Re: Useful factoid

2011-10-13 Thread Robert J. Hansen
On 10/11/2011 05:14 PM, Jean-David Beyer wrote:
 Let us assume you are the bad guy

Okay.

 Unless you have my encrypted keys, you have to access my computer
 (unless you have already stolen it, in which case there are much
 easier ways to invade the machine), you will have to try logging in
 through the Internet (in the case of my machine), and the first thing
 you will hit is the login program.

Hold on a second there.  You seem to be making some extremely
unwarranted assumptions.

If I want your secret key material, I'm not going to steal your
computer.  I'm going to use an exploit to bypass your login, plant a
Trojaned version of GnuPG, and laugh all the way to the bank.

Modern-day operating systems are frightening -- terrifyingly --
insecure.  A while ago Vint Cerf estimated that about one desktop PC in
five was already pwn3d.  That's a number that keeps me awake at night.

___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-13 Thread Jerome Baum
 Hold on a second there.  You seem to be making some extremely 
 unwarranted assumptions.

Take a look:

 Unless you have my encrypted keys, you have to access my computer 
 (unless you have already stolen it, in which case there are much 
 easier ways to invade the machine), you will have to try logging in
 through the Internet (in the case of my machine),

 If I want your secret key material, I'm not going to steal your 
 computer.

-- 
PGP: A0E4 B2D4 94E6 20EE 85BA E45B 63E4 2BD8 C58C 753A
PGP: 2C23 EBFF DF1A 840D 2351 F5F5 F25B A03F 2152 36DA

___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-13 Thread Robert J. Hansen
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256

On 10/13/11 7:51 AM, Jerome Baum wrote:
 Take a look:

I did.  You said I have to access your computer, to try logging in
through the Internet.  I don't.  I just have to find an exploit.

Saying my front door is locked is great, but it's not so great when
you consider a good thief knows how to pick locks.  Against that kind
of adversary a lock isn't much of a prevention device: at best it
delays the thief by a minute.
-BEGIN PGP SIGNATURE-
Comment: GPGTools - http://gpgtools.org

iFYEAREIAAYFAk6W1jQACgkQI4Br5da5jhCe5ADfalUs2A9esxyUzm5PSR7jHtxA
X070BaWnnBbdxwDffCSbpoq3miWq1SPMnqZ7TLaodIZTveFA6ez4dYkBHAQBAQgA
BgUCTpbWNAAKCRAHLcdEK4m9RVdTB/9VHB187mWBkW1z9XLvLyp5/aN9mv4x8nsY
XQyMwNCRfzpseplMaEMT7JFPsTH87xxzhhu853ebdMajN+QnPeva0ipaLwCeNXGu
Y3DDJ0EgtD1Hw4CQyMeRCHI8OPuJRCgzfaVul9KFUvcUmgSDUp3DNCrPUzR+GCQZ
WqDfvMlaBZjnqexwBxhWJtY3mfpm463qSgMRHuidpx7BBxe0K/jLc1AC8f2ytCCn
DXmmEQt7AHIZq6jQbZ9sZAqhRR0ceu7qmXefvL6lTXealopyARgnlyU7PSHC0p0M
yjwTynL85BnC89eUdlIWmVU7/FovpnLqg8lPj38jovAdsJ5Vtrn7
=/Uvi
-END PGP SIGNATURE-

___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-13 Thread Jerome Baum
On 2011-10-13 14:14, Robert J. Hansen wrote:
 On 10/13/11 7:51 AM, Jerome Baum wrote:
 Take a look:
 
 I did.  You said I have to access your computer, to try logging in
 through the Internet.  I don't.  I just have to find an exploit.

I didn't say anything (modulo Take a look).

 Saying my front door is locked is great, but it's not so great when
 you consider a good thief knows how to pick locks.  Against that kind
 of adversary a lock isn't much of a prevention device: at best it
 delays the thief by a minute.

You have to access my computer would be you have to enter my house.
Nobody ever said you have to enter my house via the front door.

Also, a thief that picks my front door would be someone who brute-forces
my login (assuming the front door is my login). You probably meant a
thief who just smashes a window or climbs through one that is open.

-- 
PGP: A0E4 B2D4 94E6 20EE 85BA E45B 63E4 2BD8 C58C 753A
PGP: 2C23 EBFF DF1A 840D 2351 F5F5 F25B A03F 2152 36DA

___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-13 Thread Robert J. Hansen
On 10/13/2011 8:29 AM, Jerome Baum wrote:
 I didn't say anything (modulo Take a look).

At this point it seems to me you're being deliberately obtuse.  Have a
nice day.

___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-13 Thread Melvin Carvalho
On 11 October 2011 22:32, Robert J. Hansen r...@sixdemonbag.org wrote:
 Accurate to 6%, there are 2**25 seconds in a year.  Worth remembering:
 it makes certain kinds of computations much easier.  (It follows there
 would be about 2**35 seconds in a thousand years, or 2**45 seconds in a
 million.)

