Justin said:
I get it.
I don't need any escalation to admin. The problem is, what if there's
some malware. Some malware named "winenhancer." The user sees the UAC
prompt "Winenhancer must access the internet!" and the user clicks on yes.
So UAC only works when the user knows everything about the PC, which is
unrealistic for a standard dumb user whose job is to type out proposals
and reports.
Oh, I get it. It's not the responsibility of the dumb user to know what
he or she is dumbly clicking on as they point and click. It's their
responsibly to know the situation, but they don't and most never will.
However, network admins take that responsibly for this type of worker
by using a network proxy that only allows the users to go to approved
sites closing the attack vector and mitigating such damage, as its their
responsibility to protect company's interest and not some office clerk,
lock them down.
Just like with Linux which has the same kind of an approval process
within its O/S, they point, click, approve and it's all bets are off.
But with UAC enabled when one does this, the damages are mitigated to a
certain degree as UAC protects critical areas and also not allowing the
malware to continuously run under the context of the user-admin
full-rights access token, to spread damage.
But rather with UAC enabled, the compromise runs under the context of
the admin's Standard user token, because admin user on Vista is returned
to using that token upon privileged escalation completion, and it's a
limit rights token, which mitigates/limits damage.
Like I said, nothing is bulletproof not even god's O/S Linux, but UAC on
the MS platform is better than have nothing at all, which is the case in
fact with the previous versions of the NT based O/S platform, open by
default O/S(s), to help protect the O/S.
