Basically, you should avoid emails if you care about privacy, there is nothing private about emails. Your email provider can read your sent/received emails and each recipient provider can do the same.
ProtonMail, Tuta, Posteo, Mailbox.org and custom encryption at rest will not fix the statement above. The encryption at rest done only prevent them from reading currently stored emails.
End-to-end encryption can help but they still know who send emails to who, at which frequency and each email size.
@marmarta I use restic, borg and wyng, basically I want this for a backup solution:
- encryption - remote storage (ssh or s3) that does not require an agent on the remote - ability to verify a backup - ability to check the repository integrity - ability to restore a backup (that's the point of the tool though 😅 ) - ability to exclude data (when applicable) - deduplication of data within repository - being able to configure auto deletion/pruning of old data using retention rules - bonus if you can backup multiple hosts to the same repository and use deduplication - show me some stats
The biggest issues I had with backups tools:
- inability to prune old data from a repo because the storage is full (restic problem) - a not clear documentation about integrity check: there are repository checks and data check, the former is quick because it just validates metadata while the other requires validating the whole repository data.
@farooqkz the security features are named pledge and unveil, the first restricts system calls the process can do, the latter restricts filesystem access.
Chromium and Firefox received both on OpenBSD, they are limited in term of system calls (not really visible for the end user), but also in term of filesystem, they can write in the directory ~/Downloads/ and read a ton of other directories useful for the runtime (settings, fonts, shared libraries etc...), nothing more.
FreeBSD has an equivalent named capsicum but they did not implement it into web browsers (checked a few months ago).
Linux has sandboxing using multiple methods:
- web browsers installed with snap or flatpak are sandboxed - web browsers started with firejail are sandboxed
@pitrh I think what @gyptazy meant, if I understood, was that it's easy to add someone's server to the ban list by filling a form sending an email to your honeypot address, through a registration form on a web service for instance
@pitrh when do you remove an entry from the blocklist? If an entry was due to a server being hacked, and it has been freshly assigned to a new customer inheriting the IP, how should that person proceed to not be blocked?
Also, a note about logs, due to GDPR it may not be legal to keep emails logs more than 2 or 3 years 😅 , but I'm not an expert and different laws of countries or EU often collide. It's still something to think about though.
Retention by default is pretty low, this can be configured on the command line parameter of the service (using rcctl set victoriametrics flags), everything is configured as flags in VM
@pthane@solene@matthew I've only occasionally run into issues with my business email address not being accepted in online accounts/stores. So far, it's always been because my domain doesn't use a historical core gTLD (`.tech`, in my case). Similarly, I have enough email accounts that I can use one that I prefer less, if necessary, but annoying to troubleshoot.
@karadoc I used a mail address that is not gmail and it worked fine, it's just that the domain I was using initially (an alias service) was blocked silently
My things are #OpenBSD, Transhumanism, Lovecraft and gamingI speak :fr: and :gb: and I'm 18+I type on my keyboard to do stuff, mostly promotimg self hosting, decentralization and libre software.GPG key https://perso.pw/solene.ascGPG fingerprint 4398 3BAD 3EDC B35C 9B8F 2442 8CD4 2DFD 57F0 A909