And what insane key length will prevent the people still running "normal" CPUs to build a big Distributed Computing cracking Network to crack the key before buying one of these things (after getting the public key somehow like you described)?
As soon as one Version of Office supports only DRM the transition from an old version of MS Office to another Office will be just as easy as the transition to the new DRM-only version. MS will lose the lock-in effect if they do that that keeps most companies on MS Office.
The moment the American Industry tries to introduce something like this (even in the computer sector only) it will lose most of it's business to industries in countries like China that don't care about American Laws and will use it to gain Worldwide market share.
1. You keep talking about "unbreakable DRM" which is simply BS (see each and every copy protection out there)
2. What you describe is basically plain old Pay-per-View which isn't attractive in any way to the consumer as long as there are alternatives
3. If there are no alternatives they would charge 12$ or 120$/hour not 0.12$/hour (see mobile phone ring tones,...)
And why would anyone create websites that work for the 1% of trusted computers and not for anyone else? Websites usually want more customers, not less. It will not work for the same reason we got huge problems with the ipv4 to ipv6 migration. You can either support one or the other or have more work without any significant advantages. Which one will you support, the 99% one (untrusted) or 1% (trusted)?
This is simply a bad design decision. Running everything (including GUI and other things where this is totally unnecessary) as root/Admin is not a good design. Running as few things as possible with maximum rights (and everything with minimum rights necessary) is the best design from the security point of view.
Not entirely wrong but you forget that the Linux versions of the traditional Unix System Calls are still new implementations of the same call which means they can contain bugs other implementations of the same calls do not.
But still two steps are needed compared to a remote root exploit (basically all Windows exploits that work without user intervention) where one step is enough.
Just use a livecd, mount your harddisk linux install, chroot to it and use passwd to set a new password (assuming you use the root account of the livecd)
Local root exploits can be used via ssh if you have a normal user account on that machine so it is not the same as physical access (in which case you just reboot with init=/bin/bash as kernel parameter or use a livecd to get root)
I live in Germany and I can't even remember the last time we had a real power blackout. Sometimes we have these under-a-second ones that restart the computer and the lights flicker once but even those are extremely rare.
And learning to enjoy learning is something that should help anyone in any career
I enjoyed learning before college (and still do) and my college courses do anything possible to destroy this feeling in me by testing the wrong things at the end (the things I would normally categorize as "I can look it up when I need it").
The command line will only be around as long as there is a keyboard... and the keyboard won't live forever.
And why should there ever be an interface which can be used to program but not for the command line? The command line uses extremely short and powerful constructs that are simply not available on a GUI so it will not die because of lack of usage either.
And then you will have "uuuhh, nice tits" and things like that in your source code matching the time when the hot secretary passed your open office door (cubicle for you Americans).
Slashdot Mirror
And what insane key length will prevent the people still running "normal" CPUs to build a big Distributed Computing cracking Network to crack the key before buying one of these things (after getting the public key somehow like you described)?
...which in turn will lead to the device being returned as defective by geeks and "normal" people experiencing false positives.
...and let them know why you switch.
As soon as one Version of Office supports only DRM the transition from an old version of MS Office to another Office will be just as easy as the transition to the new DRM-only version. MS will lose the lock-in effect if they do that that keeps most companies on MS Office.
The moment the American Industry tries to introduce something like this (even in the computer sector only) it will lose most of it's business to industries in countries like China that don't care about American Laws and will use it to gain Worldwide market share.
No to mention that you have other platforms to choose from especially when using Linux.
More likely there will be portals for those using DRM Hardware since they will be a minority (even the whole US would be a minority on the Internet)
Anyone who does not get it should look here: http://www.paranoia-live.net/
1. You keep talking about "unbreakable DRM" which is simply BS (see each and every copy protection out there)
2. What you describe is basically plain old Pay-per-View which isn't attractive in any way to the consumer as long as there are alternatives
3. If there are no alternatives they would charge 12$ or 120$/hour not 0.12$/hour (see mobile phone ring tones,...)
+1 Good Analogy
And why would anyone create websites that work for the 1% of trusted computers and not for anyone else? Websites usually want more customers, not less. It will not work for the same reason we got huge problems with the ipv4 to ipv6 migration. You can either support one or the other or have more work without any significant advantages. Which one will you support, the 99% one (untrusted) or 1% (trusted)?
This is simply a bad design decision. Running everything (including GUI and other things where this is totally unnecessary) as root/Admin is not a good design. Running as few things as possible with maximum rights (and everything with minimum rights necessary) is the best design from the security point of view.
Not entirely wrong but you forget that the Linux versions of the traditional Unix System Calls are still new implementations of the same call which means they can contain bugs other implementations of the same calls do not.
But still two steps are needed compared to a remote root exploit (basically all Windows exploits that work without user intervention) where one step is enough.
Just use a livecd, mount your harddisk linux install, chroot to it and use passwd to set a new password (assuming you use the root account of the livecd)
Local root exploits can be used via ssh if you have a normal user account on that machine so it is not the same as physical access (in which case you just reboot with init=/bin/bash as kernel parameter or use a livecd to get root)
I live in Germany and I can't even remember the last time we had a real power blackout. Sometimes we have these under-a-second ones that restart the computer and the lights flicker once but even those are extremely rare.
Not to mention me walking in his room with a "rm -Rf /" T-Shirt.
You know there are OS besides Windows?
You know, there are these things called 'books' that work without power I've heard...
With Fvwm you can't get work done because you are always tweaking the config file to match your needs (see also 70+ pages thread in the Gentoo Forum)
And then you will have "uuuhh, nice tits" and things like that in your source code matching the time when the hot secretary passed your open office door (cubicle for you Americans).