>>>>> In <26554.1137393402@juniper.net> Mark D. Baushke wrote:

> Hmmm... 

> I am not seeing any difference after I patch the
> gnus-art.el and mm-uu.el files and byte-compile
> them.

In fact I don't know how they work with MH-E + Gnus, but I guess
they might work differently from Gnus.

> The message with the

>    Content-Type: text/plain; charset="us-ascii"; format=flowed

> still shows the various PGP markers:

> '-----BEGIN PGP SIGNED MESSAGE-----', 
> <the message>
> '-----BEGIN PGP SIGNATURE-----' 
> <the base64 signature block>
> '-----END PGP SIGNATURE-----'

> instead of actually recognizing that the message is a
> signed text and should be desplayed as:

>  [[PGP Signed Part:Undecided]]
>   This is a test.

>  [[End of PGP Signed Part]]

> or, after the button has been executed displayed as:

>   [[PGP Signed Part:Mark D Baushke <mdb@gnu.org>]]
>   This is a test.

>   [[End of PGP Signed Part]]

I verified that the patched Gnus 5.10.6 works (it is Gnus, not
MH-E + Gnus).  Could you try the following Lisp form in the
buffer which visits the first (raw) message?

(mm-uu-dissect-text-parts (mm-dissect-buffer nil t))

It is expected that you will see "multipart/signed" in the
result.  If not, I need to try MH-E + Gnus.

> The second message works fine and the third message
> also shows no differences before and after the patch.

To make sure, I attach ``the first message'' that I use: