Re: How can I identify a system uniquely using MFC code
- From: "Ed Weir \(ComCast\)" <Anon@xxxxxxxx>
- Date: Wed, 14 Jun 2006 21:53:36 -0700
"Joseph M. Newcomer" <newcomer@xxxxxxxxxxxx> wrote in message
news:gnl192hh5jcadri296lpgfnksgt4nvtj3c@xxxxxxxxxx
| If my laptop is stolen, there is no security issue, because I make it a
point to never
| keep proprietary data on my laptop. I carry a little external hard drive,
which I keep in
| my pocket, and that's where I keep proprietary information; it is only
plugged into the
| laptop if I'm using the laptop. Otherwise, the two are kept separate.
Fine - somebody steals your drive. Now what?? Smaller than a laptop, I'll
bet.
I use AES-256 in CBC mode and a 4GB USB as backup on my key chain.
Passphrase in head and nowhere else.
| The Internet solution doesn't help if you're trying to use a product at
35,000 feet over
| Nebraska. At least not until airlines start supporting reasonbly-priced
broadband (and I
| hope they don't because of VoIP).
You're going to install software at 35,00 feet? Can't it wait? Maybe you
should do that before you take off. Or - "oops I forgot my dongle, guess I
can't do your books today".
Or how about: "Dog ate my dongle, you'll have to wait for him to shit"...
All symptoms of a poorly designed cryptosystem and ample ignorance.
| The official definition of "Copy Protection" is "A class of methods for
preventing
| incompetent pirates from stealing software and legitimate customers from
using it.
| Considered silly."
|
| Don't worry--NGSCB will save us all!
|
| (and if you believe that the Next Generation Secure Computing Base will
actually make
| systems more secure, you are almost certainly interested in my wonderful
beachfront
| property in New Orleans. Well, it isn't beachfront YET, but it will be
after the next
| hurricane. Buy now!)
That's OK - I already have a large water park south of Florida - keeps
getting alien trespassers in it.
|
| The TCB (Trusted Computing Base) does solve some important problems in
computer security,
| but it will NOT make our computers safer, or prevent viruses, or any
number of other myths
| I've heard people claim will be true (alas, some of them even work for
companies involved
| in trusted computing projects...and I can't say too much about one of them
because I just
| invested two years in a project I can't talk about yet)
| joe
Heh we could be on the same team... but I have seniority by at least three
years
| On Wed, 14 Jun 2006 16:12:39 -0700, "Ed Weir \(ComCast\)" <Anon@xxxxxxxx>
wrote:
|
| >"Joseph M. Newcomer" <newcomer@xxxxxxxxxxxx> wrote in message
| >news:r0ft82drh8qdsk83lii8ltdke71nc42bpd@xxxxxxxxxx
| >| You point out the absurdity of most hand-rolled "copy protection"
schemes.
| >Look how much
| >| more complex your solution is, and I suspect it is just a beginning,
based
| >on some of the
| >| consulting I've done. The actual copy protection means in most
| >proprietary products is
| >| considered secret for obvious reasons, but I've worked with several
| >clients who thought
| >| they'd "solved" the problem on their own ("we don't need to spend money
on
| >some commerical
| >| product!"), and I was able to demonstrate ways of cracking their
schemes
| >within a day (I
| >| used to have an anonymous account that I used to lurk on cracker
| >newsgroups. Anyone who
| >| thinks they can roll their own copy protection scheme should do this
| >first. Discover how
| >| trivial it is to crack ANY software-only scheme. Realize that even the
| >best commercial
| >| software-only products, written by experts in these areas, are highly
| >vulnerable. Then
| >| give up and get a hardware-related product with a dynamic
| >challenge-response mechanism and
| >| high-level encryption, and you MIGHT have a chance of keeping your
| >software secure...)
| >| joe
| >|
| >| On Mon, 12 Jun 2006 20:06:05 -0700, "Ed Weir \(ComCast\)"
<Anon@xxxxxxxx>
| >wrote:
| >|
| >| >"Bruno van Dooren" <bruno_nos_pam_van_dooren@xxxxxxxxxxx> wrote in
| >message
| >| >news:ug75QYgjGHA.3440@xxxxxxxxxxxxxxxxxxxxxxx
| >| >|> I am working on a MFC application. I need to identify
every
| >| >| > system which uses this application. How can I do this using MFC?
