Like if I edit a post, can they see what the pre-edited post was?
Based on this closed issue, I don’t believe Lemmy natively keeps an edit history, the comment is overwritten. However, admins have full access to all data on the server database so they can (and do) keep backups and look over them at any time. You should consider everything you do on your instance available to your admin, including private messages.
Simple rule of thumb: if you don’t own it, assume the owner has access.
If it’s been online for any amount of time, always assume that someone has seen it, screenshotted it and asked the wayback machine to archive it.
Chances are they haven’t, but if you’ve leaked a password or key, better change that right away.
Everything.
Well if it’s a post, that’s not private information anyway, even if you edited it. As for actual private data, they can see your email address.
Everything you send to the server
Don’t forget it may have also been replicated to other instances before you edit or delete it where those admins also have access to the database.
Can they see passwords? If so, I recommend not using the same stuff for every instance.
Lemmy is open-source. Anyone can modify it as they wish, so do consider that a yes.
In fact, it is better to assume every service you use stores passwords in plaintext.
Always assume they can. Use a password manager and have different passwords for each service.
Very unlikely. As Lemmy is open source you can see how it talks and what functions it does. If they were storing passwords in clear text someone would have raised an alarm siren about it.
Best case is they are using a salt and hash method to store passwords making them incredibly hard to brute force.
And every lemmy instance can change the source. For every website there is, assume every worker from CEO to the janitor can read your password in plaintext.
This is also true.
Can you verify the software running on an instance is the same as the one in the source code repository? You can’t. Can you verify the instance isn’t running code to read passwords from your login requests even if the code is the original open source code? You can’t.
That’s why (and for other reasons) you should never use a password for more than one site/service/instance.
Lemmy admins (admins in the Lemmy application) probably can’t read your password. But everyone with admin rights on the server operating system can.
That is why I said unlikely. Yes it can happen. All it takes is one admin to look and go ‘Why are we storing passwords in clear text?’ and that instance is burnt.
