IPMI/BMC have been known problems since 2015, if not before.

Matthew Garrett: IPMI - because ACPI and UEFI weren’t terrifying enough

I feel the same way, but can you explain how a VPN would help achieve that in the face of Google’s proposed developer verification scheme?

How would that help you?

The article says, “unverified apps will only be easily installable in the sanctioned countries where verification doesn’t exist.”

So even if you used a VPN to trick Google into thinking you live in a sanctioned country, you wouldn’t be able to distribute apps to those countries without breaking sanctions.

Progress report 1

• I’m ruling out HMAC-SHA1, because:
  • Plain vanilla HMAC-SHA1 uses a shared secret. I don’t want the PC/server to have a copy of the credential. May also be susceptible to MITM/relay attacks.
  • Rolling-code HMAC-SHA1 only partially solves those problems.
  • Encrypted rolling code HMAC-SHA1 solves those problems, but I haven’t found a reliable source for using it with LUKS.
• If using systemd, pick FIDO2:
  • Avoids flaws of HMAC-SHA1.
  • Native support in systemd, so “future-proof”.
  • Wider support than OpenPGP. More HST vendors to choose from, including cheaper options than NitroKey or Yubikey: useful if each sysadmin (or colleague, or relative) needs an HST.
  • Compatible with QubesOS.
• Otherwise, OpenPGP:
  • Like FIDO2, solves HMAC-SHA1 flaws.
  • However:
    • smartcard-key-luks seems unmaintained on GitHub and on GitLab.
    • LUKS with OpenPGP isn’t well-documented for non-Debian-based distros.
• TBD: Clevis/Tang:
  • Remote/network-based unlocking.

Thank you for this! That thread is helpful in itself, and also links to other relevant resources - including by Lennart Poettering (controversial guy, but the canonical source on systemd).

This coverage has a better headline: Trump blames Minnesota governor for cyberattacks against the state. Cites no evidence.

Funding restrictions and denialism, plus starting a war against a cyberattack-capable nation, made infrastructure attacks inevitable.

Security researchers have been publicly raising the alarm about SCADA vulnerabilities for 20+ years.

If it’s a server for self hosting you definitely don’t want anything that requires interaction at boot.

Depends on use-case. If you only plan to boot it when you’re physically present, it’s fine.

i believe a much better secure layer is something similar to what Novacustoms, Purism attempt to do: verify if somebody else not you try to access the laptop.

You’re thinking of Heads, which I agree is ideal for supported motherboards.

tang

Thanks. TIL about Clevis/Tang.

I read them before writing my OP. I’m still not sure what you’re getting at.

I would be grateful if you could say what you mean, instead of initiating an oblique guessing game.

Yes. Here are some common self-hosting scenarios:

• Home server containing family files: scans, photos, device backups, …
• Office server containing business files: sensitive documents, device backups, …
• Web or email server containing websites, Fediverse instances, emails, etc

In all those cases, full disk encryption (FDE) is a sensible precaution to protect the data in case the server is physically stolen.

Linux is probably the most common OS kernel for self-hosting. On Linux, LUKS (Linux Unified Key Setup) is probably the best FDE system. It’s mature and reliable. But anyone self-hosting a Linux server with LUKS FDE is faced with the question of where to store the keys.

Hardware security tokens (HSTs) are widely considered a safer place for keys than SSDs, HDDs, or USB storage. They follow the smartcard principle: a private key can be written to an HST but not read from it (security vulnerabilities excepted). Instead, they implement cryptographic algorithms to prove possession of the private key. So, anyone self-hosting a Linux server with LUKS FDE should strongly consider storing their private key(s) on an HST.

However, there is more than one way to do that. Hence the question in my OP.

Which of the 4 recipes I posted are you referring to as “this”?

For plain text & Markdown, you don’t even need Collabora. Since 2019, NextCloud has a built-in collaborative text editor.

it’s the only way!

TBF I was trying to follow the docs, aber ich spreche kein Deutsch.

I overlooked this line from the docs:

Benötigt werden cmake, git, libgd und libusb - bei manchen Distributionen (z.B. Ubuntu) sind die Header-Files in separaten Paketen “libusb-dev” die ggf. noch nachinstalliert werden müssen.

I should have done sudo apt install libgd-dev libusb-1.0-0-dev before attempting compilation.

@[email protected]

It would be trivially detected by widely distributed standard forensic software including the non-Premium variant of Cellebrite able to run on a laptop.

By “duress profile”, I mean that if user has enabled a “duress profile” feature in Settings, then entering the duress PIN would:

1. Erase (the encryption key for) all profiles and storage outside the duress profile; then
2. Unlock the duress profile.

So, how would forensic software detect that the unlocked profile is a duress profile?

@[email protected] everyone seems to be clamouring for the same thing: add “duress profile” to the roadmap. Keep up the good work.

https://alternativeto.net/software/find-my-iphone/?license=opensource