Please add that a user can quote themselves

It's already there:

With all being unselected by default (but quote-boosting your own toots would always be allowed).

:)

Is "Users on other instances" basically anyone?

Please add that a user can quote themselves

It's already there:

With all being unselected by default (but quote-boosting your own toots would always be allowed).

:)

That’s why I deleted the comment 🙈 I didn’t see it at first

I very much want to see this! And one thing I'd like to add in favor regarding abuse: as the ability to display a linked post as an inline quote is purely a client-side presentation matter, it's already possible for malicious instances to do this without the quoted user's consent, and for malicious users to do it (with worse ux) via screenshot+link. Making it a first-class feature with opt-in gating makes it so that instances that bypass the flag are explicitly doing something abusive that would be grounds for not federating with them, rather than offering an "extension feature".

Is "Users on other instances" basically anyone?

Yep. That was the intent. One could do it with radio (either-or) boxes instead, e.g.:

• Only users that follow me
• Any user on my instance and any user that follows me (from another instance)
• Any user on any (federated) instance

This might be clearer, but the control would be less fine-grained. I would be fine with either.

Cool! Yeah, I primarily would like to avoid specific instances, so maybe something that inherits my/my instances defed/blocks would be amazing.

Cool! Yeah, I primarily would like to avoid specific instances, so maybe something that inherits my/my instances defed/blocks would be amazing.

I think that for instances blocked / not federated with my instance, I wouldn't get notifications from quote-boosts anyway? At least that's what I assumed.

Is "Users on other instances" basically anyone?

Yep. That was the intent. One could do it with radio (either-or) boxes instead, e.g.:

• Only users that follow me
• Any user on my instance and any user that follows me (from another instance)
• Any user on any (federated) instance

This might be clearer, but the control would be less fine-grained. I would be fine with either.

Actually I don’t think the first one should be “follows me” (because I didn’t pick them, there is not that high a trust there) - I’d prefer “that I follow” (because I chose them, so higher trust, and if they misuse, I can just unfollow) ref my post here:

1. I can quote myself
2. People I follow can quote post me

https://pleroma.patricia.no/notice/APbJEHeNcqwrKZVEO0

I’m fine with more fine grained after that

Actually I don’t think the first one should be “follows me” (because I didn’t pick them, there is not that high a trust there) - I’d prefer “that I follow” (because I chose them, so higher trust, and if they misuse, I can just unfollow) ref my post here:

The rationale for offering both choices in my original post was:

• users that follow me being a dedicated choice
  • This is the most likely group to want to quote-boost you
  • You can block people to prevent them from following you
• users that I follow not being the "least constrained choice"
  • Being a user with small follower-base, I might want to follow people in a certain field with many more followers, simply because they are important in that field.
  • This does not imply that I trust those people, and in particular that might quote-boost a regular reply of mine, exposing me to the much larger visibility of their user-base.

So I think there are valid use-cases for both options. I assume that if you are a popular person with more people following you than the other way around, you will likely prefer "people I follow". OTOH if you are a person who follows much more people than you have followers, then "people who follow me" is likely the more trusted group.

edit: I think this is a good reason for maintaining the individual checkboxes proposed at the top, but of course I am open to other suggestions.

So I think there are valid use-cases for both options. I assume that if you are a popular person with more people following you than the other way around, you will likely prefer "people I follow". OTOH if you are a person who follows much more people than you have followers, then "people who follow me" is likely the more trusted group.

Good points, so I have changed my mind to “both” 😂

I want to add a comment here on the supposed “truth” that quote posting is inherently toxic (modified from my reply here https://pleroma.patricia.no/notice/APbYd7wiLFDZmlNmz2):

I never quote post to dunk. I explicitly don’t do that. Not because I’m nice, but because I would be amplifying what I think is a bad person. I would be giving them a new audience. So no. I fundamentally disagree. I quote post to digress. To make another point. To expand. And at least 50% of the time I quote myself, because I want to make another related point. I also quote tweet good points in the replies into threads to give attribution. I’m sorry people are used to toxic uses. What I see here are the worst type of dunk: the screenshot dunk. The person being dunked on are not even notified. And it’s been happening to me here all week. If someone is going to dunk on me unfairly I’d like to know, I want to block that person.

This is sort of a soft "no" argument, and I don't mean to be argumentative.

I think quote boosting contributes to negative parts of the culture on other platforms, where people spend their time trying to build an audience rather than engaging with others. If you say something interesting, my options on Mastodon are "spread this to others so they can also engage with you", or "engage with you myself". My options do not include "spread this to others with a perfunctory comment stuck on top of it, so their replies and engagement will go to me instead of you."

