Blocking a user behaves more like mute. Proper blocking needed to prevent harassment.

The simple fact of the matter is, the Fediverse is public. It's a space specifically built on sharing. Finding your posts is trivial, and that's by design. Blocking another account from viewing your posts is an incredibly weak hurdle for someone to overcome, and it informs them that you have blocked them.

If anything, doing so risks the possibility of escalating or accelerating harassment.

All the more reason to rename the function from block to mute so that users don't get a false sense of security.
- A better way. When users need a sense of security, they mute someone and are immediately banned from the Fedverse. This way no one can bother them and no one can harass them again, it's for their own good.
Mastodon still implements privacy and blocking to an extend: https://nerdschalk.com/make-mastodon-account-private/

So it's still possible even within the Fediverse to have more granularity.
- Mastodon's blocking can fuck right off. Very fascist neoliberals there routinely use it to censor anyone with an opposing viewpoint and because being blocked literally locks you out of content people are literally afraid to post anything lest they offend the creator's very fragile egos.
  
  As a result the entire platform is basically you just shouting into the void where people who like you have a huge incentive not to engage. You want to know how how many conversations have had even a response? 2. Two. My entire time there, two times. People either agree with what you say and like it or they don't and block you on sight. Nobody wants to risk being left out so they just keep to themselves.
On other platforms like Twitter and Mastodon the block functionality is pretty helpful. Using block lists improves the experience of those that get most often targeted tremendously.
The simple fact of the matter is, the Fediverse is public. It’s a space specifically built on sharing.

You're thinking about it wrong, a good blocking system doesn't need to hide the content but rather block interaction from the offending user, like a softer form of a ban, but only for that specific user. They can still see all content from the user but they just aren't allowed to interact anymore. Could they bypass it with alt accounts? Yes but they can also bypass bans as well using that same method, so it's not a good argument against something like that.

Currently I can apply this to a user and they can still see all my posts

Yeah, that makes sense. You blocked them, not the other way around.

Yeah, I block a lot of people. Mostly because I don't ever want to get roped in an exchange with them again.

Lots of them follow me around and keep replying, but who cares?

I don't see anything from them, which is what blocking is supposed to do.

It never made sense on Reddit where you'd block someone and still see their comments but they can't see yours.

For better or worse “block” is the ActivityPub terminology for this, and ActivityPub is what the fediverse is built on. If subsections of the fediverse stray from our shared ActivityPub vocabulary, we start talking past each other. (But don’t take my opinion as a proclamation on account of my being an admin.)

Situation 1: Person A annoys person B. Person B "mutes" person A. Person A rants at Person B, gets no reply, and moves on.

Situation 2: Person A annoys person B. Person B "blocks" person A. Person A notices Person B has blocked them, logs on to an alt account, and annoys them from there.

One of these methods defuses harassment, the other has a far higher risk of escalating it.

I'm curious about what makes people think the second method would work better.

Agreed. If anything, if someone gets muted enough times, lemmy should generate a report to the mods and admins for review. That way serial bullys can be blocked from a community or instance, which is far more effective.

Would a two way block be even practically possible on a network like this? Whatever server they are on would see your posts, so all it would take is for that server to use a slightly modified version that doesn't hide your posts to blocked users and they'd see them anyway.

If a server is going so far as to modify their code to better enable harassment, then that is a bad server and should probably be defederated from.
- You have to know that they're doing this though. Suppose some troll is self-hosting, or part of a very small instance? You'd only know they'd do this if they told you
A good two way blocking system should mainly focus on preventing interaction from the blocked user (The specs of Activitypub mention interraction, as opposed to viewing), even if they don't hide it, the interactions from the blocked user wouldn't be federated as if they were a banned user.

It wouldn't work anyway, all they would have to do is log out, or view your user page from any other instance.

I agree. I've seen cases of harassment before when one person would block the other, but then the blocked person would comment on most of the blocker comments to harass them further.

ActivityPub implementations generally don't allow this.

This comment will, when I click 'Reply', be sent to your instance (dormi.zone), that instance should then run it's filter/block checks on it and if it's happy it will forward it onto the lemmy.ml instance for further disemination amongst the subscribers of the group.

If you were to have blocked me then my reply will appear on my instance only (which is admitedly tiny - at 1 user) and go no further. This kind of falls apart if I were to be on a bigger instance as more people would see the reply.

