Problem with faked-system-time option

[David Shaw]

On Jun 15, 2011, at 6:02 PM, Jerome Baum wrote:

>> Out of curiosity, as long as we're talking about things that current code will reject, does the 0x50 signature meet the semantics desired here?  This all sounds vaguely notary-like ("I saw this document on such-and-such date") to me, and the intent of 0x50 is a notary signature.  The nice thing about a 0x50 signature is that it is a signature on a signature, so the timestamp service doesn't need to see the document - just the (detached) signature.
> 
> My understanding of a notary's job would include "I trust this key to
> be valid, in possession only of the person named in the uid, while
> that person was in sufficient mental state, not being threatened at
> gun-point, ..."

The 0x50 signature should not be interpreted as the output of a real-world notary (whose task varies in different locations anyway).  It is merely analogous to a notary in that the "notary" sees a signature, and affixes a seal to it indicating "I saw this" (oversimplification, but forgive me).

OpenPGP calls this signature a "Third-Party Confirmation signature".  It is merely a signature on a signature for whatever purpose is desired by the signer.

> -- why should we use a signature type that could be
> misinterpreted, when there is a "timestamp" signature type that fits
> our needs exactly?

Because as already noted, the 0x40 signature is not fully specified in the standard.  There is not enough information to know how to generate one.

David