I think even allowing this as an option would enable this kind of anti-community sentiment to build, even if the majority of uses are completely harmless or even worthwhile, as many people have outlined here. The culture could shift significantly enough that people just don't engage with you if you don't have quote boosting enabled, because they can go find someone else's post to QB, and that will give them better engagement. (Even if that's not how they're consciously thinking about what they're doing.) Communities could form on Mastodon that all allow QBs, and then they sit around QBing each other all day in a way that attracts attention and followers and sucks the air out of more vibrant spaces. Once people start following a bunch of popular folks, they're less likely to engage with small accounts.

So... soft no. I'm usually for giving users more options to control their own experience, but I also have feelings about keeping Mastodon weird.

As I elaborated here, unless it's unrestricted by default (because most people won't change defaults) it's going to limit commentary and talking e.g. about news or tech as a sort of "forwarder"/educational introducer to another group. Regular replies just aren't that great for that, since they signal to the original poster it's meant to be a direct discussion with them when it's not.

I prefer opt-out too (by default, others are allowed to quote you) but I can see more room for objection to that, and I'd be happy with either. I don't want the choice between opt-in and opt-out to be a sticking point that blocks this. The reason I'm not really concerned about folks not opting in is that, a lot of the time and probably most of the time, the folks you want to quote in a positive way actively want to be quoted. Either they're yourself (citing past posts in a new context) or they're seeking broader exposure for the ideas they're posting outside their own circles - journalists, activists, researchers, public health advocates, etc.

I'd like to add some thoughts here on why I would use quoting rather than just boosting for many of these. If I boost something from outside my audience's circles to my audience, on that first boost, folks can see that it's coming through me and it's introduced to them with my credibility. Ideally they can assume I have some reason to believe the message I'm boosting is true and important. Except they can't, because I might have just boosted it from someone else I trusted to have vetted it in that way, who in turn boosted it from someone else they trusted to have vetted it in that way, etc. And as such, they're less likely to re-boost. I know I'm less likely to re-boost something I see boosted by someone I trust, but haven't researched myself, for this reason.

With quoting, the dynamic changes completely. If I quote something from outside my circles and explain why it's important to me and/or to my audience, now my followers have something they can boost with my name attached to it, without feeling like they're taking on personal responsibility for having vetted the quoted post, because I'm the one putting my credibility out there as the source of vetting. This allows the quoted information to spread further than it otherwise would, and in a way that preserves authorship credit and connects people who discover it indirectly through me back to the original source, in a way that, over time, builds trust and community.

(Full disclosure: I am not well aware of inner workings of Mastodon and I'm acting like a layman here, knowing only basics of signing)

I was thinking earlier today about this issue and had a thought that just like admins patch their instances to have more than 500 characters, or that they notify users if they get blocked by someone, they can also patch Mastodon to be able to quote anyone, ignoring those restrictions. And this is definitely something worth fixing.

So I was wondering how it can be done, and I think the solution could be to make quotes require other instance to sign the post with quotation, which ensures that: 1) the instance from where quote comes is in total control of quotations (e.g. it can check if user quoting has permission to do so, and if they don't, then reject signing that quotation, which makes all other instances that would to receive such quote also reject it since it's unsigned or displayed with a warning), 2) the integrity of quote is preserved and it's unmodified (so nobody can modify your ‘I hate this job’ to, say, ‘I hate kittens’).

However, this solution comes with a huge drawback: any instance that receives the quote must contact the instance where the quote comes from for a public key in order verify that signature is indeed correct. Don't think you can really just attach public key to a post, because then malicious instances can just use their own key for signing, so perhaps some trust keyring / cache can ever so slightly mitigate this? That is given that the key is created per instance, rather than per user on that instance.

This is how it would look in real world:

1. When Alice is about to post a quotation of Bob's post, Alice's instance (let's call it ‘allium.social’) sends a request to Bob's instance (let it be ‘begonia.io’), asking to sign Alice's new post that includes quotation of Bob's post.
2. Bob's settings are set to allow only those who he follows to quote his posts, so begonia.io checks if Alice from allium.social is indeed followed by Bob.
3. Since Bob does follow Alice, begonia.io now does additional verification to ensure that quotation in post received from allium.social fully matches Bob's original post.
4. It does, so begonia.io now uses its public key and signs Alice's post in full (with its attachments, contents, etc; effectively sealing it from any context manipulations) and sends back to allium.social.
5. allium.social now propagates that signed post to other instances, among which is camellia.xyz, on which user Charlie is following Alice.
6. Upon receiving this post camellia.xyz, which was unaware of Bob's existence before this very moment, checks if it has a public key stored for begonia.io.
7. Unfortunately, it does not, so camellia.xyz now contacts begonia.io to receive its public key. Begonia.io gladly shares the key, and camellia.xyz uses it to check signature of Alice's post.
8. It matches. Camellia.xyz now shows post normally: in Charlie's feed, and in federated feed (since Alice posted it with public privacy).

