I do not want to block all bots. I only want to block bots from specific instance. More specifically, the @alien.top instance is using most, if not all, bot accounts with random usernames. It uses that instance to post in communities of other instances. I thought about blocking other instances. But the main issue lies with random bot accounts from @alien.top. And I cannot block individually by usernames.
What shall I do to block posts posted by random bot accounts of @alien.top instance?
Update
I looked up potentially affiliated sites connected to that instance and found these.
https://level-up.zone/ (@communick, high post rate)
https://gearhead.town/?dataType=Post&page=1&sort=New (same admin, high post rate)
https://viewfinder.pro/?dataType=Post&page=1&sort=New (same admin, medium post rate)
https://netheads.online/?dataType=Post&page=1&sort=New (same admin, low post rate)
https://healthy.community/?dataType=Post&page=1&sort=New (same admin, low post rate)
https://sfw.community/?dataType=Post&listingType=Local&page=1&sort=New (same admin, no post yet)
https://blockchained.world/?dataType=Post&page=1&sort=New (same admin, low post rate)
https://matchpoint.zone/?dataType=Post&page=1&sort=New (same admin, low post rate)
https://poweruser.forum/?dataType=Post&page=1&sort=New (same admin, high post rate)
https://hi-fi.community/?dataType=Post&page=1&sort=New (same admin, medium post rate)
https://expats.zone/?dataType=Post&page=1&sort=New (same admin, medium post rate)
https://nba.space/?dataType=Post&page=1&sort=New (same admin, high post rate)
https://athletic.center/?dataType=Post&listingType=Local&page=1&sort=New (same admin, no post yet)
https://style.land/?dataType=Post&page=1&sort=New (same admin, high post rate)
https://metacritics.zone/?dataType=Post&page=1&sort=New (same admin, high post rate)
https://academy.garden/?dataType=Post&page=1&sort=New (same admin, medium post rate)
https://foodie.rehab/?dataType=Post&page=1&sort=New (same admin, no post yet)
https://indiehackers.space/?dataType=Post&page=1&sort=New (same admin, high post rate)
https://hardware.watch/?dataType=Post&page=1&sort=New (same admin, high post rate)
https://nfl.community/?dataType=Post&page=1&sort=New (same admin, high post rate)
The instances may be hours to two months old and are run by the same admin. Some have high post rate, while others have no post yet. The ones with posts are all using @alien.top’s random bot accounts. I suspect this problem will amplify.
Update2
Blocking by instance is finally working. I blocked @alien.top and @hardware.watch. Consider this issue resolved.
It kind of changes things if accounts are automatically made and the user on reddit has no idea. From earlier discussion I thought you were saying these posts were from people who explicitly went to alien.top and signed up. But it turns out no, the content is automatically duplicated and the user on reddit has no idea this is happening - so it IS largely a waste of time to try to respond to them as they’re unlikely to ever read it.
The suggestion I’ve seen you make elsewhere for Lemmy users to message people on reddit and invite them to Lemmy is rather unrealistic, and I’m pretty sure reddit would tolerate this being automated for about 1 day after they found out about it. Reddit isn’t going to let you scrape the site and autospam members (in their view) to support what’s essentially an unauthorized 3rd party client that replaces the core function of their site.
If the content being posted had a real person behind it from the beginning, then it would be one less reason to treat the whole thing as “spam”, wouldn’t it?
It should be simple to discern: users from alien.top that have not claimed their account there are marked as bots. Once an user from reddit claims their account, they are then unflagged as bots and none of their content gets automatically mirrored (because once they claim their account, they control the password and there is no way for my service to post on their behalf)
I am not even saying it has to be automated. At the moment, I’m actually suggesting people to go do what I am doing and writing DMs on reddit to tell the original poster about the mirrored link and to let them know that they can get more answers on Lemmy.
I know, I addressed the manual messaging by saying
One problem is the communities aren’t very engaging for people on Lemmy since it’s this huge list of posts with no responses and no guarantee the original poster will ever see your helpful response.
So… it’s not automatically mirrored after they sign up on Lemmy? I thought that was a key feature.
Why? During the summer, at the height of the protest, I was taking 15 minutes every day and sending 10-20 DMs to people on the emacs community. For better or worse, I think around 40 people responded positively.
Why couldn’t a fraction of us do something similar? Specially that now we all could leverage the fact that fediverser provides a really easy way to migrate?
The key feature is the login with the automatic subscription to the corresponding communities, not the automated mirroring of the posts.
okay, I guess I see what problems you’re trying to solve, but I don’t see how it really does them that well.
