Can you expand on this?
All modern VM solutions support snapshotting (let's disregard zones/jails/lxc/openvz containers which work a bit differently, but the fundamental problem is the same - host can do something without guest ever knowing).
Snapshots is exact replica, both disk, ram and system state of a live guest system, with that data it can be resumed - "forked" anew somewhere else, the guest never "reboots". When LEA writes a gag order for a VPS, they no longer ask for mere disk images, but for a live VM snapshot. Snapshots work atomically, without the guest system ever knowing. Normally it is used for live system migrations, load balancing and more.
Wrt encfs, same situation as dmcrypt - are you keeping it mounted, or not? If it is mounted, the raw encryption key ends up in the snapshot - OS has to be able to access the disk somehow.
As for ssh, it decrypts private keys with passphrase only momentarily during auth phase, but keeping keys on the server as such (even if passphrase protected) is really bad idea anyway (what if the box gets actively backdoored to snoop passphrases?). Just use agent forwarding where the server becomes literally a 'dumb pipe'.
Good luck finding one that will allow you to encrypt your whole partition.
Pretty much all providers with KVM technically support this, whether they intend it or not.
The problem is more fundamental. With VPS (as apposed to physical box where one can be certain that RAM snapshots are not easily made) - encrypting a partition won't help, just waste the already constrained cpu cycles. The provider still can trivially make a snapshot of your running VM, including live keys in memory, so what's the point? This is why businesses are not particularly fond of cloud things, and tend to replicate that stuff in-house. There's a field of ZK utility computing, but it is still quite inefficient.
Yet you can't be bothered to explain this alleged complexity. My view is that it really is that simple.
Realistic conflict theory is much more than just "duh, ig/og". The complexity is of why competing cliques are formed in the first place - the explanations in terms of racial/gender (both left and right) theories appear indeed simple, but biased to the point those also completely contradict empirical research in social sciences.
But when you start picking and choosing which ethnic groups to protect while other ethnic groups aren't even recognized, it's quite clear that's the usual in-group/out-group dynamics going on.
I agree it's a problem. Right wingers are concerned about reverse racism while social left is concerned that protecting classes does not protect certain groups sufficiently (thus you get AA, which antagonizes right, in endless cycle).
You're correct that protected class is simply protection of attribute as such, regardless of value of that attribute. Can't discriminate based on sex or ethnicum, regardless of what actual gender or race one is.
Next should be currency that any of us are allowed to create
Already the case with most of keynesian currencies, aka elastic supply. While not really "anyone" can "print" the IOUs, when borrowing money, reserve banking "prints" those out of thin air.
The only way to make it even more fair would be simply scrap the IOUs and print naked (which is what "negative interest rate"s pretty much boil down to) - but as a basic income.
This basic income/QE infinity hybrid will coincide with immense deflationary pressure (already somewhat present) because of capital concentration and people being pissed that majority of cheap credit ("printing") currently ends up fueling illiquid ponzis such as housing bubbles and corporate buybacks, instead of consumer economy.
"Pascal is Pascal is Pascal is dog meat." -- M. Devine and P. Larson, Computer Science 340