​

1. Unfortunately, some people can't just live without drama. One of those people is Mallory. She saw Bob's post and decided to commit a mischief. Thanks to her skills, she already got a set up on her instance (mallow.sh) in which she can modify various aspects of her posts before they go live.
2. So she does exactly that, she quotes Bob, but changes quote to include various spelling mistakes. She then adds in her post, ‘Bob, can you type normally? Having a hard time understanding what you even write’.
3. She publishes it. mallow.sh now asks Bob's instance, begonia.io to sign Mallory's post.
4. begonia.io respects Bob's settings and checks if Mallory is followed by Bob.
5. No, Bob does not follow Mallory, who'd want to follow people like that anyway. begonia.io rejects signing Mallory's post.
6. You would think the story would end here, but Mallory knew this would happen, but just forgot disable sending of that redundant request. Mallow.sh ignores the error and proceeds to propagate Mallory's post to other instances where people follow Mallory.
7. One of the people following Mallory is Dave on dracula.fr, so his instance receives Mallory's post.
8. Immediately dracula.fr notices that Mallory's post has a quote, yet lacks any signature. It still shows the post, but displays a big red warning on top of it, stating: ‘dracula.fr cannot confirm the authenticity of the quote below. Be careful: it may have been faked or modified.’
9. And this is how Dave saw Mallory's post, so he immediately calls her out, ‘why did you fake the post?’ and after not receiving a good answer decides to unfollow Mallory, as well as reports her posts so dracula.fr moderators ban mallow.sh as malicious instance. Mallory totally deserved that.

Questions

• It's not very clear what to do when quotes post updates, though: should the quote be left unchanged or should the instance that has the quote be aware of all the instances that quoted the post and notify them, so they can update their post accordingly and request to re-sign it?
• And even some of more important matters: can one revoke quotations of their posts per quote post basics (e.g. allow Bob to remove quote from Alice's post)?
• Accordingly: can the preferences be applied retroactively? In which case I assume Mastodon would have to store and serve per-user signing keys, so when preferences change old key gets revoked and new one is created and all the quoting instances have to re-sign their posts containing the quote.

With that being said, this feature really needs a lot of thought to be put into it to ensure it's not abused and works as one would expect... The scale is completely different than on birdsite, which has central authority and control over all posts ever made on there. For fediverse feature like this has to be carefully considered and discussed with developers of other federative software to ensure the compatibility and to prevent abuse. Imagine Mallory's post was received by someone using Pleroma and they would never have guessed that Mallory modified the quote because Pleroma never implemented any of these signature checks. Yikes, definitely not a good situation.

Displaying a quotation is fundamentally a client side presentation matter. There is no amount of protocol level or cryptographic stuff you can do to prevent it. That's impossible for the same reason working DRM is impossible.

The status quo is that instances could just start offering quoting, with no opt-out or opt-in flag, and present that as a feature rather than something malicious. The proposal here is that we codify how quoting is supposed to work and create an expectation that instances honor the flag. Then, if instances pop up with "you can quote anyone!" as a feature, the rest of the community can react and say "nope, that's an instance trying to be abusive, defederate!"

Displaying a quotation is fundamentally a client side presentation matter. There is no amount of protocol level or cryptographic stuff you can do to prevent it.

This much is true. Though part of the issue here is not whether it's possible to link to another post, or how that post is displayed, but whether clients should provide a button to make doing so trivial.

The status quo is that instances could just start offering quoting, with no opt-out or opt-in flag, and present that as a feature rather than something malicious. The proposal here is that we codify how quoting is supposed to work and create an expectation that instances honor the flag.

I disagree. The status quo is that (the overwhelming majority of) instances don't offer quoting. You're talking about changing that to a situation where most instances offer a limited form of quoting and the Fediverse accepts the consequences, both good and bad.

With respect to whether instances could offer quoting, individual instances striking out on their own are subject to the response of the community in exactly the same way now as they would be under your hypothetical scenario. There would be "an expectation that instances honor the flag", but right now the expectation is that instances don't offer quoting!

If you feel strongly that the status quo is not determinate enough, the implication of that is not that we have to codify some form of quoting. Rather it means we need to consider several possibilities, one of which is establishing a consensus that clients and instances providing quoting functionality is not acceptable, and defederating from those that do. That's a perfectly legitimate way to resolve the problem, if there is one. In fact it's probably the most straight forward way to codify the status quo.

