I know that people like to dump on Cloudflare, but it’s incredibly easy to enable a built-in CSAM scanner with CloudFlare.
On that note, I’d like to see built-in moderation tools using something like PDQ and TMK+PDQF and a shared hashtable of CSAM and other material that may be outlawed or desirable to filter out in different regions (e.g. terrorist content, Nazi content in Germany, etc.)
I don’t want much, I just want deletion to be propagated reliably across the fediverse. If someone got banned for CSAM and their contents purged, I want those action propagated across all federated instances. I can’t even delete my comment reliably here on Lemmy since many instances doesn’t seem to get the deletion requests.
People are wary about how internet got more and more centralized behind cloudlare. If you’re ever getting caught in cloudlare’s captcha hell because they flag your IP as suspicious, you’ll get wary too because you suddenly realized how big cloudlare now when half of the internet suddenly ask you to solve cloudlare captcha.
But, no, however the sentiment makes sense and as I am trying to disperse / decentralize most everything I can these days, including getting away from Google services, for example, this does make sense as well.
Lemmy.world had to start using CloudFlare because some script kiddies were DDOSing it. Some people were complaining that it encourages centralization, etc.
Personally, I love it. The service you get even at the lowest level of payment ($20/mo) is great. And what you get for free can’t be compared.
It looks like it scans and flags on the outbound (user download of the image), so as long as it sits in front of your instance, it should work just fine.
You’re still responsible for removing the material, complying with any preservation requirements, and any other legal obligations, and notifying CloudFlare that it’s been removed.
It would be ideal if it could block on upload, so the material never makes it to your instance, but that would likely be something else like integration with PhotoDNA or something similar.
I know that people like to dump on Cloudflare, but it’s incredibly easy to enable a built-in CSAM scanner with CloudFlare.
On that note, I’d like to see built-in moderation tools using something like PDQ and TMK+PDQF and a shared hashtable of CSAM and other material that may be outlawed or desirable to filter out in different regions (e.g. terrorist content, Nazi content in Germany, etc.)
I don’t want much, I just want deletion to be propagated reliably across the fediverse. If someone got banned for CSAM and their contents purged, I want those action propagated across all federated instances. I can’t even delete my comment reliably here on Lemmy since many instances doesn’t seem to get the deletion requests.
Wait, why do people like to dump on CloudFlare? I must be out of the loop.
People are wary about how internet got more and more centralized behind cloudlare. If you’re ever getting caught in cloudlare’s captcha hell because they flag your IP as suspicious, you’ll get wary too because you suddenly realized how big cloudlare now when half of the internet suddenly ask you to solve cloudlare captcha.
Got it. Damn. I just switched some services to CloudFlare DNS today…now I guess I’ll change them back.
Are you so easily swayed by a single post from a rando internet person?
Absolutely.
But, no, however the sentiment makes sense and as I am trying to disperse / decentralize most everything I can these days, including getting away from Google services, for example, this does make sense as well.
He actually looks at others’ opinions and weighs the pros and cons unlike almost everyone here on lemmy.
Cloudflare DDoSes about 23% of the internet. Stop defending Cloudflare.
Lemmy.world had to start using CloudFlare because some script kiddies were DDOSing it. Some people were complaining that it encourages centralization, etc.
Personally, I love it. The service you get even at the lowest level of payment ($20/mo) is great. And what you get for free can’t be compared.
Does the CSAM scanner hook into lemmy properly though?
It looks like it scans and flags on the outbound (user download of the image), so as long as it sits in front of your instance, it should work just fine.
You’re still responsible for removing the material, complying with any preservation requirements, and any other legal obligations, and notifying CloudFlare that it’s been removed.
It would be ideal if it could block on upload, so the material never makes it to your instance, but that would likely be something else like integration with PhotoDNA or something similar.