I’m also thinking of things from a moderator/instance owner point of view where trying to defend your instance from illegal activity, spam, and harmful content is greatly increased by having open sign ups. If the majority of engagement on your posts are from pussy in bio accounts then that is a put off from the end user as well.
I don’t particularly agree that it’s “greatly increased” by it. I’m on a server with them and I don’t see that much many more spammers. It seems like fearful hyperbole, I’m not saying there aren’t any I’ve certainly seen a few but it’s not as common as most people think it is.
Your perspective has a lot of valid points as well, I think that there is a compromise somewhere that is slightly more effective than a simple captcha that could also streamline onboarding and validating users. I don’t know if it’s as easy as you suggest to ban spammers automatically but I’m all ears for hearing solutions. I think the fediverse is especially equipped for cooperation that could make that kind of automoderation more successful than on legacy socials
Proof of Work Captchas are very effective when it comes to stopping or slowing down bots, while not impacting regular users much if at all.
As for banning spammers automatically it is very easy to set up scripts that ban people for rapid posting, malicious links (bio, posts, comments, DMs), as well as words or phrases associated with spam (buy my Tshirt, check out my OF, see my pussy in my bio). For illegal content, image hashes can be used to automatically purge images and data that is known from the server, while simultaneously banning the people doing it. Rate limiting of suspicious IPs are also another way.
There’s probably some more ones that can be done, as well as ways to improve these methods, point is they aren’t new, and they improve the situation for new users who can sign up easily and instantly without worry of rejection or too high expectation.
The big thing about these (except for captchas and PoW) is that they are post-signup methods that catch the people doing it in the act. It might seem like a loss for some since you don’t stop them from signing up at all but you need to remember. Spammers don’t give up, they don’t stop, they don’t care, real normie users do though. You can increase the bar and make them seem to disappear, but then you also lose the flow of real users, as most of them are filtered out or quit. Detecting and purging them automatically is the best way to combat them, captcha and PoW stop people using scripts for batch signup but banning them automatically when they start. If done right, spammers would be auto-banned after a few seconds of spamming or account setup depending which flags they trip first.
I don’t particularly agree that it’s “greatly increased” by it. I’m on a server with them and I don’t see that much many more spammers. It seems like fearful hyperbole, I’m not saying there aren’t any I’ve certainly seen a few but it’s not as common as most people think it is.
My mastodon instance rarely ever sees spam despite having open signups either. Maybe my problem is I’m worried about future problems rather than the current state of how things are. My concern comes from personal experience with legacy socials like facebook and twitter when it comes to spam that I report and the frustration of when I get the automoderation notification back and they determine it didn’t break their guidelines.
I don’t want the cesspool of spam to become a problem on the fediverse as these experiences were a major reason I decided to join it.
I appreciate your comments again. You’ve softened how I feel on the challenge of balancing open user sign ups with moderating malicious users.
At least on the Fediverse when it comes to that most platforms will take spam more seriously than monopolized media, since even if they ignore their users they’ll listen to other admins, and if they don’t they become less popular. But there needs to be an initiative to improve on what big tech as done, rather than trying to reinvent the wheel.
I don’t particularly agree that it’s “greatly increased” by it. I’m on a server with them and I don’t see that much many more spammers. It seems like fearful hyperbole, I’m not saying there aren’t any I’ve certainly seen a few but it’s not as common as most people think it is.
Proof of Work Captchas are very effective when it comes to stopping or slowing down bots, while not impacting regular users much if at all. As for banning spammers automatically it is very easy to set up scripts that ban people for rapid posting, malicious links (bio, posts, comments, DMs), as well as words or phrases associated with spam (buy my Tshirt, check out my OF, see my pussy in my bio). For illegal content, image hashes can be used to automatically purge images and data that is known from the server, while simultaneously banning the people doing it. Rate limiting of suspicious IPs are also another way. There’s probably some more ones that can be done, as well as ways to improve these methods, point is they aren’t new, and they improve the situation for new users who can sign up easily and instantly without worry of rejection or too high expectation.
The big thing about these (except for captchas and PoW) is that they are post-signup methods that catch the people doing it in the act. It might seem like a loss for some since you don’t stop them from signing up at all but you need to remember. Spammers don’t give up, they don’t stop, they don’t care, real normie users do though. You can increase the bar and make them seem to disappear, but then you also lose the flow of real users, as most of them are filtered out or quit. Detecting and purging them automatically is the best way to combat them, captcha and PoW stop people using scripts for batch signup but banning them automatically when they start. If done right, spammers would be auto-banned after a few seconds of spamming or account setup depending which flags they trip first.
My mastodon instance rarely ever sees spam despite having open signups either. Maybe my problem is I’m worried about future problems rather than the current state of how things are. My concern comes from personal experience with legacy socials like facebook and twitter when it comes to spam that I report and the frustration of when I get the automoderation notification back and they determine it didn’t break their guidelines.
I don’t want the cesspool of spam to become a problem on the fediverse as these experiences were a major reason I decided to join it.
I appreciate your comments again. You’ve softened how I feel on the challenge of balancing open user sign ups with moderating malicious users.
At least on the Fediverse when it comes to that most platforms will take spam more seriously than monopolized media, since even if they ignore their users they’ll listen to other admins, and if they don’t they become less popular. But there needs to be an initiative to improve on what big tech as done, rather than trying to reinvent the wheel.