The status quo is that (the overwhelming majority of) instances don't offer quoting.

This isn’t actually true. I see the equivalent of quote posting all the time. People are linking to other posts. There is a not-too-great preview in many clients. What is missing is the fairness to the quoted: to get notified and then review the quote post.

They will then have the option to at least block the quote poster. And if this feature existed, change their settings.

This isn’t actually true. I see the equivalent of quote posting all the time. People are linking to other posts. There is a not-too-great preview in many clients. What is missing is the fairness to the quoted: to get notified and then review the quote post.

They will then have the option to at least block the quote poster.

I think that's moving the goalposts more than a little. After all, I don't think the people who are advocating for this feature in this issue and elsewhere would be satisfied if the devs came in and said "we already have 'the equivalent of quote posting', we're going to change this by allowing users to block linking to their posts on Mastodon." Surely that's not what you have in mind!

The request, as I understand it, involves three things:

1. On Mastodon, a quote-boosted post will show the entire text content of the linked post, instead of (at best) a tiny preview.
2. On Mastodon, the interface will include a button that will allow quickly creating a new quote-boost, instead of having to manually copy and paste the URL into a new post yourself.
3. Users will have the ability to opt out of allowing the above two features to be applied to their posts.

I think that's a significant change, and has to be debated on the merits. I expect that (1) and (2) would have negative effects on the culture and community of Mastodon and the Fediverse more generally, and I don't think that (3) would be sufficient to prevent those effects.

À propos, one feature I would support is an opt-in option to receive a notification if your instance sees a post containing a URL to one of your posts. This makes sense because it's more akin to a @username mention. I think that would be useful and would provide some of the "ability to block" parts of this request, without requiring the implementation of quote boosts. Maybe that's worth a separate feature request?

I dislike the quote options, even under restrictions due to one key aspect: The fediverse.

With the way Mastodon works can someone start their own instance and just quote everyone and everything without any real limitation (outside of being blocked perhaps).

Quoting doesn't really bring anything useful tbh.
Just boost and reply... done.

Perhaps one way/solution could be that the instance can recognize a link to a post, and if it doesn't belong to blocked users/instances would render the content of the post in its entirety.

This could have benefits:

• No modified posts. It's pulled directly from the source through the link.
• Something else I forgot...

Tho, downsides are also possible:

• Mastodon needs to recognize a post link from any other one. Usually not an issue as those are normally https://domain.tld/@user/timestamp, but you never know.
• No real option to avoid quoting like this, unless there was some complex communication and verification of settings bin place.

Displaying a quotation is fundamentally a client side presentation matter. There is no amount of protocol level or cryptographic stuff you can do to prevent it. That's impossible for the same reason working DRM is impossible.

Yes.

Perhaps one way/solution could be that the instance can recognize a link to a post, and if it doesn't belong to blocked users/instances would render the content of the post in its entirety.

Quotes would obviously not copy the content, but instead refer to the original by ways of some message-id or link.

But I would really like to not discuss implementation detail in this thread. It's about the merits of the feature and raising awareness of it, so that it might come to mastodon at some point.

As I elaborated #12753 (comment), unless it's unrestricted by default (because most people won't change defaults) it's going to limit commentary and talking e.g. about news or tech as a sort of "forwarder"/educational introducer to another group. Regular replies just aren't that great for that, since they signal to the original poster it's meant to be a direct discussion with them when it's not.

News sources and similar users will likely allow everyone to quote-boost. Mastodon instances can choose whether they do opt-in or opt-out, and for specific fedi communities, it might make sense to allow instance-wide quoting by default. But I am strongly suggesting opt-in, because this is quite a sensitive issue for many current mastodon users.

The reason I'm not really concerned about folks not opting in is that, a lot of the time and probably most of the time, the folks you want to quote in a positive way actively want to be quoted.

I agree!

News sources and similar users will likely allow everyone to quote-boost.

Can also be relevant for memes, like quoting a meme but with a new, sub group specific joke, e.g. a generic meme with a new tech context joke. I'm sure there are plenty other non-malicious uses I'm forgetting right now.

I don't think opt-in makes much sense, it's so limiting and well, outdated. Like we're all still watch centralized TV and not everyone might be potentially making interesting and quote-worthy content now.

@Andre601

As someone who was harassed through quoting feature on Twitter (both with calls for harassment and through context omission (where person extracts one tweet from the thread and pulls strawman)), I don't really want to see quotes being available on Mastodon without any options for me where I can prevent them in the first place or do that retroactively.

Mastodon didn't add them precisely for this reason: if you make quotes freely available, you open a huge opportunity for abuse.

