So far Lemmy is vibing. Everyone here is excited and optimistic and willing to put up with a few rough spots to be part of something.
When the Eternal September comes, which it will, how does a Lemmy instance deal with bad actors?
I think it’s important to enable account portability across instances, like what Mastodon has. It should be easy for people to move to a different community, back up their data so they can re-substantiate their known persona if their instance goes poof, etc. This will help a lot with encouraging people into communities that suit them and with people who might stay in a community they are unhappy with because they don’t want to start over.
Strongly agree. This could also be good if you join an instance and it winds up being toxic or not vibing with your beliefs.
I was thinking about this, actually. Wouldn’t it be better to have users-only instances and content-only instances? That way you can have an instance with a policy towards certain subjects (e.g.: bigotry, racism, sex openness), but you chose the content you want. Just like if it were a cable or streaming service. You choose the content you want.
BTW, is there a place to discuss this? How to improve Lemmy and next steps? Also as a way to know how to contribute.
Self-hosting might be the only way to do this, I imagine any instance with enough users will have people wanting to post locally
definitely. account migration and maybe community migration (unsure how that’d work exactly) would be great. losing history every time an instance shuts down isn’t very fun
Some thoughts —
The original “Eternal September” (on Usenet) wasn’t an influx of abusers. It was an influx of new users who didn’t know how to do things properly yet.
Most of the new users were from the America Online (AOL) private service, and known as “AOLers”. (As it happens, I joined Usenet around the same time, but from a local dial-up Unix BBS in the Washington DC area.)
The AOLers didn’t know which aspects of the service as they saw it were due to the AOL custom client software, which were due to the AOL local server, which were due to the newsgroup (forum) they were looking at, and which were due to the global Usenet consensus. So when they had a problem, they didn’t know where to address that problem. They complained on public newsgroups about UI issues with their local client, because they didn’t know what was what.
And the existing users didn’t have the time or capacity to help them. The AOLers were added to Usenet en-masse without preparation. Nobody had signed up to help them. The AOLers were accustomed to AOL chat rooms that had staff helpers and moderators; most of Usenet did not have any — just regularly-posted FAQ documents, which the AOLers did not know to look for, and grouchy users who angrily told them to read the goddamn FAQ before posting.
Another consequence of the influx of new folks was that Usenet suddenly just had a lot more people. This made it a tasty target for commercial spammers and other abusers; which led to the eventual spampocalypse and a lot of people abandoning Usenet for web forums or other services.
It wasn’t long into Eternal September that the hardcore abusers showed up, though. That, I think, is the harder problem to deal with.
“Good” Usenet servers did not reliably disconnect themselves from the servers that were accepting and forwarding spam. It was not generally acknowledged that a good server needs to block bad servers: the free-speech ideal was assumed to mean “accept anything from anyone; let the client decide what to filter out” — which meant that new users who had not written any filters necessarily saw all the spam.
And because nothing was secured by strong encryption, forgery was rampant; with a little cleverness, anyone could pretend to be anyone from any server.
There were many, many efforts to fix the spam problem. Unfortunately, as things turned out, it wasn’t enough. Eventually folks noticed that the NNTP facility offered by their ISPs was a great means for sharing pirated porn …
I’ve just gotta know was that local dial up in DC Digex?
I worked with Tale@UUNET during the Eternal September, providing NNTP support to our customers. God that was hell.
I’ve just gotta know was that local dial up in DC Digex?
CapAccess.
Agreed on all points! It turns out Lemmy has a mechanism for federating block lists. What will be interesting is when instances disagree about bans. If you get banned from an instance because - hypothetically - you disagree with the actions of one government or another, it’s not obvious to me that other instances should repeat the ban.
Will we end up with islands of trust?
Yes, as we always do, digital systems should represent the real world, not be a distortion of it. Protocols are meant to standardize communication but the rights to re-distribution have never been guaranteed . Now many understand why this may not even be feasible in a real way.
There will never be just “one zone” and there shouldn’t be, however control over your interaction with these zones should be up to you not brokered by a proxy. To a degree we do this out of necessity though IMO the larger goal would be to give the user the ultimate option even if deployed infra is helping make it happen.
Yes, as we always do, digital systems should represent the real world, not be a distortion of it.
It’s OK for online systems to represent a projection of the real world. Not every feature of the real world needs to be represented in every online system.
