Comments on new examples from Getronics

["Blake Ramsdell" <blake@xxxxxxxxxxxxxxxxxx>]

John, thank you for providing the revised examples.  I have some comments:

In my testing, example 5.6 appears to have switched the order of the
SignerInfos.  I do not have access to the examples draft, so I don't know
what the spirit of this test was.  It is most likely not an issue.


I have the following comments about 5.8.eml and 5.9.eml.  Most of them are
MIME formatting issues, and will not affect the CMS portion of the messages.

* Both files are missing the MIME-Version header.  I believe that the spirit
of these files is that they are valid MIME messages, and thus require this
header.

* The timezone for 5.9 doesn't match the comment (-0300 and Eastern Standard
Time).  This is not required for correct operation, however, and is not
critical to fix.

* The Content-Type header for 5.9 is not correct:

Content-Type: Application/pkcs7-mime;name=smime.p7m;filename=smime.p7m;
        micalg=SHA-1;
        protocol=application/pkcs7-signature

I believe it should be replaced with:

Content-Type: application/pkcs7-mime;name=smime.p7m
Content-Disposition: attachment; filename=smime.p7m

Which fixes a few issues, including the inclusion of the micalg and protocol
parameter, as well as the "name" and "filename" parameter specified on the
Content-Type.  The capitalized A in Application/pkcs7-mime is not an error,
but I made it lowercase just for the heck of it.

* I am still having some issues with ExContent.bin matching data in MIME
messages.  5.8.eml now contains the following first part of the
multipart/signed:


Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 7bit


This is some sample content.


Even after removing the text/plain MIME wrapping for this content, there is
now an extra CRLF at the beginning of the data, which fails the data
comparison with ExContent.bin.  I may be misunderstanding the spirit of what
we are going to be using ExContent.bin for.

Blake
--
Blake Ramsdell
Brute Squad Labs http://www.brutesquadlabs.com