Thanks for sharing the work-in-progress. on quote posts It's great to see you're approaching it in such a thoughtful way, and I especially appreciate the work @Claire has done both researching and writing up the background document.

My biggest question at this point is what work you're doing to get perspectives on the design from Black users -- and from former Black users and Black non-users who have cited the lack of any equivalent to quote tweets as contributing to the whiteness of Mastodon. Of course these aren't the only perspectives that need to be listened to, but given quote tweets' importance in Black digital practices -- and the racialized nature of so much quote boost discourse here over the years -- they're particularly important.

For example, common use cases for quote tweets include calling out racism., calling people's attention to something worth seeing and highlighting why they're important, addressing a different audience (or changing the subject while still keeping) a link to the previous conversation, and Putting the quote before the comment potentially interferes with these -- centering the racism, not the call-out, etc etc. How much does this matter relative to other concerns? Opinions may well differ, so please make sure to get Black perspectives here!

@MastodonEngineering @andypiper @renchap

And a couple of more specific questions -- apologies if these were in the docs somewhere and I just missed them.

how can people dho don't want to see quote posts filter them? One option is to reusing the existing "hide boosts" mechanisms (for the timeline as a whole and for individuals); another option is complementary "hide quote posts".

Bluesky treats any link to a post as a quote. Are you planning on a similar approach?

I dont have opinions either way -- there are pluses and minuses and implementation complexities -- but they're worth including in the spec and getting feedback on.

Also, if it's useful, here's the threat modeling work I did on this in early 2023.
https://privacy.thenexus.today/social-threat-modeling-and-quote-boosts/

And, I listed various postive use cases for quoting in the section on "Does quoting "inevitably" add toxicity?" in https://privacy.thenexus.today/black-twitter-quoting-and-white-toxicity-on-mastodon/ , so that also points to feedback to get on the design. It wouldn't surprise me if it turns out that comment before vs. after quote posts are useful in different situations for different people, and so it'll be useful to provide flexibility both to the person doing the quoting and the persion viewing the quotes. But, that's just a guess on my part, getting diverse perspectives on it is the key!

@Claire @MastodonEngineering @andypiper @renchap

@jdp23 Regarding the link => quote question, I think currently we plan to ask the user when they paste a link so they can transform it as a quote. There are a lot of complex interactions here and being really decentralised makes some things much harder. We do not want to provide a safe sense of safety where we can, as other AP software might not implement things the same way (or not implement some controls at all).

@andypiper

Thanks for the reply @renchap! Agreed that there are a lot of complexities here, and that decentralization makes some things much harder. Glad to hear that the user-facing part isn't set in stone yet ... I misunderstood the comments in the spec doc to mean that the decision on the quoted text before comment had been made. FYI @laurenshof

In terms of the link-as-quote, that's how Sharkey does it today (so presumably also how Misskey does it?). It's a good point that how no matter what Mastodon does or doesn't do, other AP software is likely to support links that aren't quotes.

And agreed about not providing a false sense of safety ... just like any other functionality), even though quotes usually aren't used for abuse, they certainly can be. And, determined attackers can route around them. Still, speed bumps are useful, both for reducing low-effort drive-by attacks and the chances of quoting unintentionally leading to dogpiling. So, an interesting design challenge!

@andypiper

@jdp23 Yes, due to the diversity of ActivityPub implementations, it will be hard (or impossible) to 100% control what is happening. And you can always link to a status, or screenshot it. But any speed bump adds friction, and friction is an effective tool against most harassment.

@laurenshof @andypiper

@renchap

Yeah I understand how it is impossible for any AP software to 100% control what's happening. But I think focusing on people using links/screenshots to sidestep limitations people set is a bit of a red herring: these are wellknown patterns, commonly used on other platforms as well.

