Hacker News new | past | comments | ask | show | jobs | submit login

Forcing you to query IA at least reduces the frequency of feedback, as they're taking snapshots instead of giving a live feed. You could also shadowban IA. You can also do things like guess based on IP address or browser fingerprinting, or require a login from various IP ranges.

Of course your main point - that this is all terribly imperfect and won't stop a determined, sophisticated user, who has realized what's happening - is spot on. That, however, is perhaps a rare combination, rare enough to simply continue dealing with manually.






The feedback doesn't have to be very fast. If I'm botting correctly, my accounts will almost never be banned. Even once every 24 hours will be more than enough.

IA was just an example, and Tor would be easier. But anyway, I think it shows the flaw in doing so:

> You could also shadowban IA.

If the spammer manages to get all the IPs hellbanned just by looking at things, he gets more eyeballs on his spam.

My point is, you can't get much better than normal shadowbans, which are trivially detectable for moderately sophisticated users (just log out and try to check your profile) but not anyone else. "Hellbanning" is a stupid extension of this concept which only works in video games.

Also, shadowbanning is a spineless and deeply unethical move. If I get banned, I know what I did wrong and can reflect on that. If I get shadowbanned, I'm just screaming into the ether. That is not a Good Thing™, it is atrocious.


> Even once every 24 hours will be more than enough.

Depends on the use case. Once every 24 hours is a lot easier to moderate than a minute by minute spam wave.

> IA was just an example, and Tor would be easier.

TOR would indeed be easier... assuming it's not already blocked through other means, as it frequently is. There's a whole ecosystem around blocking TOR and other proxy mechanisms - imperfect and permeable though they may be.

> My point is, you can't get much better than normal shadowbans

Not sure I agree or that you've supported your point - however, even shadowbans are often unnecessary. The goal is never perfect moderation, merely to stack the deck in favor of the moderators for blocking problematic content in terms of time effectiveness until either the moderation effort available can handle it, or until the spammer moves onto easier, more cost effective targets (which even basic shadowbanning can achieve, mooting the need for better tools even if they're available.)

> Also, shadowbanning is a spineless and deeply unethical move.

As a first line of defense against mere rules breakers, I might agree. As a second, third, or nth line of defense against particularly problematic ban evaders and spambots, I will gladly resort to such tools - or worse - and sleep soundly at night.


> 'As a first line of defense against mere rules breakers, I might agree. As a second, third, or nth line of defense against particularly problematic ban evaders and spambots, I will gladly resort to such tools - or worse - and sleep soundly at night.'

So then you will have no problem recognizing where I'm coming from when I say that the Hacker News 'mods' are some of the most evil Nazis on the planet. After all, I created my first account here only a few months ago, and was shadowbanned in less than 15 posts despite posting with honest intentions. Is that what decent people do to other people?


I guess we know who gets banned from places!



Applications are open for YC Summer 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: