Re: Evil PGP sigs thread must die. was Re: Stop it with putting your e-mail body in my MUA OT
On Wed, 10 Jul 2002, Jordyn A. Buchanan wrote:
Your facts are correct, but you're missing one so your conclusion is wrong.
You need to verify the signature in order to be able to rely on it. However, if one usually does not consistently sign their messages, then it becomes entirely plausible that a spoofed message lacks a signature not because the forger does not have the capability to generate the signature, but simply because the sender simply neglected to attach a signature (yet again). In this case, unsigned data is accorded roughly the same level of authenticity as signed data.
Yes, but once again you must consider content, given that most mail clients don't automatically verify signatures. Most of us will have to make a judgement call as to whether or not to bother to check the signature. The higher the degree of "importance" of the content, the more likely I am to check the signature, and the more likely I am to take verification steps if not signed. If the content is not "important", I won't bother checking the signature. Lest anybody confuse my argument, I think PGP signatures are a good thing. I just don't think people need to sign everything they send. And I'm talking about posts to Nanog here, not private communication. In private communication, it's reasonable to sign most everything sent with official business purpose. If the majority of mail clients automatically verified pgp signatures, I would be totally in favor of signing every single email. But the simple fact is that not only do most mail clients not support that, many mail clients can't even display the signed text inline! Surely a compromise is needed for now. Andy xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Andy Dills 301-682-9972 Xecunet, LLC www.xecu.net xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx Dialup * Webhosting * E-Commerce * High-Speed Access
On Wed, Jul 10, 2002 at 03:45:41PM -0400, andy@xecu.net said: [snip]
Yes, but once again you must consider content, given that most mail clients don't automatically verify signatures. Most of us will have to
And _therein_ lies the problem. And if the clued among us do not pressure vendors to Do The Right Thing, history has proven they will Do The Profitable Thing instead.
Lest anybody confuse my argument, I think PGP signatures are a good thing. I just don't think people need to sign everything they send. And I'm talking about posts to Nanog here, not private communication. In private communication, it's reasonable to sign most everything sent with official business purpose.
As I mentioned before - sign everything. It's just a good habit to be in, and there are no good arguments against it (except for the preceding one about MUAs not supporting PGP commands, and _that_ can be changed with pressure from those purchasing the software. Which (surprise!) many of us are either in charge of, or have influence over.)
If the majority of mail clients automatically verified pgp signatures, I would be totally in favor of signing every single email. But the simple
So _help create that majority_. Sitting around complaining on NANOG that support doesn't exist will not improve the situation. Let your vendor know that this is an important^W^Wa critical feature for you.
fact is that not only do most mail clients not support that, many mail clients can't even display the signed text inline! Surely a compromise is needed for now.
As the mutt homepage says: All mail clients suck. This one sucks less. If we want things to Not Suck, the only way it's going to happen is if we put some effort into making it so. -- -= Scott Francis || darkuncle (at) darkuncle (dot) net =- GPG key CB33CCA7 has been revoked; I am now 5537F527 illum oportet crescere me autem minui
Thus spake "Andy Dills" <andy@xecu.net>
Yes, but once again you must consider content, given that most mail clients don't automatically verify signatures. Most of us will have to make a judgement call as to whether or not to bother to check the signature.
The higher the degree of "importance" of the content, the more likely I am to check the signature, and the more likely I am to take verification steps if not signed.
If the content is not "important", I won't bother checking the signature.
Why not just upgrade to a modern MUA and not have to worry? OE only supports S/MIME for now, but it does automatically verify every message, including checking that the From: line matches the key. It makes a big stink if the signature doesn't match, but just displays a simple little icon if it's verified correctly. How can you prefer to check messages manually and therefore cause the problems you describe?
Lest anybody confuse my argument, I think PGP signatures are a good thing. I just don't think people need to sign everything they send. And I'm talking about posts to Nanog here, not private communication. In private communication, it's reasonable to sign most everything sent with official business purpose.
Ironically, there's no need to sign intrabusiness email because it's trackable by trusted authorities and therefore implicitly trusted for non-legal matters. It's personal email that needs a trust mechanism.
If the majority of mail clients automatically verified pgp signatures, I would be totally in favor of signing every single email. But the simple fact is that not only do most mail clients not support that, many mail clients can't even display the signed text inline! Surely a compromise is needed for now.
Sure. Use old-style signatures if you're going to sign every message, and we can transition to new-style signatures once most people upgrade. S
On 01:30 PM 7/10/02, Stephen Sprunk wrote:
Thus spake "Andy Dills" <andy@xecu.net>
Yes, but once again you must consider content, given that most mail clients don't automatically verify signatures. Most of us will have to make a judgement call as to whether or not to bother to check the signature.
The higher the degree of "importance" of the content, the more likely I am to check the signature, and the more likely I am to take verification steps if not signed.
If the content is not "important", I won't bother checking the signature.
Why not just upgrade to a modern MUA and not have to worry?
OE only supports S/MIME for now, but it does automatically verify every
including checking that the From: line matches the key. It makes a big stink if the signature doesn't match, but just displays a simple little icon if it's verified correctly. How can you prefer to check messages manually and
message, therefore
cause the problems you describe?