That said, Lemmy may not be doing that quite right as the whole Groups/Communities thing is sort of an extension of the main protocol. I hope it's doing it the right way.
- That's what I thought too, but then I had a look at Sharkey (and I guess Mastodon does the same), and it is possible to have private profile, that needs to accept requests before people are able to follow them.
  
  https://nerdschalk.com/make-mastodon-account-private/
  
  So I guess it's still possible even with ActivityPub?
- Unfortunately Lemmy isn't like that and does not follow activitypub spec in that regard, in their current form the block doesn't seem to do that at all and simply hides the blocked user from the blocking user as if the blocked user didn't exist. There are no checks on interactions.
  
  Also if you're wondering how it works with Mastodon, Lemmy basically ignores Mastodon's blocking system and freely allows interraction with Mastodon accounts in the thread even if they blocked the user replying, and also the community actor.
Yup this happened a lot on Reddit. As much as people complain about the newer two way blocking system on Reddit this type of harassment disappeared basically overnight when that rolled out. It largely was a good thing because for every user who was legitimately being abused by it, there were a lot who were benefiting from it by stopping harassment from others.

I wish blocking communities worked more like mute. There are some I do not want to see on my feed but from time to time may feel like browsing them.

On the other hand blocks/mutes are there for a reason, and I can't think of a good reason outside digital self-harm to remove them.
That's what "subscribed" is for, no?

Honest question. If you can't see them because you blocked them, then how can they harass you?

Create a new account to do so, interact in a way to spur aggression from other users (lying in comments you can't read, being aggressive to supporters of your content, etc.), using the info you post in an attempt to dox you or otherwise stalk you, screenshot everything you post and send it to a discord that's made to harass people like you, an quite a few more.

You'de be surprised how quietly never seeing a person again stops this interaction, but just no longer engaging won't as the person will be constantly reminded of your existance. Simply disengaging while they still see your posts will quite often make then get more violent or direct in order to get a responce fron you. I've seen this before from kiwifarm members.

A real block feature may not stop all harrassment as they can still create a new account, but it's less likely to prompt it's continuation as the other user simply sees a quiet end to your existence.

The way blocking works on Reddit is awful and is frequently abused, so I'd rather not see Lemmy go in that direction.

A person on reddit once cried to me that "blocking is only for when people are harassing you" and I just laughed and blocked them. If reddit's mods/admins would rather ban people calling out nazis than the nazis themselves, than you're well in your right to so for yourself.

I agree with OP that block should be labelled mute, and agree with others in this thread that terminology should be consistent across ActivityPub applications and that the current Lemmy behaviour is the best behaviour for a block/mute button given Lemmy's public nature.

Could Lemmy (if user consensus agrees with it) raise the topic to the ActivityPub developers?

if you post things on a public website what do you expect ?

I think the bigger thing is, blocked people can even still respond to your comments and posts, right?

Now Lemmy can implement anything but nothing could ever prevent blocked/muted user to create another account in order to continue harassment.

But it's not specific to Lemmy and same with anything open on internet.

I think the only way to prevent such issue would be a system which would require to prove identity in some way in order to create a single account. But this is completely against the openness of a federated network.

Now Lemmy can implement anything but nothing could ever prevent blocked/muted user to create another account in order to continue harassment.

Not a great argument because the same could also apply to community and site bans.

I think that having more tools to fight harassment is ultimately a good thing, are these tools perfect? Of course not, but they are still better than having nothing.

I think the only way to prevent such issue would be a system which would require to prove identity in some way in order to create a single account. But this is completely against the openness of a federated network.

Indeed it is, plus it doesn't stop those malicious enough to commit a felony just to harass someone but that is neither here nor there, this discussion is about protective measures that can be done before ban evasion.
- Indeed, I was focusing mainly in the fact that there is not a easy solution to completely avoid this problem. Now indeed being able to mute, to ban or to "report" an user to his instance admin could be useful tools even if they are not the full solution.

I don't think that blocking people from seeing posts really makes sense since they can just log out to see them, it's all public. So I don't really feel like hiding content from them is very useful since they can view it without an account. What is really needed is a way to restrict users from interacting, I made an issue addressing this it would be treated as a kind of profile ban, similar to community bans but for individual users but even when banned people can still view and read, like they always could before.

Definitely want a hard block like other platforms. The fact that 'block' is treated as 'mute' is really damaging in terms of enabling harassment.

I might be missing something here but cant you make you posts followers only? - Am I about to have a learning moment?

There is no such thing as following users on Lemmy
- Gottya. Yes I was thinking Mastodon. i guess Lemmy is by its nature public. I guess you would need to be able to address your posts to either public, users who are community members or become one, only current community members or potentially either of those two but excluding blocked users.
  
  I feel like their are some common group communication design patterns that could be adopted that might make working through these options simpler - e.g public vs private groups, broadcast vs multicast communication, forward vs backward accessibility. But perhaps I just cant sleep and ita 330am...