(My original reply went just to Philipp. My apologies.)
On 7/24/2013 1:53 AM, Philipp Klaus Krause wrote: > Maybe having a look at the following scenario will help: Unfortunately, this is not casting very much light on things. The use of phrases like CONFIDENTIAL, SECRET and TOP SECRET have very specific meanings in NATO countries, and you're using them here in ways that are at odds with their NATO meanings. Let me try this rephrasing: You have three machines: Fry, Leela and Bender. Fry is your smartphone, Leela is your desktop and Bender is your "secure" desktop. Email to [email protected] goes to ... what, all three of them? Email to [email protected] goes to Leela and Bender, and [email protected] goes only to Bender. Further, each piece of traffic can receive any of three classifications: C, S or TS. You can send C traffic to Bender: the necessary keys to decrypt it are held there. However, although you can technically send TS traffic to Fry, Fry can't decrypt it: the keys aren't there. If I have this right, then you've walked straight into the Bell-LaPadula security model. You'll be well-served by reading up on it: a good academic reference will answer many of your questions. The short answer is, "OpenPGP by itself will not be sufficient for your purposes. It might be able to provide a couple of tools, but what you want to achieve is far beyond the scope of OpenPGP." _______________________________________________ Gnupg-users mailing list [email protected] http://lists.gnupg.org/mailman/listinfo/gnupg-users
