Re: Outstanding question - rule on cleartext signing last line

[David Shaw <[e-mail]>]

On Mon, Dec 26, 2005 at 04:20:28PM +0000, Ian G wrote:
> David Shaw wrote:
> 
> > Can you give an example of why you feel this is a bug?
> 
> I think it is a bug if we decide that the
> process of cleartext signing is reversable.

Can you please give an example of how the current system is not
reversible?

As far as I can see, the current system is quite reversible.  For
example, given a document reading "this is a test" (no line ending,
and the last character in the file is the second t from test), here's
a clear signature:

  -----BEGIN PGP SIGNED MESSAGE-----
  Hash: SHA1
  
  test  <<-----  There is a CRLF here, but it is NOT PART OF THE SIG
  -----BEGIN PGP SIGNATURE-----

  blah blah blah
  -----END PGP SIGNATURE-----

Now given a document reading "this is a test<LINEENDING>", here's a
clear signature:

  -----BEGIN PGP SIGNED MESSAGE-----
  Hash: SHA1
 
  test  <<----- There is a CRLF here and it IS PART OF THE SIG
  <<-----  There is a CRLF here, but it is NOT PART OF THE SIG
  -----BEGIN PGP SIGNATURE-----

  blah blah blah
  -----END PGP SIGNATURE-----

In what way is this not reversible?

> So, if the process of verification creates
> a file that is the pre-signed version, then
> it has to decide whether the final CRLF has
> to be preserved or stripped in the unsigned
> output.

The final CRLF is not part of the document.  If a user/implementation
wants a final CRLF in there that is part of the document, they need to
add one.

Think of the "BEGIN PGP SIGNATURE" string as actually being
"CRLF-----BEGIN PGP SIGNATURE".  It's part of the message structure
and not part of the signed text.

David