If harrasment isn't enough for you, another problem on the birdsite is PQRTs where people quote you through their private accounts for the sole purpose to (excuse my French) sh-t on you (e.g. homophobes/transphobes) or (don't know if that's better or worse) sexually objectify you (mostly a problem for women), so there are many people who are uncomfortable with these too, and understandably so.

I cannot support anything remotely similar to that and will advocate against freely available quote reposts on Mastodon. If we are going to add quotes, they must be controllable to the extent possible within the federated context.

@Andre601

As someone who was harassed through quoting feature on Twitter (both with calls for harassment and through context omission (where person extracts one tweet from the thread and pulls strawman)), I don't really want to see quotes being available on Mastodon without any options for me where I can prevent them in the first place or do that retroactively.

Mastodon didn't add them precisely for this reason: if you make quotes freely available, you open a huge opportunity for abuse.

If harrasment isn't enough for you, another problem on the birdsite is PQRTs where people quote you through their private accounts for the sole purpose to (excuse my French) sh-t on you (e.g. homophobes/transphobes) or (don't know if that's better or worse) sexually objectify you (mostly a problem for women), so there are many people who are uncomfortable with these too, and understandably so.

I cannot support anything remotely similar to that and will advocate against freely available quote reposts on Mastodon. If we are going to add quotes, they must be controllable to the extent possible within the federated context.

Well, the sad truth here is, that this feature will be in Mastodon sooner or later, just not in base Mastodon itself.
It's a sad downside of open source, especially with one that allows different instances to communicate with each other.
Someone will implement it in a fork and someone will abuse it for sure. It's only a matter of time here. It may not render on normal Mastodon, but that wouldn't matter for the person using it as they have it on their instance.

Harassmend is obviously a bad thing and it's sad that there isn't any real good solution here that would make this usable without this risk... it's part of the nature of these things sadly.

But tbh, everything can become a potential case of abuse.
Character limit? Someone can post some 10k spam into your timeline or DMs or similar.
Embedding links (Which from what I saw Mastodon supports)? someone can abuse that to link to scam pages or similar.
Heck, even gifs can be abused by intentionally posting seizure-causing stuff.

Again, I don't want to underplay the potential harassment this could cause. I just wanted to point out other sources that can become a source of harassment and abuse simply due to the system allowing free configuration and alteration of how everything works in one instance... It's a double-edge sword of sorts that comes with the whole Federation stuff here.

Isn't that kind of how all source links and quotes usually work? It's a quote, not a reply, so I don't see why there would need to be "parent context". This ties back into me saying if people want a post that looks like reply, they're probably going to use the reply function for that which already exists. Maybe that's just me, but I don't fully get it.

A quote-boost won't look like a reply, which does not show the parent context (i.e. the quote) in your timeline unless you click on the reply. The quote-boost is a reply, but shows the parent context (i.e. the quote) in the timeline. That is the visual difference, and what makes it a "quote" and not the same as a "reply". They don't look the same.

BoredDan: The thing about making a quote boost a reply is that it's not always what people actually want from the feature. Part of the point is often to create a new thread that is addressed to your followers.

Replies are already new threads, and you can also set the privacy scope on the reply (i.e. to your followers). The thing you can't do is evade the moderations rights of the original poster.

The majority (I would argue vast majority) use for what you're asking is to post something that the person would get blocked for if they said it in the original thread. That's exactly the majority use of the feature on twitter, to evade the moderation rights of the original poster. Something mastodon has never, I think rightfully, supported.

The majority (I would argue vast majority)

That seems a bit of a big claim to make without numbers. That it happens is undisputed.

Just as a note, if you let users block quotes of their public posts (rather than addressing removal of toxic quotes purely via instance moderators and defederation) you get these questions which at least if I didn't miss it weren't really answered yet.

That seems a bit of a big claim to make without numbers. That it happens is undisputed.

Honestly, I didn't think anyone would question the cesspool that is current twitter. Virtually all of the "quote-tweets" I see in the media are exactly quotes from people who would certainly be blocked if they made those comments in the original thread. Websites create entire articles of quote-tweets "throwing shade" and worse because that apparently drives more views of their site.

Just as a note, if you let users block quotes of their public posts (rather than addressing removal of toxic quotes purely via instance moderators and defederation) you get these questions which at least if I didn't miss it weren't really answered yet.

Under the proposal I'm endorsing a "quote-boost" would just be a reply with an enhanced UI, all moderation remains the same. So there should be no new questions as no new post, boost, or moderation functionality is being proposed, just a UI enhancement.

"Virtually all of the "quote-tweets" I see in the media…" is the equivalent of "Virtually all of the left hand turns I see in the police blotter are pedestrian crashes."