It’s OK for the furries to have their server where everyone pretends to be tigers and dragons.
its also ok for them to go to private residences and dress the part, im usually speaking of data, trust and execution realms. These need to represent the real world since things like giving up your ownership of your data and systems should not be a requirement to use a novel app. This is not how the internet was intended to operate and in the days of 6ghz silicon and ultrafast dram the cryptographic overhead of doing things in a way where you own your digital domain in the same way you might own a house is very real.
Where you want the technology to not represent the real world is in its abilities to scale, and that’s what’s really crazy where were are with technology today individuals can be companies and small teams are international orgs. This is not just a concept for entrepreneurs but a concept for anyone who wants to take more control over thier presence.
It’s still okay for people who don’t dress the part to pretend to be tigers and dragons online.
thats not what i mean either, just like in the real world you can wear masks and costumes or not. you can even wear masks that arent obvious simply pretending to be entirely different people. what else are you looking for, hit me with it.
The same Elder Internet that spawned Usenet also spawned furries, which seem to have become a standard test case for “so just how tolerant is your community?”
this is worthy of a BestOf, do we have a BestOf?
We do now. https://lemmy.world/c/bestof
I’ve been on reddit long enough that I remember the mantra…
Do not talk about Reddit on other sites
Do not link to Reddit from other sitesThey understood the concept of “Eternal September” and wanted to hold it off for as long as possible.
It kinda worked for a very long time. Like a good 8-10 years. Sure, there was a slow decline but reddit was still pretty good up until new reddit was introduced.
I remember being embarrassed to discuss reddit irl in a way that I wasn’t embarrassed to discuss Facebook, for example. Reddit was the dirty little secret.
Downvote they ass
on your instance you shud ban but
haha evil wins again!
Jokes aside, most offensive posts mostly originates from different instances with vastly different user culture. Downvoting posts works in the way that it lowers the visibility on your server but the offending poster might be on an instance that disregards downvotes so they “won’t get the message”.
It’s much more effective to just block the poster, or the whole community if one so desires.
If a server admin turns out to be a giant asshole (present company excepted, of course), is there a way to migrate your identity to another instance?
If a server admin gets hit by a bus and their instance goes away, do all the users just cease to exist?
It’s a planned feature if I’m not mistaken.
Mastodon has that feature, but Lemmy has not added that feature yet. From a technical perspective, I don’t think there’s anything preventing it, the developers just need to code it. I’m sure they have their hands full dealing with the reddit explosion right now though.
My understanding, based on what I’ve seen with Mastodon, is that, yes, all users will just cease to exist if an instance admin decides to pull the plug. There was some stupid drama with a particular Mastodon admin for a really popular instance a while ago (I forget which server exactly), and they decided to just kill the server. Poof, 100k+ users gone
The potential for accounts to vanish if the instance they started on is, to me, the single biggest hurdle that Lemmy will face with casual users. I think that the devs need to really consider figuring out a way to make user logins global.
I said this the other day, but I think it may, unironically, be one of the first times I’ve ever seen a genuine use for a blockchain, but I have no idea how to implement it.
The reason that the big social media companies came to exist is precisely because people didn’t like having to have a dozen accounts for all their different communities. Lemmy fixes that problem through federation, which is great, but introduces a new problem of “your account could just disappear, making all your contributions vanish.” I know that was technically a problem before big social media companies appeared and everyone was using forums, but it’s a big plus of the current social media giants- you don’t have to worry too much about the company failing so completely that the website gets shut down, which is the only way you’d lose your account, any time soon. People are used to that stability, and will not be happy if they join an instance in the fediverse only to have the rug yanked out from under them.
If we want this to be a true alternative to big social media, it needs that stability.
The other consideration is that impersonation might be pretty possible by making your own server called lemmy.mi or something and then stealing peoples username’s verbatim. IDK if that’ll ever become an issue but I do think its an avenue of attack for bad actors.
Yeah, it’s basically like email. Though I imagine an instance like that would get defedded pretty quick
Oh it’ll definitely become an issue - Help help my local community! A calamity has befallen me and I need cash now! - Posted by @[email protected]
that’s…really unfortunate
It’s part of the reason I chose to host my own rather than depend on another server somewhere. That way when I do fuck it up at least the only person to blame is me
Yay federation and activitypub!