 E.g., let's say you want to brute-force an 64-bit key on a CPU that can
 do a million (2**20) attempts per second.  This requires, on average,
 2**63 attempts.  2**63 / 2**20 = 2**43 seconds: 2**43 / 2**45 = 2**-2 =
 a quarter of a million years.

 I don't know why it took me so long to notice that: seems like the sort
 of thing I should've noticed a decade ago.  It makes certain kinds of
 computations so much easier.

 Anyway, figured I'd throw it out on the off chance there were others who
 hadn't noticed it.

I used to think of there being roughly 2^32 seconds in a lifetime :)


 ___
 Gnupg-users mailing list
 Gnupg-users@gnupg.org
 http://lists.gnupg.org/mailman/listinfo/gnupg-users


___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-13 Thread Jean-David Beyer
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Robert J. Hansen wrote:
 On 10/11/2011 05:14 PM, Jean-David Beyer wrote:
 Let us assume you are the bad guy
 
 Okay.
 
 Unless you have my encrypted keys, you have to access my computer 
 (unless you have already stolen it, in which case there are much 
 easier ways to invade the machine), you will have to try logging in
  through the Internet (in the case of my machine), and the first
 thing you will hit is the login program.
 
 Hold on a second there.  You seem to be making some extremely 
 unwarranted assumptions.

Quite possibly. And unwarranted assumptions are especially pernicious
because those are typically those I am unaware of making.

I am not a security expert anymore. I really was never a security
expert, though I was once put in charge of security for 10 VAX machines
running UNIX, but this was around 30 years ago almost before the
Internet. Some of us were using uucp on dialup, but that was about it.
In those days it was almost impossible to get the users to use passwords
on their accounts.
 
 If I want your secret key material, I'm not going to steal your 
 computer.  I'm going to use an exploit to bypass your login, plant a 
 Trojaned version of GnuPG, and laugh all the way to the bank.

I realize if you stole my computer that I would notice it.
If you broke into my house skillfully enough that I did not notice it,
you could install a key logger, or copy my hard drives, steal my backup
tapes, ... . But you could also remove all protections by getting in as
the root user (on UNIX-Linux). And I might not notice that.

The trick is to do that from the Internet. I have some safeguards to
protect me, and they may protect me from amateurs, but an expert might
be able to defeat me.

It seems to me that to do much damage to my machine, you need to get a
shell with root access. And to do that, do you not pretty much need the
root password? Or hijack a program that is currently running with the
root privileges?

I never run a web browser as root. But there are demons that run and
some have root privileges. Such as the download mechanism to download
updates from Red Hat. My nameserver does not run as root. I do not run
telnet. ssh will talk only to specified IP addresses on my LAN. My
firewall will not accept messages from outside unless in reply to
something I sent out, so I believe it would take a man-in-the-middle
attack to get past that unless the firewall is defective. I actually
have two firewalls; a primitive one in the router that comes with
Verizon's FiOS service, and another one using iptables. These, too,
could have bugs, especially if I made a mistake in programming the
iptables firewall.
 
 Modern-day operating systems are frightening -- terrifyingly -- 
 insecure.  A while ago Vint Cerf estimated that about one desktop PC
 in five was already pwn3d.  That's a number that keeps me awake at
 night.
 
At one extreme, the only way to be pretty safe is to have a machine that
is not connected to the Internet, and have U.S.Marines to guard the
hardware and access to it. I do not choose to defend myself against
threats that would reasonably require that. I want my security to be
weak enough that the black hats would not resort to torture to get the
information they want.

The friends of mine that even know what computer security might mean do
not even encrypt their e-mails, though they worry about it's being
intercepted. Friends complain if I digitally sign my e-mails. I assume
if they could accept encrypted e-mails, that they would save them in
clear form on their machines anyway. So maybe I am kidding myself.

I do not think my machine has been taken over. For one thing, I can
pretty much see the Internet traffic from it, and when I am not doing
anything, not much goes down the Internet. A friend whose machine was
hacked (Windows ME) had lots of Internet traffic and the machine got
impossibly slow. The hard drives never stopped clicking. I do not have
that, though the hard drives on this machine do not click, but the
Xosview program shows that when nothing is going on, nothing except
BOINC programs run. The demons do, but they do not use any processor time.

If I ran this machine as a server, my problems would surely be worse.

- --
  .~.  Jean-David Beyer  Registered Linux User 85642.
  /V\  PGP-Key: 9A2FC99A Registered Machine   241939.
 /( )\ Shrewsbury, New Jerseyhttp://counter.li.org
 ^^-^^ 08:50:01 up 6 days, 17:23, 4 users, load average: 5.14, 4.93, 4.94
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.5 (GNU/Linux)
Comment: Using GnuPG with CentOS - http://enigmail.mozdev.org/

iD8DBQFOlu/MPtu2XpovyZoRArvUAKC022RLKvUmsbM1XD5shR+xrB06kQCdEDE+
gx/6aDndO7obVhfgZVEMk6o=
=yjMn
-END PGP SIGNATURE-

___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-13 Thread Robert J. Hansen
On 10/13/11 10:03 AM, Jean-David Beyer wrote:
 It seems to me that to do much damage to my machine, you need to get a
 shell with root access. And to do that, do you not pretty much need the
 root password?

