Look, we can debate the proper and private way to do Captchas all day, but if we remove the existing implementation we will be plunged into a world of hurt.
I run tucson.social - a tiny instance with barely any users and I find myself really ticked off at other Admin’s abdication of duty when it comes to engaging with the developers.
For all the Fediverse discussion on this, where are the github issue comments? Where is our attempt to convince the devs in this.
No, seriously WHERE ARE THEY?
Oh, you think that just because an “Issue” exists to bring back Captchas is the best you can do?
NO it is not the best we can do, we need to be applying some pressure to the developers here and that requires EVERYONE to do their part.
The Devs can’t make Lemmy an awesome place for us if us admins refuse to meaningfully engage with the project and provide feedback on crucial things like this.
So are you an admin? If so, we need more comments here: https://github.com/LemmyNet/lemmy/issues/3200
We need to make it VERY clear that Captcha is required before v0.18’s release. Not after when we’ll all be scrambling…
EDIT: To be clear I’m talking to all instance admins, not just Beehaw’s.
UPDATE: Our voices were heard! https://github.com/LemmyNet/lemmy/issues/3200#issuecomment-1600505757
The important part was that this was a decision to re-implement the old (if imperfect) solution in time for the upcoming release. mCaptcha and better techs are indeed the better solution, but at least we won’t make ourselves more vulnerable at this critical juncture.
There’s nothing stopping instance owners from incorporating their own security measures into their infrastructure as they see fit, such as a reverse proxy with a modern web application firewall, solutions such as Cloudflare and the free captcha capabilities they offer, or a combination of those and/or various other protective measures. If you’re hosting your own Lemmy instance and exposing it to the public, and you don’t understand what would be involved in the above examples or have no idea where to start, then you probably shouldn’t be hosting a public Lemmy instance in the first place.
It’s generally not a good idea to rely primarily on security to be baked into application code and call it a day. I’m not up to date on this news and all of the nuances yet, I’ll look into it after I’ve posted this, but what I said above holds true regardless.
The responsibility of security of any publicly hosted web application or service rests squarely on the owner of the instance. It’s up to you to secure your infrastructure, and there are very good and accepted best practice ways of doing that outside of application code. Something like losing baked in captcha in a web application should come as no big deal to those who have the appropriate level of knowledge to responsibly host their instance.
From what this seems to be about, it seems like a non-issue, unless you’re someone who is relying on baked in security to cover for your lack of expertise in properly securing your instance and mitigating exploitation by bots yourself.
I’m not trying to demean anyone or sound holier than thou, but honestly, please don’t rely on the devs for all of your security needs. There are ways to keep your instance secure that doesn’t require their involvement, and that are best practice anyways. Please seek to educate yourself if this applies to you, and shore up the security of your own instances by way of the surrounding infrastructure.
I think that’s a heck of a loaded assumption there that I’m relying on the Devs here
Cloudflare ✅ Strict Firewall Rules ✅ Hosting on an actual cloud provider rather than my home ✅ CSAM Scanner ✅
However, that’s come with other tradeoffs in useability, speed, and fediration experience.
Just today I found that the OWASP managed rules in Cloudflare end up blocking certain functions of the site, sure I’ll be adding an exception/rule for that, but it’s not a straight forward task. Heck, the removal of websockets will require quite a few changes in my Cloudflare config.
Sure, someone truly concerned with security knows to do this, but that’s definitely not going to be everyone, and now with the current spam situation we’re turning individual instance problems into “everyone problems”.
Like what? If properly configured none of the things listed should negatively impact hosting a Lemmy instance.
It honestly should be to someone who would be hosting any public web application using Cloudflare. Cloudflare makes all of this quite easy, even to those with less experience.
What config are you referring to? In the Cloudflare console? For websockets changing to a REST API implementation there should be nothing at all you need to do.
And it shouldn’t have to be everyone, only those who take on the responsibility of hosting a public web application such as a Lemmy instance.
No matter the capabilities inherent in what you choose to host, the onus rests on the owner of the infrastructure to secure it.
Everyone should be free to host anything they want at whatever level of security (even none) if that’s what they want to do. But it’s not reasonable nor appropriate to expect it to be done for you by way of application code. It’s great if security is baked in, that’s wonderful. But it doesn’t replace other mitigations that according to best practices should rightfully be in place and configured in the surrounding infrastructure.
In the case of the captcha issue we’re discussing here, there’s more than enough appropriate, free solutions that you can use to cover yourself.
Can you elaborate which functions are blocked by the managed rules? I haven’t noticed anything legit being blocked yet, just a bunch of obviously malicious things.
Yeah, I can’t seem to upload photos without whitelisting /pictrs/ from the OWASP managed ruleset. It wasn’t being “blocked” but it was trying to do a managed challenge and the lemmy-ui’s code didn’t really understand what to do with it. so it would just throw an error on upload.
I would recommend reconsidering that solution - I’ve already seen some malicious image uploads which Cloudflare has caught and prevented. For example:
Maybe you can check which specific rule from the ruleset was being triggered? For me, legit uploads are still working with the default ruleset (as you can see by the screenshot I uploaded in this very comment), so maybe you enabled some extra rules?
Interesting, well, I guess I sound vague because the error was pretty vague:
Cloudflare OWASP Core Ruleset
949110: Inbound Anomaly Score Exceeded
So yeah, your example is from the Standard Managed Ruleset, I wouldn’t even think of Disabling that, and I think this issue is limited to this OWASP one only. I think I’m still safe here, but I think I can just exclude only this one particular rule as you noted.
EDIT: Nope that’s the one you cannot edit. Strange, I’m fairly certain these files are fine and it’s probably not that - I may have to exclude the entire ruleset here.
Double Update: Yeah, so this OWASP one is far to sensitive. I’ve validated my files with AV and other solutions and tried other machines and such. Apparently it’s tripping some XSS rules, SQL injection detection rules, and a few other things. Mostly seem like false positives on account of the EXIF and other file header data.
Triple Update: Found the proper way to exclude from specific rules in the managed rulesets. There’s like multiple ways that you appear to be able to do it, but only one way that works.
I’m surprised some large instances aren’t using Cloudflare. It takes a few minutes to setup and the added benefit of caching alone is worth it. Let alone the bot/ddos protection.
I know right? The free tier would be enough to handle most anything and would take a tremendous load off of the origin server with proper Cache Rules in place. I can’t remember which instance it was, but one of the big ones started to use Cloudflare but then backtracked because of “problems”. When I saw that, I couldn’t help but think that they just didn’t know what they were doing. My instance is currently behind Cloudflare, and I’ve had no problem whatsoever with anything.