Is
| >| >there
| >| >| > any unique identifier for a PC which can be retrieved using MFC?
| >| >|
| >| >| You can find the computer SID in the registry.
| >| >| More details over here:
| >| >| http://www.sysinternals.com/Utilities/NewSid.html
| >| >|
| >| >| --
| >| >|
| >| >| Kind regards,
| >| >| Bruno van Dooren
| >| >| bruno_nos_pam_van_dooren@xxxxxxxxxxx
| >| >| Remove only "_nos_pam"
| >| >
| >| >This use of an SID is woefully inadequate for security, as the site
you
| >have
| >| >referenced illustrates so well; anyone can change the SID of a
computer,
| >so
| >| >how can we expect the data on our hard disk to be secure? This is
just
| >one
| >| >more example of dangerous security assumptions and coding done by the
| >| >'experts' at MSFT.
| >| >
| >| >A more secure method would be to create a one-way sha-256 or AES-256
hash
| >| >of:
| >| >1.) volume ID + SN
| >| >2.) User ID
| >| >3.) User domain
| >| >4.) Machine name
| >| >5.) OSINFO
| >| >6.) User PIN or passphrase
| >| >
| >| >to lock the user to the hardware in use. In the event of a hardware
| >failure
| >| >the last element can be used to recover the ID if necessary. There is
of
| >| >course more to it than this, certain code which needs to be written to
| >| >implement the hash and to later validate it as well as to recover the
ID
| >| >after hardware failure or machine migration.
| >| >
| >| >Point is, the SID is anything BUT secure...
| >| >
| >| >-- Ed.
| >| >
| >| >-----------------------------------------------------
| >| >hex->bin->b64
| >| >F9E7707A2AF502D0A899C6ACB43A2D35EB7E
| >| Joseph M. Newcomer [MVP]
| >| email: newcomer@xxxxxxxxxxxx
| >| Web: http://www.flounder.com
| >| MVP Tips: http://www.flounder.com/mvp_tips.htm
| >
| >You nailed it. Only use the internet connection as your 'hardware', and
a
| >server you control as your 'dongle'. Since most applications are
internet
| >related, this will solve the problem for a good percentage of
applications
| >needing protection against piracy.
| >
| >Secret methods are kept secret in complete ignorance of the principles of
| >Applied Cryptography; you must always assume that an attacker has full
| >knowledge of how your scheme works, and will (quickly) devise an attack
if
| >the target is worth the effort. You are only as secure as the cost of
the
| >attack exceeds the worth of the prize. You have succeeded in attacking
your
| >targets because of the relative ignorance of the victims. Unfortunately
| >(for all of us), not a great feat.
| >
| >The more complete solution would be to:
| >1) Educate every user in secure usage and practices
| >or
| >2) Design systems that are intrinsically secure* without having to
depend
| >on user competence in security
| >
| >E.G.; What if your laptop gets stolen - is your data on it secure from
| >attack?
| >
| >I just got a letter from the Gov't that my service information has been
| >compromised. How did this happen?
| >
| >-- Ed.
| >
| >-----------------------------------------------------
| >hex->bin->b64
| >F9E7707A2AF502D0A899C6ACB43A2D35EB7E
| >
| >* The cost of attack exceeding the value of the prize
| Joseph M. Newcomer [MVP]
| email: newcomer@xxxxxxxxxxxx
| Web: http://www.flounder.com
| MVP Tips: http://www.flounder.com/mvp_tips.htm
.
- Follow-Ups:
- Re: How can I identify a system uniquely using MFC code
- From: Tom Serface
- Re: How can I identify a system uniquely using MFC code
- References:
- How can I identify a system uniquely using MFC code
- From: Sachin Bhave
- Re: How can I identify a system uniquely using MFC code
- From: Bruno van Dooren
- Re: How can I identify a system uniquely using MFC code
- From: Ed Weir \(ComCast\)
- Re: How can I identify a system uniquely using MFC code
- From: Joseph M . Newcomer
- Re: How can I identify a system uniquely using MFC code
- From: Ed Weir \(ComCast\)
- Re: How can I identify a system uniquely using MFC code
- From: Joseph M . Newcomer
- How can I identify a system uniquely using MFC code
- Prev by Date: Re: does VS C++ 2005 actually work????
- Next by Date: Re: multi-thread problem
- Previous by thread: Re: How can I identify a system uniquely using MFC code
- Next by thread: Re: How can I identify a system uniquely using MFC code
- Index(es):
Relevant Pages
|