Nope.  Local exploits are enough.

Take a look at the kernel.org exploit as an example.  The current belief
is that one of kernel.org's legitimate users was sshing in from a
compromised box.  That compromised box was running a keylogger.  From
that keylogger, the attacker discovered this user's login name and ssh
credentials.  The attacker then logged into kernel.org as this user and
ran a local exploit to gain root access.  The attacker dropped a
rootkit, a Trojaned ssh/sshd that was harvesting passwords, and all
other kinds of goodness.

Then, since one of the users on my box sshed in from kernel.org, the
attacker got a login credential on my box.  The attacker logged in using
this stolen credential, used a local exploit, and the next thing I know
sixdemonbag.org was rooted.

As you can guess, I'm not talking about some abstract theory here.  This
was a real attack that really compromised my web server.

People tend to grossly underestimate the risks of malware and pwnage.
We talk about it very little to almost none at all, and honestly, I
think it's the eight hundred pound gorilla in the room that everybody is
trying very hard not to notice in the hopes that if we just pretend not
to see it that it will go away.

___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-13 Thread Michel Messerschmidt
On Thu, Oct 13, 2011 at 10:03:56AM -0400, Jean-David Beyer wrote:
 It seems to me that to do much damage to my machine, you need to get a
 shell with root access. 

Depends on what you regard as damage.
Do you need root privileges to use your private gpg keys ???


 I never run a web browser as root. 

If you run your web browser under the same account that you use for
gpg, vulnerabilities in your browser *potentially* allow an attacker 
to access your private keys.

Same is true for your mail program, PDF reader, messaging client, ...



signature.asc
Description: Digital signature
___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-11 Thread Jean-David Beyer
Robert J. Hansen wrote:
 Accurate to 6%, there are 2**25 seconds in a year.  Worth remembering:
 it makes certain kinds of computations much easier.  (It follows there
 would be about 2**35 seconds in a thousand years, or 2**45 seconds in a
 million.)
 
 E.g., let's say you want to brute-force an 64-bit key on a CPU that can
 do a million (2**20) attempts per second.  This requires, on average,
 2**63 attempts.  2**63 / 2**20 = 2**43 seconds: 2**43 / 2**45 = 2**-2 =
 a quarter of a million years.

Let us assume you are the bad guy and have computing power that can do
an arbitrarily large number of key attempts per second. Unless you have
my encrypted keys, you have to access my computer (unless you have
already stolen it, in which case there are much easier ways to invade
the machine), you will have to try logging in through the Internet (in
the case of my machine), and the first thing you will hit is the login
program. This can probably handle only a few attempts per second, and if
I were serious about security, I would have it double the time to reply
each time it got a failed login on that connection. In the days of
dialup, I would have the machine hang up on the connection with too many
failed login attempts.

Of course, if you could get into my machine and login as the only user
with access to my encrypted password file, you could copy that file to
your high speed facility and crack it at your leisure. But if you could
do that, you could already do anything you wanted with my machine --
install trojan horse keyloggers, defeat the security in the login
program, etc.

 
 I don't know why it took me so long to notice that: seems like the sort
 of thing I should've noticed a decade ago.  It makes certain kinds of
 computations so much easier.
 
 Anyway, figured I'd throw it out on the off chance there were others who
 hadn't noticed it.



-- 
  .~.  Jean-David Beyer  Registered Linux User 85642.
  /V\  PGP-Key: 9A2FC99A Registered Machine   241939.
 /( )\ Shrewsbury, New Jerseyhttp://counter.li.org
 ^^-^^ 17:05:02 up 5 days, 1:38, 4 users, load average: 4.73, 4.76, 4.82

___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users


Re: Useful factoid

2011-10-11 Thread MFPA
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA512

Hi


On Tuesday 11 October 2011 at 9:32:18 PM, in
mid:4e94a7d2.7060...@sixdemonbag.org, Robert J. Hansen wrote:


 Accurate to 6%, there are 2**25 seconds in a year.


[...]

 I don't know why it took me so long to notice that:
 seems like the sort of thing I should've noticed a
 decade ago.

I suppose you didn't need to notice it because you already remembered
pi seconds in a nano-century

- --
Best regards

MFPAmailto:expires2...@ymail.com

A nod is as good as a wink to a blind bat!
-BEGIN PGP SIGNATURE-

iQCVAwUBTpTlNaipC46tDG5pAQo8NgP/f/etxoSVmn5rhWCc/mUxaoO4U4HD/9TB
snAV8qD1mZU2dzvkzrlZXMlIgr3pYzEXTImSGfsmjBLH90Q/hGdvAvlC2smW8Ezw
Net+bV/vw6r8TFKbwoF7ubIK4/27A3bSoq3up5t0PrEK2dOGIpTYnPgfEY5pIfe/
jz1JYCPJNhE=
=/wzd
-END PGP SIGNATURE-


___
Gnupg-users mailing list
Gnupg-users@gnupg.org
http://lists.gnupg.org/mailman/listinfo/gnupg-users