A video I recently saw about this problem: YouTube's Science Spam Crisis
Spam videos have always been a huge problem just from simple automated scripts that create bulk quantities of useless videos. But now with ai the videos are a bigger problem because they seem real enough at first glance and people are less likely to click away immediately and will have their time wasted more.
Doesn't matter that much. Personally I'd just go with virtualbox because it's open source.
If I recall correctly, one of the things that led to it finally being shut down was that having community of people interested in that content inevitably led to them sharing the real thing via dms
.ml tends to be one of the most abused top level domains for malware, spam, etc (in terms of ratio of malicious to non malicious domains) similar to .top, .buzz, .club, etc. So, many DNS filters on company networks simply filter all domains of these TLDs (and maybe whitelist a few known good ones) since they tend to be almost certainly malicious.
I filter them on my home network too via pihole (though not .ml)
Reflector is broken now because it's hardcoded to test the community.db file, but that file is now just an empy stub as that repo was merged into extra. Reflector is just a python script so you could edit it to replace mention of community.db with extra.db until the devs fix it.
The poneytelecom IPs would just constantly remain connected to me without actually downloading or uploading anything, which is quite unusual because torrent clients normally are supposed to disconnect from peers that they have no use for. And there would be like 15-30 IPs doing the same thing on the same few torrents. They were using Deluge, a legitimate client, which is quite weird, so maybe their shit was just misconfigured accidentally somehow. I looked up one of them on iknowwhatyoudownload.com and it was active on thosands of random torrents (including lots of CP apparently). I also recall in the past another IP from that range repeatedly downloading the same 80 GiB torrent which I am the only seed on, wasting my bandwidth for no apparent reason. So I just banned the entire IP range since clearly it's not doing anything legitimate to me and is just acting strangely in all sorts of ways. It's sort of a mini DDoS attack (intentionally or not) since I have my qBittorrent configured with a max number of connections.
The Xunlei IPs aren't really attackers per se, but the client doesn't follow the BitTorrent protocol standard and seeding to them is useless since they are incapable of seeding to other people. Some people just ban China entirely but I can't do that because there are lots of legitimate Chinese users on the torrents I have and I don't want to cut them off over something other people do
I've found that the block lists on the net tend to contain extremely outdated information and blocks a lot of legitimate activity, while ultimately being ineffective at actually blocking copyright trolls sufficiently. Best to have a vpn to prevent that. Since I have a vpn, I don't care who downloads from me so long as they aren't abusing my resources. So I manually create a blocklist for IP blocks I've observed malicious activity from. The blocklist file syntax is a note and an IP or IP range (not cidr notation) on each line, separated by a colon. for example, to block 195.154.0.0/16:
Poneytelecom:195.154.0.0-195.154.255.255
(That's an IP range I actually block, belonging to poneytelecom, a very low reputation hosting provider I was getting some weird denial of service looking activity, like 40+ simultaneous connections who wouldn't actually download anything)
Also, if you download torrents popular in China you may come across the Xunlei client, which always reports its progress as 0% and never seeds. Banning these would be impractical game of whack a mole. So instead, simply enable super seeding mode on those torrents. Gone instantly. Might be slower at seeding, but at least now you can seed to legitimate users.
@Supermariofan67
@lemmy.fmhy.ml