get their own instance and add the feature. It won't render properly unless other instances support the feature though.

(I think. I haven't really dug into the ActivityPub format and looked at how its fallbacks work)

@lolzac @annaleen

Any content delivered can be reproduced, so unless WP is blocked completely, they get posts delivered/can pull and as such quote-toot to their hearts content and have the means and qualification to implement quote-toots to their instance if they want to.
Most likely it will retoot just like long toots do, which some instances seem to have.

@lolzac @annaleen

May I point out that truthSocial is a (defederated) mastodon instance.

The great thing about the fediverse is, you don't need to agree with anybody. There are always plenty you can agree with. You can choose the amount of friction and griefers you want to have. Though the size of your playground may depend on your behavior

@annaleen That's not really how it works though. Quote tweeting is just putting the letters QT at the end of your toot and then a link to the other toot. A quote toot basically just shows that link inline.
@ada @annaleen I guess If all ActivityPub implementations agreed not to embed notes with an "embedable: false" property, maybe it could work... but you'd have to extend the note type from ActivityPub and add that property that everyone can technically still ignore... so yeah... not an easy fix anyway, and I don't see implementations like SoapBox agreing with that any time soon...
@cody What is not implemented through protocol can be implemented through policy: We can ban instances that don't want to play nice.
@annaleen agree with the sentiment, agree with the goal, not sure about "the answer is". Though, i hope one is found.

@jeffrizzo @annaleen

Probably worth op editing it to ā€œanā€ answer is eh?

@annaleen
Remember, though, that this is also telling some people that they cannot use a particular tool for self-promotion (while others can) because they are too easily victimized, which is patently inequitable.
@wcbdata No. It is allowing users to choose whether a post can go viral or not. They can set a post to be quotable or not quotable.
@annaleen
I respectfully disagree. I think that a world where some people have to self-censor or forgo particular activities or spaces in order to avoid abuse is an abusive world, and I don't think that's the right answer. I have less firsthand experience with that sort of risk management than many - probably most - do, but I hesitate to make that the core of a solution...
@annaleen
Let me place a caveat on that: if being racist, sexist, anti-LGBTQ, abusive, etm. is the practice someone is being told to forgo, they can just fuck right off. My right to an opinion ends where it harms someone else.
@wcbdata 1. Nothing stops screenshot-sharing or copy-past link-quoting now. 2. This is the reason the Fediverse has both moderation and deferation. @annaleen
@reneestephen @annaleen
Agreed, but the general opinion I'm seeing from POC and non-cis-male folk is that moderation and defederation really only covers ya if you're white and male.
@wcbdata @reneestephen @annaleen You mean moderators refuse to protect users who aren’t white and male? That’s horrible. Are there particular instances that are standing down or is it all of them?

@dudleysaunders @wcbdata @reneestephen @annaleen

Definitely not all of them. Federal, worldwide, fairly easy to set up for the average webmasters => hence very diverse.

The fediverse covers all sides of the spectrum!

@wcbdata It's not self-censoring to set a post to not go viral. It's a way of limiting its reach. As a journalist and author, I feel that I have to censor myself far more when I think a post might go viral, and get into the eyeballs of people who might misunderstand it or weaponize it against me.
@annaleen
I'm sure... And I doubt it's always an easy choice whether to share your likes or expertise or skill and risk it being twisted. I think that some of the folks concerned about this feature are feeling like they don't have to be quite so constrained, but I also recognize that even in the face of a terrible choice - it's still a choice. I wish I had an answer...

@annaleen @wcbdata
"It's not self-censoring to set a post to not go viral."

There isn't a thing you can do about it aside from going offline or behind password protection.

As a community, not having a convenient button helps keeping the community civil. And your personal tools to moderate here are extensive, as well as the moderation tools for admins.

I don't think to go too restrictive is helpful. At some point you introduce self-censureship for fear to upset the community admin.

@wcbdata @annaleen

I'm not getting the censorship tie-in either. I think that might be from another conversation you were having elsewhere.

From what I've seen, the speed of unintentional consequences is what makes quote posting powerful and dangerous at the same time b/c of the telephone game dynamics of human communications.

I like giving people who feel like they can handle that danger the chance to use the utility, but also appreciate proceeding with caution b/c it could alter culture.

@wcbdata @annaleen Oh, then you just make the default setting for a new user no-qt. Opt in then has the friction, not opt-out.