As an experiment I'm running through a several lists on Twitter (one I maintain, one I subscribe two, and two generated by the algorithm) and tallying QT's

Public health and epidemiology: https://twitter.com/i/lists/1233779498031575040

Friendly endorsement |||||||||||||||
Self QT or organizing own thread ||||||||||||||
Encourage action or comment in response to neutral news ||||||||
Provide source for data ||
Description of RT'd link |
RT other's tweet about self |
Critique public official's statement |

— Things below this line might be called "negativity"

QT as public reply (including agreement) |||||
Rebut misinformation |
Hostile dunk |

Ukraine/Russia crisis list: https://twitter.com/i/lists/1486852497104908293

Friendly endorsement |||||||||||||||
Encourage action or comment in response to neutral news ||||||||||
Self QT or organizing own thread |

QT as public reply (including agreement) |||||
Rebut misinformation ||
Hostile dunk (including reframing news w/ insults) |

Okay, here's the current trending topic, ClimateEmergency, looking at "Top Tweets." Same rubric.

Friendly endorsement ||||
Critique public official's statement |
Encourage action or comment in response to neutral news |
Self QT or organizing own thread |

Hostile dunk |

There's a good amount of hostility in this list, but actually very few QT's so I'm moving on after less data. The hostility was largely in the non-QT posts.

And trending topic Ben Shapiro, looking at "Latest Tweets." Probably our best chance for detecting hostility, and we do.

Description of RT'd link |

QT as public reply |
Hostile dunk |||||||||||

Okay, what have we learned? QT's have many uses. They can be a major tool in productive discussion and a major tool in a pile-on.

In my view, disabling QT's is not the right solution to the dunking problem, given the many types of constructive interactions it enables.

shrug I don't think your experiment is more valid than mine. I think most of your examples don't fit with the underlying premise of my assertion.

In my view, disabling QT's is not the right solution to the dunking problem, given the many types of constructive interactions it enables.

QT's are not enabled on mastodon, so it isn't a question of disabling anything. The discussion is whether or not (for as far as I can see, the first time) mastodon should allow the transfer of moderation rights from the original poster.

For what it's worth, in your anecdotal sample I didn't see a single instance of a positive QT that wouldn't work just as well as a "quote-boost" reply in mastodon. That is, with an enhanced UI that provided the "quote" context in the timeline display.

For what it's worth, in your anecdotal sample I didn't see a single instance of a positive QT that wouldn't work just as well as a "quote-boost" reply in mastodon. That is, with an enhanced UI that provided the "quote" context in the timeline display.

This proposal to make all Quote Boosts into replies sacrifices the user's power to organize a thread. The quote boost becomes a part of the original thread, instead of part of the meaning-making done by the quoter. Here, from the first list I investigated, is a 31-post threat with seven such QT's. https://twitter.com/jeffgilchrist/status/1607379781892575234

I can say without hesitation that much of what I've learned about COVID in the past three years has come from effective use of this tool by experts to organize, explain, and cite their thoughts.

My own work on documenting evolving situations of violent conflict in Bolivia relies extensively on this tool to provide comprehensive updates, with options for readers to see background information in separate threads. Again, this cannot be achieved by threading these QB/QT's as replies. https://twitter.com/CarwilBJ/status/1196877235455696897

Basically, the difference here is between being able to quote someone (or translate them) in the middle of your own document, or only on a one-page flyer about their quote. Both are actually useful.

The discussion is whether or not (for as far as I can see, the first time) mastodon should allow the transfer of moderation rights from the original poster.

What moderation rights are you talking about here? You keep talking about transferring moderation rights, but right now a user has no rights to moderate their replies, only to report to their and/or the repliers server, and to mute and block.

I'm assuming the moderation of replies refers to the ability to disallow them.

Under the proposal I'm endorsing a "quote-boost" would just be a reply with an enhanced UI, all moderation remains the same

Quoting is a different and maybe in some ways more important task than responding, again see my questions. These don't arise if you can't reply, you just don't reply and interact then. But that doesn't work if you want to more generally talk about a public item, unless we accept that things strangely can't even be talked about (or we need to avoid sourcing & maybe even paraphrasing them when we do). That doesn't seem like a good climate, and the moderation situation doesn't seem "the same".

Edit: to clarify, right now, sourcing is always possible but might be clumsy via e.g. screenshot. But if quoting exists but then is explicitly turned off by someone on a public thing, would using a screenshot now be a rule violation by circumventing the obvious wish? If yes, how do you even then refer to things? If no, shouldn't quoting just work, maybe? It seems confusing to me.

