With the recent hack, there is now irrefutable proof of malicious actors trying to break Lemmy and steal user accounts. Please be careful about entering your password into random Lemmy apps!
I think for a while leading up to the recent session stealing hack, there has been a massive amount of positivity from Lemmy users around all kinds of new Lemmy apps, frontends, and tools that have been popping up lately.
Positivity is great, but please be aware that basically all of these things work by asking for complete access to your account. When you enter your Lemmy password into any third party tool, they are not just getting access to your session (which is what was stolen from some users during the recent hack), they also get the ability to generate more sessions in the future without your knowledge. This means that even if an admin resets all sessions and kicks all users out, anybody with your password can of course still take over your account!
This isn't to say that any current Lemmy app developers are for sure out to get you, but at this point, it's quite clear that there are malicious folks out there. Creating a Lemmy app seems like a completely easy vector to attack users right now, considering how trusting everybody has been. So please be careful about what code you run on your devices, and who you trust with your credentials!
This is why password managers are so heavily pushed. Imagine if you used the same password for Lemmy that you used for your email? Both are now compromised. A unique password for all accounts is the bare minimum you must do.
Although it is worth noting that the recent Lemmy hack didn't come from a password compromise, but from session token harvesting, which a password change would not really protect against.
Lol this is not how the hack worked. JWT cookies are encrypted. They don’t contain your password at all. There was no way to reverse engineer from your cookie to your password.
I literally just got one last week. It's so nice! I guess I was hesitant because centralized passwords always put me off, but it's not hard to setup 2FA on all your important stuff.
Sorry, but that's literally every online service. For example if you buy a new virtual server it takes like 5 minutes till a Chinese IP starts to try root passwords.
If someone actually wanted to harm Lemmy they'd just DDOS the biggest instances for a month (which would be easy, it's mostly single servers after all) or attack it with so much spam and large images that storage would break.
I'm getting CS-nerd excited about how this is all going to play out. Federated moderation is hard and so many awful, clunky things have been tried before. Are we actually going to see a web-of-trust or reputation system that reaches widespread adoption? It's gotta be silent and noninteractive as there's no way to expect normal people to put up with the complexity.
The difference is that when you buy a vps you aren't handing over all your access creds to random developers.
And "harming lemmy" may be an intent that sparks a DDoS but there are other intentions that should make users wary. Harvesting creds of people who reuse passwords across accounts is an easy example that could have more serious implications to the individual user.
Dude, you can't trust any Lemmy instance at all. It doesn't even matter that the code is open source, the instance owner could just compile their own version that sends them every password in plaintext. There is zero guarantee that your password is safe.
Anyone who reuses passwords has been pwned a dozen times already. Just check your own logins here: https://haveibeenpwned.com/
If you reuse passwords online you have a problem, it's simple as that. Even big companies had breaches that leaked user data, no company is safe. For example one of my old passwords got stolen from Adobe. One from Unreal Engine. And my old logins are currently shared in 2,844 separate data breaches. Not using a password manager with a random password per service nowadays is madness.
It helps, but it's still not a silver bullet. For example, a Lemmy app could contain no malicious code in its open source repository, but malicious code could still be added to a binary release in an app store.
Voyager (formerly wefwef) is a self-hostable web app, so it doesn't have this problem. Of course this only means you can inspect the code you're running. You still have to able to understand the code to be sure it's not doing anything malicious.
That's why F-Droid is the safest Android app repository. If I'm not mistaken, every app they offer is rebuilt from the public source code by the repo package maintainer.
Also if it’s a desktop app they could just put the malicious code in the binary download 99% of people will use, or if it’s a web app, they just put it in their hosted version, etc.
The safest option would be for Lemmy to implement OAuth and apps that aren't in some "official front end for xyz website mode" to authorize via OAuth with the backend instead of via credentials.
No, because open source apps need to have enough eyes on them to spot malicious code. And highly complex ones need proper audits and even that might not be enough to catch every fancy vulnerability.
Well I recommend to stay away from any closed source apps. Also make sure your password is unique to lemmy and use an email address that is not your primary email.
Well I recommend to stay away from any closed source apps. Also make sure your password is unique to lemmy and use an email address that is not your primary email.
Be careful of clicking links and DMs. If using the browser use noscript and block scripts by default.
But in this case, if an app is open source, there is a higher chance of discovering that it sends your credentials somewhere else than in closed source app
I really want this. I haven't tried any of the apps (other than the one I made) yet because I don't want to give anyone my password. Oauth support would be so nice.
