[Sohcahtoa82]

This carries the same energy as "People will break into your car no matter what, so just leave your doors unlocked."

[bigbuppo]

You say that, but I know someone whose house had their front door kicked in by burglars even though it wasn't even locked.

[yencabulator]

This actually happened to me. The seats had moved and glovebox was open one morning. Then a second break-in a few days later, and this one damaged the door panel near the lock. I left the doors unlocked for a couple of weeks after, to decrease the break-in damage -- there was never anything of value in the vehicle.

[tptacek]

The energy here is "so don't leave anything valuable in your car".

[angry_octet]

Unfortunately that is not what they proposed. To stretch the automotive analogy too far, you could say: if you invite a carjacker in, their seatbelt is not going to stop them from carjacking you.

[tptacek]

"Avoid shared-kernel attack surfaces" is not an unreasonable proposition in 2026.

[JackSlateur]

Virtual machines are still the best design and has been for something like 20 years

Containers are good, as long as they all share the same purpose (read: same application, no multi-tenant)

We all know that multi-users systems (and thus, containers) have a very wide attack surface, while VM attack surface is very limited ..

This is why I am totally convinced that:

  - redhat and friends are a terrible idea (licencing forces collocation which reduces segmentation)
  - per-instance pricing (read: cloud public, but not only that) are terrible: for the same reason. Paying per consumed CPU/ram is sane, paying per VM unit is damageful

[angry_octet]

Yes that is reasonable, but dispensing with all on machine controls is not.

[0123456789ABCDE]

isn't root level access one of the selling points of the cloud vm product line?

[angry_octet]

That doesn't mean you should run your services as root, it means other users are not sharing your machine/ kernel.

[__float]

It is very good practical advice.

It also saddens me greatly, imagining what computing could look like if systems evolved differently.