[E-voting] Re: FC: Electronic voting in
jwkckid1 at ix.netcom.com
Fri Jun 6 06:14:21 IST 2003
Adam and all,
Greatpp comments and remarks here Adam. I have some remarks
to add to yours below.
adam beecher wrote:
> > How many desktop computers were infected by ILY - a trojan developed
> > by individuals with no funding? How many eggs are in one basket wrt
> > OS software and application software? How easy is it to get arbitrary
> > code running on a large number of desktop systems? How easy is it to
> > hide that code so it doesn't do anything stupid 'til the appropriate
> > time? How easy is it to mask the origin of that code?
> Aren't these questions proving my point?
Indeed we have come full circle. But hiding or masking code is
sometimes done it is not the norm. It can be found, and masking
is not usually difficult to unmask. What is needed is good
independent code review. As a long time develoer/code writer
myself for various security systems I know this to work and work
well and also to be true.
> Haven't we come full circle, back
> to mindset? I mean, sure, you can look at this from a purely technological
> point of view, you can blame the software and Microsoft's OS monopoly, but
> how do you propose to fix that? You /can/ fix it, of course. You can lock
> down development processes, clean-room code, audit it line-by-line, try to
> underrun and overrun the buffers, etc. You won't catch absolutely
> everything, but you should be able to get five nines.
> But what happens when your software leaves the clean-room? What will the OEM
> do to it? What about the user? Ok, so we have to DRM our code to make sure
> it can't be hax0red. Lock it down. And lock down the operating system too,
> so the hax0rs can't install reverse-engineering tools and the like. Load the
> OS into ROM, so it can't be modified. DRM it. Lock down the hardware, put it
> in a sealed titanium box. No USB ports, we can't trust the user. No DVD or
> CD ROM's. Hardwire the monitor. Forget the snazzy soundcard, you'll take
> onboard sound and like it.
> Forget about Linux, and FreeBSD, and all the other wonderful OS's that have
> been thrown at us for free. Forget about the tools at the very core of our
> network: Apache, sendmail, and bind. PHP and Perl and Python are all gone.
> Qmail and Exim, procmail and SpamAssassin, emacs and vi, MySQL and Postgres,
> all dead. You can't have them. I'm not exaggerating here, that's what you
> have to do if you want a secure system, and /even then/ there's an element
> of risk, because you didn't throw one particular string at the code in the
> clean-room; and oops, you've got an underrun, j00r 0wn3d. Unlikely, but
> Do we really want that flavour of security, everyone with the same box, the
> same software? And it won't be open source or free software mind, it'll be
> proprietry, because as soon as you lock those systems down, you lock open
> source and free software developers out of their trade. Bill will own you.
> And of course that introduces further risk, because you've got one entity
> holding all the keys. Again though, that's not a technological issue, that's
> a societal issue. It's not technology owning you, it's the guy that owns the
> Again, I'm not saying we /should/ do remote electronic voting now, I'm
> saying we can. Technologically, it's possible. It's our current society that
> prevents it. We need to change the way society perceives electronic
> security. We need to get "journalists" to stop using the work "hacker" when
> they mean "cracker". We need people to understand that security is about
> perception, about paranoia, about assuming the worst. We need them to
> understand that when their computer is broken into, 90% of the blame lies
> with them, not the people that wrote the software. THEY opened the email
> attachment, not Microsoft.
> > > There are technological ways around this, for example by
> > > automatically closing - maybe evening uninstalling - the voting
> > > application as soon as someone has voted.
> > And this does what?
> I would have thought it was obvious: It prevents people proving who they
> voted for.
> > Or they set up an unofficial polling station in some CyberCafes, or
> You have the same difficulties, as this would be fraud on a grand scale. You
> don't get a couple of hundred people gathering in one spot to be threatened
> or cajoled out of their vote without someone talking about it.
> > they set up a mechanism to transfer voting credentials so that they
> > can act as proxy voters. Not all of the attacks here assume a hostile
> > target. We are not just talking about vote rigging, but also
> > vote selling.
> Voting credentials here are keys though, and again, this is a societal
> thing. You don't give your front door key to every Tom, Dick and Harry that
> asks for them or threatens you for them. We need to give people keys and
> teach them how to use them, teach them that they're the electronic
> equivalent of house keys, which need to be kept safe and secure. And sure,
> houses do get robbed occasionally, but to have a significant affect on an
> election, you need access to thousands, tens of thousands of votes. Again,
> we're talk and risk and effort, and society and mindsets.
> > > This one is the most troublesome for me, although I think that
> > > a properly implemented system with trusted oversight shouldn't
> > > need second channel
> > > verification.
> > Do you know what all of the software on your computer does? I don't.
> Depends on your definition of "does" in this case. If you mean, do I know
> exactly what the bits and bytes on my rig do when I open an app or change
> something, then no, of course not; I'd imagine there are only a few hundred
> people in the world that run systems they can understand from the On button
> up. If your meaning is more basic though, then yes, of course I do.
> Everybody should.
> I can see where you're going with this, but to be perfectly frank I'm not
> going to answer an unasked question.
> E-voting mailing list
> E-voting at lists.stdlib.net
Jeffrey A. Williams
Spokesman for INEGroup LLA. - (Over 131k members/stakeholders strong!)
"Be precise in the use of words and expect precision from others" -
CEO/DIR. Internet Network Eng. SR. Eng. Network data security
Information Network Eng. Group. INEG. INC.
E-Mail jwkckid1 at ix.netcom.com
Contact Number: 214-244-4827 or 214-244-3801
More information about the E-voting