1/2

@jdp23 @andypiper

@renchap

What concerns is me the potential situation where a Mastodon user sets a post to dontquote, and another user on another fedi software that has a different qp implementation that doesnt interoperate with dontquote rule then still quotes the post. More specifically, what will that social dynamic be like, and who will users blame for who's at fault here

@jdp23 @andypiper

@laurenshof the FEP mandates that you implement and respect the controls if you support quotes according to the FEP. If you write invalid software with potential malicious intent, I suspect you will quickly end up on many blocklist (like it already happens).
If you don’t support the FEP, then Mastodon won’t display it as a quote post, for example we don’t plan to support the current quote post « implementations » (RE: <url>) and such
@jdp23 @andypiper

@renchap @jdp23 @andypiper

this is not about software with malicious intent, this is about platforms like misskey that have had quote posts 5 year before mastodon had

People have been quote posting Mastodon posts using *key platforms for a long while now, this is a settled pattern, I did this regularly when using firefish. But now Mastodon is introducing a 'dontquoteme' feature thats not supported by *key.

1/2

@renchap

How are people on *key supposed to know that a Mastodon post is not supposed to be quoted? That fully depends on the willingness of *key devs to implement your other noncompatible implementation of quoteposts. And in order to get them to do it you need to do federated diplomacy.

That's why I'm concerned with Mastodons declarative stance of 'this is how we are going to bring quote posts to the entire fediverse', because it makes the diplomacy harder

@jdp23 @andypiper

It's tricky. I agree on the importance of federated diplomacy, and certainly hope that Mastodon is working with other platforms. But on the other hand, today's fediverse wasn't designd with safety in mind, and to move forward there are going to have to be incompatibilities.

It's like GoToSocial's interaction controls ... they're not a priority for other projects (including Mastodon), I'm not sure whether the FEP has stabilized, and challenging for to implement, so until that happens they're likely to lead to unexpected behavior which will result in some conflict. But they''e still a good thing!

For that matter it's similar to post visibility and blocking when Mastodon first introduced it in 2016/7. And sure enough it led to huge fights with GnuSocial and other impletations that *didn't* support post visibility -- if I recall correctly, Robek even talks about that in the post about tensions between Masodon and other platforms. But guess what, people wanted that functionality, and it' helped lead to Mastodon risking to dominance.

Of course it's not an exact analogy: Mastodon now has much broader usage than these other platforms, so needs to be careful about being seen as the big elephant in the room throwing its weight around. And the current team, through no fault of their own, also inherits the mistrust from Eugen's "Mastodon-first" attitude. So it's complex. I haven't been tracking what's been going on in terms of the FEP discussions, if it's really Mastodon making a unilateral declarative stance then I agree with Laurens that it's not a helpful way to go about it.

But on the other hand, I'm currently seeing a post by Claire on socialhub described as a "pre-FEP." So that doesn't seem like a unilateral declaration to me. So maybe one of the issues here is the tone of the blog post coming across as more absolutist then the reality. But then again maybe the tone of blog post actively reflects the reality! (1/2)

@laurenshof
@renchap @andypiper

@jdp23 @laurenshof @renchap re blog post, I think given the socialhub discussion and this thread, I've learned that no matter how mindful I thought we had been, there is still more work to do to clarify and demonstrate that we're listening and learning (and working with others).

@andypiper @jdp23 @renchap

I think its not so much a matter of being as careful and mindful as possible (while thats obviously a great thing to do), but that fedi is lacking the political layer for diplomacy between different fediverse platforms. 'Listening and Learning' is a framing between a platform and its end-users. But what I think is going wrong here is not seeing other platforms as equal partners that need to be negotiated with

@andypiper @jdp23 @renchap

this is not only a Mastodon problem, but a wider fediverse problem. We only have FEPs as a tool for this, but those are technical specifications that are being mangled into a diplomacy framework. Which, not surprisingly, is kind of difficult to do well

@laurenshof This is indeed a big "issue" (is it really an issue? idk) as there is no formal diplomacy process. We (Mastodon) do not plan to solve this issue for everyone, but we have established some (closed) places to have discussions with other implementors. Our FEP draft was shared & commented on those places for several weeks/months before being made public.

@andypiper @jdp23