@annaleen @wcbdata I don't think this addresses the concern, which isn't about someone else choosing your user settings. It's that the result of people making that choice for themselves is still inequitable access, because people who are more frequent targets of harassment will face more pressure to turn it off

As I see it, this is an argument against QTs, not against giving users a choice if QTs are implemented

@pompelon @annaleen @wcbdata Quotes can be done no matter what, as long as a person is posting publicly. I believe the only solution is Friction as Deterrence—don’t make it easier to quote. Quoting will still happen, but it won’t be so easy.
@annaleen @wcbdata So long as, as with the post itself, the setting may be edited after the fact. Plenty of people on the bird site have posted things they honestly believed was perfectly innocent, until someone saw it, saw a different meaning, and QTed it.
@eyrea @annaleen @wcbdata This is why my preferred solution would include an option for users to approve each QT, like you can choose to approve follows (approved QTs could still be boosted of course). It wouldn't scale to viral level but would be good at the level of community discussion.
@outeast @annaleen @wcbdata That's an interesting idea. I see "it wouldn't scale to viral levels" as a feature, not a blocker. If the OP throws up their hands and says, "I can't keep up with this", it cools down the torrent of hot takes. So many viral posts with QTs are just different people saying the same thing over and over again. And, of course, sometimes they're orchestrated that way, where all these "individual" hot takes just happen to have the exact same wording.
@annaleen @wcbdata Couple of things about this idea:
When you post anything online, anywhere on the internet, you’ve essentially given up control over that content. There’s a difference between controlled use of a cloud database and social media, obviously, but social media’s nature of posting to a wide audience means what you say is ā€œpublic.ā€
All the legal niceties in the world won’t keep someone from screenshotting your post and using that to repost and link to you.
@laurajhmarshall @annaleen @[email protected] There’s also all the difference in the world between a screenshot and a #qt though.
@rabbigabriel @annaleen True, but it’s a quick search away to find the person whose post is featured in the screenshot. If someone truly wants to harass or intimidate, they’ll make that effort.
@laurajhmarshall That’s no reason to make harassment easier…

@rabbigabriel @laurajhmarshall I've never really got that argument.

Using screenshots is good for preventing vitality of the original post, and thus disincentivising trolling for clicks (a different case against QTs). That's why people use them to call out bigotry etc.

However, screenshots mean the QTer cannot be identified, so they are also effective for malicious purposes. Incite a pile-on and (unlike with QTs) no one can see the source of the harassment.

@annaleen can't someone nonconsensually viralate you with screenshoots, though? i like the idea of mastodon users controlling whether they can be quote-posted but i'm not sure it would do much to solve this problem
@bro @annaleen Yes, but the amount of friction involved in these actions is incredibly important. If the poster has to take a screenshot, and then the viewers have to manually type a username in rather than clicking a link, a huge proportion of potential interaction is filtered out.
@ZaneSelvans @bro I think that friction is the point. Make it slightly harder, and you can cut down on the easy drive-by mobbing. But we still need good moderation to solve the bigger problems of abuse.
@annaleen For sure! The right amount of friction where we want it, when we want it can radically reshape how the platform works. I feel like I've seen a surprising number of posts pointing out that something is already possible in theory (with a lot of friction) as if that means providing the option of making it easy is pointless.
@annaleen @ZaneSelvans @bro the fact that you would have to do so much work to violate someone’s consent also makes it a lot more obvious to others viewing the viral post that it was probably not with their consent. I rarely blocked on twitter (preferred muting since they don’t know they’re muted) but obvious non-consensual screenshot tweets got block and report. but it was hard to know for sure. :(
@r343l @annaleen @bro And I would imagine that this interpretation is also really context dependent -- like a small account screenshotting something from a big account / powerful person to hold them accountable or avoid their wrath is totally different from a big account screenshotting a nobody and punching down.
@r343l @annaleen @ZaneSelvans @bro
I'm all for not having QTs at all, or failing that, giving the poster control over whether their posts can be QTd.
It will definitely reduce drive-bys and casual bad actors,
*however* I'm completely unconvinced this will deter determined bad actors and their mobs, which are the real problem on the birdsite and elsewhere.

@annaleen @ZaneSelvans @bro

Ok, I can go with the argument of ā€œFriction as deterrenceā€.