These communities on Lemmy are unlikely to take off since they’re not building true engagement. Signing up to Lemmy with my reddit credentials and automatically being enrolled in a bunch of Lemmy communities that are not really active isn’t really better than just signing up to Lemmy normally, which is not very difficult. I wouldn’t post a question in any of these bot-filled communities on Lemmy because it’s unlikely anyone would see it as it would be covered up by 50 automatic posts from reddit before anyone even saw it.
I don’t see the lemmy content as spam, just that like other reddit-autoposters, it’s a strange form of a 3rd party reddit client. I do browse by All>New because that’s how I can discover new communities and keep my feed fresh, given Lemmy’s limited activity at the moment. However, pretty sure that reddit and many members would see messaging people unsolicited and promoting another site as spam.
I am tracking some data, and I can tell you there are “fediversed” communities (the ones with mirrored content) growing in organic activity even when Lemmy usage is overall dropping.
I know, I wish people could take 3 minutes of their time to acquire some minimal tech literacy, but the truth is that the majority of social media users simply don’t care enough about it to make the switch.
There is already some mechanisms on fediverser software to add rate-limiting of mirrored posts. Meaning that I could set it up to post at most N posts per hour (or per day). But be honest: is there a number that is low enough for you to think that posting is worthwhile, or maybe it’s just you not feeling comfortable to take initiative and post in a place where you haven’t “seen” other peers?
And there will be just as many (if not more) people that are willing to ditch reddit, but are not aware of any other alternative.
Marketing should not be such an ugly word. If we really believe that the Fediverse is a better and healthier alternative to the existing social media landscape, why can’t we do a little bit of evangelizing?
I pondered this and I think it would work great for communities where OP is not asking a question and the comments don’t matter - the SFW pic network, for instance. People would probably really appreciate seeing the content from subs like that without having to go to reddit. There are also so many of them that even going to reddit, I miss a lot of the photos the first time around. But the subs where OP is expecting tech help are perhaps the least appropriate for this current system and I think that mismatch is the main reason people on Lemmy had a problem with this. That, and might want to rethink the auto-account creation. Posting under one bot account would make it look less strange to people on Lemmy, and each post could have a byline explaining that it was mirrored from a specific reddit post. If that OP wanted to sign up for Lemmy via alien.top they still could, and to me it would seem less offputting than a mysterious account already having been made with my username.
With the concerns about storage and other system resource usage from federated instances, though, rate limiting would definitely be required. One post every couple hours would be okay. Some of the instances I saw when the mirror-bots were active had a post every 15 seconds or so.
Why would I post a question in a community that is 99% bots? It’s unlikely anybody on Lemmy would see it in the flood of posts mirrored from reddit. I’d be better off asking in a higher signal-to-noise ratio community on Lemmy or if I was desperate, just reddit. The excessive posts and unexplained nature that led people to conclude those instances were spam led many people to block them, which makes it even more unlikely someone on Lemmy would see my post there.
Marketing is great, but unsolicited DMs is NOT the way to do that. I’d suggest finding some other way of spreading the word.
Agree about SFW. So much so that I created another instance for it and was finishing setting it up: https://sfw.community
Disagree about separate bots for replicating content and also disagree about not having comments. This is exactly the approach taken by the lemmit.online dev, and he was facing pretty much the same backslash as me.
About “resource usage from federated instances”: I could argue that the real issue is caused by those people who are running scripts to subscribe to all communities (like https://github.com/wescode/lemmy_migrate). It’s key to understand that content is only copied between instances if there is one subscriber to that community. Also, it’s not like I am running a massive computing cluster or that alien.top is throwing terabytes of data every day. If the network cannot deal with the load that alien.top is generating, then it’s a sign that we are not ready for more people coming from Reddit. And if we are really that not unprepared for it, then I’d rather drop the whole project and go back to the drawing board.
Have you tried, or are you projecting your feelings towards it?
Cold call sales still exists. Direct Mailing still exists. If companies still put a lot of money to do these things, it means that it pays off.
If you have any other idea, I’m all for trying. But if they are not at least as effective as direct contact, I won’t be dropping the idea of sending DMs.
okay, that’s all cool.
I’m speaking from my experience with unsolicited messages and marketing, as well as the general impression people get from that. Direct mail and cold call sales work sometimes, but they’re often not welcome and reflect poorly on the organization doing it. I feel like doing it excessively would be likely to result in reddit suspending accounts or at least restricting message ability, plus, it would mean that for some people their first exposure to Lemmy is a message that many would see as spam (since unsolicited mail inviting you to do or buy something is the definition of spam).