You could see this as an iterative process, where various options don't conflict with each other, but some can be quicker/safer features that solve a particular set of use cases but don't preclude later work to address others. Then you can start out with the quicker and safer approaches, see how they go (and give people some tools more quickly), and evolve from there. For example a path may be:

(1) Implement "context boost" or whatever you want to call it, allowing people the option to boost their reply in a thread with the context of what they are replying to. This is quite safe, since boosting replies already exists. It is probably the fastest thing to implement (though I know nothing about the Mastodon code to judge that).

(2) Implement "safe" quote boosts, where the post you are quoting needs to opt into this one way or another. This allows you to try out quote boosts to see how they go, in a safer way than just replicating retweets.

(3) Potentially evolve in various ways to more extensive quote boosts. Maybe play around with defaults, so now the default of a post is to allow quote boosts. Maybe make those always allowed, but have a mechanism for someone to retroactively protect themselves if someone quote boosts in way that is causing them trouble.

I am suspicious of any claims about particular choices having no significant abuse concerns, or driving so much abuse that they aren't viable... this is going to depend a lot of many other characteristics of the social network, and will probably change significantly over time on Mastodon if the network keeps scaling up. So I don't think it make sense to see this as "here's the decision, and that is that" but something that will evolve over time based on the behavior on the network.

CarwilB: This proposal to make all Quote Boosts into replies sacrifices the user's power to organize a thread.

I don't think this is the case, all replies to replies are in their own thread and can have privacy individually set.

Again, this cannot be achieved by threading these QB/QT's as replies.

Given that replies each start a new thread, I don't see why this would be the case.

What moderation rights are you talking about here? You keep talking about transferring moderation rights, but right now a user has no rights to moderate their replies, only to report to their and/or the repliers server, and to mute and block.

Part of this is protection for the future. Creating a quote-boost that is not considered a reply creates a requirement to now consider both features when enhancing moderation in the future for either feature. For example, there are requests to make all replies visible to the original poster on their instance (i.e. automatically added/required as a mention) so they can even know what is being said on their post even if they aren't explicitly mentioned on every reply. Would such a feature apply to a quote-boost that is not modeled as a reply?

Implementing as a reply allows one construct to be used for both ways of displaying a reply (which as you said, this is really about enhancing the UI, not inventing new features).

CarwilB: This proposal to make all Quote Boosts into replies sacrifices the user's power to organize a thread.

I don't think this is the case, all replies to replies are in their own thread and can have privacy individually set.

Again, this cannot be achieved by threading these QB/QT's as replies.

Given that replies each start a new thread, I don't see why this would be the case.

This is just the problem. In an effective thread using QB/QT's the various quote tweets are strung together with one another to form a new thread. Like this (as linked in my previous post):

Formatting them as replies attaches them to the tweets they quote, right? Instead of each other.

Formatting them as replies attaches them to the tweets they quote, right? Instead of each other.

They contain a reference field that connects them to the parent (i.e. the quote), so you're saying what you would like is to post a QT as a reply in another post? This is the first I've heard of that as a request and I'm not sure using boost as the mechanism to create the feature would address that either.

Thank you for your patience, I'll noodle on it.

So I don't think it make sense to see this as "here's the decision, and that is that" but something that will evolve over time based on the behavior on the network.

"It will evolve" seems again like dodging how it will work in practice.

If you add a quote but people can't quote things with it (which your gradual suggestions boil down to), I think you should expect there's a risk they'll start or keep doing something else like they are now. Right now, the obvious screenshot workaround seems to be something worse with less notifications if desired for the target, less accessibility for any viewer, and more work for the quoter. If people keep using that instead, is there really even a point in trying to limit who can quote? Kind of obvious maybe to try to prevent it via rules, but will that be accepted at all because that might be seen as ban to talk about things even in benign ways, and/or how should people then talk about things instead?

I think just not caring about these questions and adding this with no answers will lead to undesired results.

Honestly, I didn't think anyone would question the cesspool that is current twitter.

Footnote: cesspool is now a mainstream term:
https://youtu.be/FhZUiGH832I

I'm not sure if this has been brought up yet, but unlike Twitter, it would be ideal if any QTs made are posted as replies to the original post, in order to keep the discussion within the same thread.

This should completely avoid the 'not talking directly to the poster and addressing your audience instead' performative aspect of QTs on Twitter. It's effectively the same as currently replying to a post on Mastodon and then boosting your own post, except you're also quoting what they said.

This is basically short for 'quote reply and boost'.

Formatting quotes visually as responses below has been suggested already, multiple times, including responses why it's a bad idea for some uses like the big one of adding a source/quote to support a claim you make that is otherwise not meant to start a discussion with the original source. Please at least scan over the above comments before chiming in.