Maybe the following has something to do with his desire to avoid using Micro$oft products as you advocate, *especially* as regards to letting a M$ email client automatically process PGP .sigs on incoming email. jc "the timing couldn't have been better" dill <http://www.eeye.com/html/Research/Advisories/AD20020710.html> Remote PGP Outlook Encryption Plug-in Vulnerability Release Date: July 10, 2002 Severity: High (Remote Code Execution) Systems Affected: NAI PGP Desktop Security 7.0.4 NAI PGP Personal Security 7.0.3 NAI PGP Freeware 7.0.3 Description: The beer is still cold, the days are still long, the exploits still start as jokes (this time over a beer with a three letter agency) and the advisories... we'll just say, "All of your SCADA are belong to us." A vulnerability in the NAI PGP Outlook plug-in can be exploited to remotely execute code on any system that uses the NAI PGP Outlook plug-in's. By sending a carefully crafted email the message decoding functionality can be manipulated to overwrite various heap structures pertinent to the PGP plug-in. This vulnerability can be exploited by a user simply selecting a "malicious" email, the opening of attachments is not required. When the attack is performed against a target system, malicious code will be executed within the context of the user receiving the email. This can lead to the compromise of the targets machine, as well as their PGP encrypted communications. It should also be noted that because of the nature of the SMTP protocol this vulnerability can be exploited anonymously. Technical Description: Exploitation: By creating a malformed email we can overwrite a section of heap memory that contains various data. By overwriting this section of heap with valid addresses of an unused section in the PEB, which is the same across all NT systems, we can walk the email parsing and eventually get to something easily exploitable: CALL DWORD PTR [ecx] This pointer addresses references a function pointer list. At the time of exploitation, an attacker controlled buffer address is the first item on the stack. By overwriting the function pointer list pointer address with the address of an Import table, we can call any imported function. Our current stack will be passed into the function for parameter use. as is. The first item on our stack is an address that points to attacker-controlled data. By overwriting the address, with the address of the SetUnhandledExceptionFilter() IAT entry, execution will redirect into this address when the default exception handler is called, After returning from SetUnhandledExceptionFilter() PGP Outlook will fail as it crawls back down the call stack, after cycling through the exception list it will call the DefaultExceptionFilter, which now contains the address of our code. This of course can also be exploited silently using frame reconstruction. Due to the large size of an example vulnerable email we are not including it in our advisory. We will be updating the research section of our website with a link to an example email. http://www.eEye.com Where do you want your secret key to go today? Vendor Status: NAI has worked quickly to safeguard customers against this vulnerability. They have released a patch, for the latest versions of the PGP Outlook plug-in, to protect systems from this flaw. You may download the patch from: http://www.nai.com/naicommon/download/upgrade/patches/patch-pgphotfix.asp Note: This issue does not affect PGP Corporate Desktop users. Discover: Marc Maiffret Exploitation: Riley Hassell Greetings: Kasia, and the hot photographer from Inc Magazine. Phil Zimmerman, the godfather of personal privacy, much respect. Copyright (c) 1998-2002 eEye Digital Security Permission is hereby granted for the redistribution of this alert electronically. It is not to be edited in any way without express consent of eEye. If you wish to reprint the whole or any part of this alert in any other medium excluding electronic medium, please e-mail alert@eEye.com for permission. Disclaimer The information within this paper may change without notice. Use of this information constitutes acceptance for use in an AS IS condition. There are NO warranties with regard to this information. In no event shall the author be liable for any damages whatsoever arising out of or in connection with the use or spread of this information. Any use of this information is at the user's own risk. Feedback Please send suggestions, updates, and comments to: eEye Digital Security http://www.eEye.com info@eEye.com
At 3:45 PM -0400 2002/07/10, Andy Dills wrote:
Lest anybody confuse my argument, I think PGP signatures are a good thing. I just don't think people need to sign everything they send. And I'm talking about posts to Nanog here, not private communication. In private communication, it's reasonable to sign most everything sent with official business purpose.
No. It is precisely the public e-mail messages which should always be signed, since they are the ones likely to reach the largest audience, and the ones that are likely to have the biggest negative impact if they are successfully spoofed. You should sign all private e-mail, too, but the public e-mail messages are the ones that need it the most. -- Brad Knowles, <brad.knowles@skynet.be> "They that can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety." -Benjamin Franklin, Historical Review of Pennsylvania.
On Mon, Jul 15, 2002 at 05:35:01PM +0200, Brad Knowles wrote:
No. It is precisely the public e-mail messages which should always be signed, since they are the ones likely to reach the largest audience, and the ones that are likely to have the biggest negative impact if they are successfully spoofed.
what is amazing is that if this was really that important.. how is it that mankind has gotten this far without the need of authenticating every piece of communication. but what this all has to do with nanog, i don't understand at all.
participants (6)
-
Andy Dills -
Brad Knowles -
Christian Kuhtz -
JC Dill -
Scott Francis -
Stephen Sprunk