Hi David,
And, the security required to ensure that it's safe against
the silly pranks and and malicious hacking of tens of thousands
of Internet denizens, ranging from junior-high-school script-
kiddy wannabes up through sociopathic cybercriminals.
That's actually not hard to do --if you *plan* on that
from the start and don't just "bolt it on", later.
E.g., you can *know* my IP address, install a packet sniffer
on my network feed and *still* not gain access to anything
inside the home.
What's more, I can hold the front door open for you, help
you cart whatever kit you want *into* the house and *still*
be safe from your attempts at hackery (short of burning
down the place).
The best you can do is deny me access to a control channel
over the same medium that you happen to be using. But,
you could ALWAYS do that: just cut the damn wire and
I can't use it! (i.e., silly to spend any effort beyond
this level of protection)
The problem with most "remotely accessible" devices (whether
they are accessed via IP, wireless, etc.) is that the developers
don't address security from the ground up.
E.g., if you design a product that *relies* on that remote access
to perform its normal function (e.g., services in the cloud!),
then you've got a built in vulnerability: if I can deny you
access to that external resource, then you won't be able to
operate properly.
Similarly, if you assume "no one would ever want to hack my device"
(e.g., TOILET) and, thus, never put any thought into protecting
against such hacks, then you'll be chagrined when someone *does*
hack it (as evidenced in recent media reports re: toilets, cars,
baby monitors, door locks, etc.).
When it comes to security, reliability, etc. there's a really
simple solution: enumerate all your assumptions. Then,
challenge *each* of them. Then, *repeat* the exercise.
"/* CAN'T HAPPEN */" usually *does*!