Same, I was on world when it got hacked, rotated the simple login email, and rotated the password post hack (and deactivated the old email) just to be on the safe side.
Indeed, this is a real weak spot with Lemmy's security. I honestly think we need to place more emphasis on implementing OAuth2, when I have the time I'll have to take a look at that again to see if I'm able to.
Not sure about the other apps, but Memmy keeps your password stored in the app. But generally speaking I don’t think Gavin is a malicious person. He spent 3 weeks with beta tests before he was even comfortable pushing it to Apple App Store and the code is reviewed by Apple before being included to watch for malicious scripting. I am not saying people don’t get malicious scripts passed through Apple, but from what I have seen of Gavin’s work, he doesn’t seem the type that would do that.
We really need a better way to log in from 3rd parties. It'd be nice if we could crowd fund features, I'd def pay up for some sort of app-password system or other
Obviously you should only input account credentials into an app you trust, but shouldn't a properly designed Lemmy app not store the credentials in plain text at all? (And definitely never send them somewhere else) Authorize the user through the API and then it's just an authenticated session, no need to store the username/password at all until you sign out.
I suppose if you have fast user switching it might need to store it. Hmmmm.
When you pass your password into an app, it can just copy and store it. If at any point you put a password through, even once, its compromised and no password manager will help in the slightest.
I make bobApp, you download bobApp, if you put a password into it, you are done.
edit: answering both of you.
Then i misunderstood. After i typed this I noticed others said what you meant. My bad.
I feel like it raises somewhat the general issue of how much we're willing to live with complete mysterious anonymity from all of the developers and admins in the fediverse. I'm not saying that every admin or developer should have their real identity revealed and linked here. But there's a tension or issue here in how much it's normal and accepted and how much the fediverse in general wants to grow and attract users that are accustomed to trusting large companies that provide a different kind of base level trustworthiness than makes sense "out here".
If not links to real life identities (however trustworthy that can be in the limit), at least some connection to a broader online presence such that it becomes more likely the actor has something to lose in acting in bad faith (the lemmy core devs being a good example).
I don't have a solution ... but it seems to be a growing pain as this whole thing kind of grows from "hacker project" to "mainstream social media".
Preface: I have no opinion. Have not installed, and will not install because I don't need it.
If you have to ask, it's probably worth investigating further.
The first thing I would do is search around for any mention of the app name and variants on the word "password".
Never install anything that you don't trust, and dissociate from your "real" email/identity for those that you aren't sure about. Whoever owns the software potentially owns your credentials.
Apple doesn’t review it that thoroughly. They could easily send people’s credentials up to some server and Apple would likely not catch it.
What I mean is don’t just trust it because it’s in the App Store.
I personally use it because it has an active GitHub and is one of the more popular mobile clients. Also I don’t really care if my accounts get hacked in the first place lol so I’m also trying out Mlem beta and Wefwef. But even with that said I wouldn’t just try out any random new client that came along.
Feels like this will be a very common occurrence with people rushing to build and use new apps, and host new servers. There are plenty of positives to fediverse vs centralized, but it doesn’t come without negatives.
I've been wondering that myself. I've only entered my pw into Jerboa, which is made by the Lemmy devs (and Liftoff once, but changed the pw since).
Now I only ever use FOSS apps, which all seem to be under some amount of scrutiny, but idk how much is enough.
I've always been particularly wary of Voyager/wefwef. Not that I wouldn't trust the devs, but the whole concept of entering a password into a 3rd page that only passes it onto the right page, damn that's just dumb on principle.
It's particularly weird since this is home for so many techies and privacy/security advocates.
Its less dumb than entering it into a regular app compiled into an apk, which is more opaque (even if it's also FOSS). Voyager you can host it yourself.
All of the apps have you enter your credentials into their page because Lemmy doesn't support OAuth2. I don't think it's fair to criticize Voyager for a problem that is currently inherent to all Lemmy apps.
Yea but it's a local form on the device and not a 3rd party server, which is another layer of insecurity. And I'm not sure how much of the rest of communication needs to get proxied too.
I dunno, that's just way beyond my comfort zone unless I really want to self-host that stuff.
Anyway, okay, nothing seems to be all that well secure at this point.
it's about time to change that password and i think lemmy team should have an option in the settings to revoke access we give to third party apps....not sure if that's possible
The greatest part of Bluesky/atProtocol vs ActivityPub is that they let you sign in to one account on any service on the protocol and it requires you to make separate app passwords.
Users getting invite codes every X weeks is nice for the server the require registration too