Edit: for a more practical example why quotes can be weird as a response, can you imagine you talked about the president to your friends and every time you did, you would be forced to discuss that with the president themselves? That seems super impractical and to miss the point of talking about something, which I would have thought is the main use of a quote. So I don't get that proposal, personally.

Edit 2: I also don't see what's bad about that "performative" quote use, as long as the quoted source isn't directly spammed and attacked. But isn't that more likely if the quote is visually a direct response rather than a more separate framing with its own discussion space? So it seems counter-productive for that, too.

@ell1e Thanks. I did skim quickly through the whole thread, but there's a lot that has been discussed including related concepts and extensive previous context that I did not have the time to read thoroughly and wasn't sure if it was specifically this point or not, so I wanted to throw this in for consideration just in case which can otherwise be ignored. That's a fair point against reply quoting which I hadn't considered before, as I was thinking about how forums and Reddit handle it where users sometimes quote someone in their reply.

I think it'd be nice to have different options such as reply quoting and just plain separate quote-posting for conversation flexibility, as each have their own use cases where it depends on the conversation and its form. Sometimes you just want to reply directly to someone, but you also want to boost it while showing your followers exactly what you're commenting on (without having them open the thread) and it's hitting two birds with one stone where you're quote posting and also replying directly to the poster and others can more easily join in on the discussion. But at other times, as you mentioned, you don't necessarily want to engage with the original poster, which is more like cross-posting on Reddit.

Ultimately, it's important to provide users with the tools and flexibility to drive the conversation and discussion as they see fit. People misuse and abuse technology and features all the time, which is more of a behavioural and moderation issue.

Right, what you're proposing I would call "Reply with Quote". I can see that being useful as an additional option, it's just not really a "Quote" in the more journalistic, distanced sense some of us are hoping will be added. (Which is more about forwarding it to a wider, different audience without pushing for interaction with the source.)

it tries to capture the idea of stapling the boosted reply to the other post for context in the home timeline

Just noting for visibility that according to this post (which has a helpful summary and screenshot), that approach is implemented by Elk, an alternative open source Mastodon client in development.

Just noting for visibility that according to this post (which has a helpful summary and screenshot), that approach is implemented by Elk, an alternative open source Mastodon client in development.

Exactly, by modeling the "quote-boost" as a reply it reduces the problem to just enhancing the UI. And you get all the existing benefits from the reply feature already built in (and all future benefits), benefits discussed in prior posts here and more discussed in the thread on the post you've linked.

For example, Thread muting is another huge benefit.

It doesn't though for the other uses of quote where a "reply quote boost" isn't helpful. And especially for those, muting can be a problem: imagine some public person simply banned journalists from quoting a public statement, that would possibly cause quite some chaos in reporting, also see my previous ponderings on that. I still think it should therefore be considered an entirely separate functionality to non-reply quote boosts like Twitter has them where details should be pondered separately. But that's just my personal thoughts, of course.

Edit: I'm also not saying disabling quotes done by the target (not a moderator) is unthinkable, just that it causes trickle down questions. Maybe that clarifies it a little!

Putting on my product owner hat, I thought it a good idea to separate out quote posting features and any settings to control quote posting. Now this issue can narrow scope to focus on quote post implementation. #25735

I want to state a personal stance here: What annoys me personally the most about QRTs (and that's not to say harassment is any less bad) is that over there on Twitter, bigger accounts have quoted me with totally unnecessary, unoriginal comments instead of retweeting and all the likes, RTs and interaction went to them instead of me. I hate that! Big accounts are loving this feature to indirectly increase their reach even more at the expense of smaller accounts' thoughts. We are just cheap content providers for them. Native QRTs are shaping the culture. This is why I hope such a feature in Mastodon will allow me to at least retract QRT allowance on toot and account level even after they have been written. Restricting from the start who’s to QRT like suggested in this issue might also prevent much of the nasty cultural side effects, but not completely.

@frumble — Yours is an interesting concern. To each their own.

In a world where our posts are recirculating via software, I support giving posters a variety of tools to intervene in that circulation. These suggestions, designed to limit harrassment, might also help your concern:

1. Users that are in at least one of the following groups may quote-boost my posts user-controlled, abuse-resistent quote-boosting #20673
2. A slider to disable "Show quote boosts" in your mentions.
3. Options to cap interaction/virality with your posts. Such as a "comments are closed" switch that either can be turned on in the middle of a pile-on, or pre-sets a number of replies/boosts/etc that you are willing to handle. Give people a medium-sized default, but also control. For example: A notification could read "Your post has been QB'ed 50 times. Allow more?" or "This post has received the maximum number of interactions set by the user."
4. Powers to limit mentions… A user might choose to limit mentions to: a) mutuals b) followers / people followed c) their instance d) others accounts with fewer than N followers.