Smart choice, do you have it set to private? The only thing I’d worry about is people trying to join my server and bogging down my internet lol
I don’t have it set to private because when I tried that before it seems to break federation entirely. I do however have to approve anyone who wants to join. At this point I’d probably allow my close friends to join if they wanted, but that’s about it.
Mostly because I am nearly 100% positive I will either lose my ZFS array, try to move the server to different hardware and bork psql, or what have you…
My homelab is mostly duck tape and bubblegum.
I’m gonna set mine up to where everyone has to be approved and approve nobody since it’ll be running on a Raspberry Pi
I’m really interested in running my personal instance. Can you point me to a tutorial on how to go about that?
Here is the official documentation: https://join-lemmy.org/docs/en/administration/administration.html
Pardon my ignorance, I’ve only just started to figure out federated sites (I think, probably not though), what’s activitypub?
It’s the protocol all the different federated services, not just Lemmy, communicate over.
Why do people care about preserving their “identity” and posts so much? This was never a thing in the old internet.
The old internet didn’t have an all encompassing issue with bots and bad actors trying to gain your trust, a public post history is basically the closest thing a person can have to a trustable identity online, it’s not a perfect solution but it helps
I am not sure I follow. I don’t see where trust comes in when you’re just reading random people’s posts. I guess if you wanted to do moderation or something. But I know a lot of people including myself purposely delete their reddit account and start over.
Hey this person is talking about this subject I have just heard of. I will at some point need to go validate their information but as a shortcut I can go look at their profile and see that they are well respected in communities dedicated to that subject. Therefore I can trust their information.
Alt
This person is asking questions that sound reasonable on the surface - but when I look at their post history I see they are active in some much more extreme communities and I’m able to form the conclusion that their apparently reasonable post may not be in good faith.
if I’m unable to detect the tone or intentions of a comment I’ll check that user’s posts to get an idea, if someone has a history of not being an asshole I’m much more likely to give them the benefit of the doubt or want to engage with them. it also helps ID spam accounts
Because social media exists. There is identity attached to your online presence for the vast majority of people.
Really? I had my reddit account for 10 years, I dont think a single person remembers/recognizes my “identity”. With smaller communities people actually knew eachother. Your name actually meant something.
I think what your_mind_aches is saying is that the mindset has changed. People who didn’t know the internet before social media are more emotionally attached to having one single identity online. Even if in the case of reddit it’s not necessarily linked to your real world identity.
Sometimes when I am unsure about a post/comment I click on the user profile and if I see 10yrs / 100K karma, it helps forms my opinion and trust of the user.
I’m wondering the same, I would guess grey listing and shadow banning to be the most effective.
Shadow banning is Orwellian . something you might expect from the CCP instead of a supposedly progressive online forum. If you’re going to ban someone at least have the decency to let them know they are banned.
China bad
All these American tech firms shadow ban people. The practice is very abusive, but it is also very American.
Woah, this is the first time I’ve pressed “all” on lemmygrad. It’s… so much bigger!
Lol yeah!. Default should be “all” imo. Also, the default sort would be “hot”.
I mean most of the time I’d still prefer to stay in local. It’s mindblowing how many libs got through.
I kind of wanted to downvote you, but I suppose ecochambers are kind of a feature in Lemmy? I’ll have to wrap my head around that.
Like reddit, it’s by design. That’s the price we pay for participating in a consensus-driven frontpage aggregator that’s divided by interests/politics/ideologies etc…
My stance is leave them alone to talk shit about me so I’m left alone to talk shit about them. Block and move on if I find the person disruptive, report them if they break server rules. And then block them and move on.
In total I think I’ve blocked more people than I’ve downvoted on Lemmy.
I hardly ever used downvotes on reddit and will probably use them even less on lemmy, though I’m still very glad to have it if I need it.
Even the most principled Lemmitor has a special downvote for that special asshole. I’ve heard.
When the Eternal September comes, which it will, how does a Lemmy instance deal with bad actors?
i’ll bully them away >:3 !!!
On the real I feel like Lemmy/the wider linkagg fediverse will prob be good at self-moderating somewhat like other fediverse software’s communities are. It’ll probably be easier for admins to noice bad actors on their instance than it was for site admins on Reddit to notice bad actors there because the admins-to-users ratio on here will probably be better, even if things are kinda concentrated on lemmy.ml, lemmy.world and beehaw right now (people will probably spread out as they get a grip on how things work), and the average user will probably grow a stronger connection with their instance admins for that reason too, making it easier to address things like that since more people will be able to comfortably contact their admins directly. And if said bad actor is from another instance, and the admins of that instance refuse to deal with them, there’s always community-level bans (I think anyways? I’m still not familiar with the comm mod tools) and, if more drastic measures are needed, defederation.
Ban them. Honestly if it’s egregious the admin staff takes care of it. If it’s just some asshattery then the mods of the communities are left to deal with it.
Individual instances will have to moderate themselves. If they become chaotic, other instances should unfederate them. But as users, you should also subscribe to communities you think are behaving well and block users/communities that are not.
Also, I have seen some users who are “grabbing” as many communities as possible, namely @[email protected]. Dude is moderating 60 communities, in an instance that started a few days ago… He is not building the communities, he is just power tripping it seems. @ruud@[email protected], something might have to be done about that in the future. I suggest some sort of “requestcommunity”, in which you can apply to become the mod of said community, if community is being badly run (or not run at all).
There should be a limit on how many communities you can create in a given time span
There should be a limit on how many communities you can run, period. This is how we got super-mods like GallowBoob on Reddit
I upvoted this post and I saw a popup “report created” this is not what I intended, I completely agree with this.
I’ve seen spontaneous “report created” moles¹ as well. It’s not clear to me that a report is actually being created; it seems like a UI bug.
¹ “mole” : a pop-up div that appears from the bottom of the page.
Mole? What’s the difference between a mole and a toast?
I’m not sure. I’m not a frontend dev myself; “mole” was the term I heard from people who were.
Oh. I think they are most commonly called “toasts”. Called such, because they pop up from the bottom like toast from a toaster. I’ve also heard them just referred to as “alerts” or “notifications”, but I think that’s a bit ambiguous. Android likes to call interactive ones “snack bars”, which is kind of silly. “Moles” is new to me as a term for them, but I think it’s quite fitting too, yeah, I like it
And how does that stop them creating multiple accounts to multiply the limit? It doesn’t.
this. Ihateany “verification” or “request” process as somebody has to do it. But saying that you cant create more than x communities per month or y communities per yearwould pretty much solve the problem.
Unfederation should not be used so cavalierly. Instead, community blocks. I know many people that chose lemmy.world because it doesnt block anything and hope it stays that way.
What about the ability for users to block instances? I’m sure there’s a way I’m not thinking of that this would be problematic.
Seems best to me
We’ll live, we’ll see. Meta is showing its interest in mastodon, so we have a reason to worry. But I think, lemmy will change according to the situation, when situation will be present, not before it.
there are multiple ways evil can behave on lemmy:
trolling
- trolling, it is annoying, if 25% of all posts are troll posts, the site can be annoying to use.
- content voting systems can mitigate this tho, but bots will eventually find a way to game this?
- the difference between trolling and spamming (imo): trolls type in their message with a physical keyboard. Spammers use bots to automate trolling
(Bot) Spamming / automated troll farms
- spamming, creates huge load on storage capacity of the server owner, not good if you host for
free- spam can be hard to detect in the age of chatgpt LLMs in general, because normal spam would be detected by how random it is. for example
adfjakjdfkl would be easily detected as spam
- spamming huge amounts of text is still better than spammers creating huge amounts of video and photographs
- proof of work algorithms can mitigate this issue somewhat, tho this also makes performance worse for everyone
any other thoughts on proof of work, or how evil doers can behave on social media sides?
- trolling, it is annoying, if 25% of all posts are troll posts, the site can be annoying to use.
thine asses shalt be downvoteth
I’m not worried about assholes. I’m more interested in being free. As long as the community mods are nice enough, I’m optimistic.
Live free or fucking die
Lol at creating a new open source platform with free speech and immediately asking how to eliminate it.
In the older, better days of the internet, “assholes” were just a part of it.
Learn to deal
In the older, better days, we used kill files, and our choice of platform was eventually overrun by spam.
No, assholes need to learn to behave within the context of the instance / community they are posting to or get downvoted / moderated. Ideally they go away to their own instance where they can be assholes to each other and be defederated.
How long until we start seeing tiktok/instagram/facebook/reddit reposts.
Then we’ll know we’ve truly “made it”.