This is an explanation such that less advanced users can follow, if they are, like I was, eager to learn how to verify emails from web.
( Just if you see any sigs when you follow below, do the:
gpg --recv-key XXXXXXXX
)
If you've been subscribed to Muut, you can follow this.
Go to you mutt maildir. Do not open any messages.
Hit '/' . You're in the command track in bottom. And now type (or paste
this):
=h 20160921191202.GB18462
and hit Enter.
The first find is the message that Claus Assmann used as example. (We only
need the message where that number is part of Message-ID string, not where it
is part of In-Reply-To or References string.)
Enter to open the message. It shows (to me and others) as having BAD
signature. Notice how it turns to be for you.
I have maildir, and if I would do the below, it would create a small maildir
folder. So I first have to do:
touch ClausAssmann.eml
(I named it after the poster of the proof-of-concept non-verbose little tar
archive at:
http://marc.info/?l=mutt-users&m=147448664132307&w=2
but for clarity I'll attach the same file here as ClausAssmann.tar, because I
can thank Claus (and Ian Zimmermann in this other email
http://marc.info/?l=mutt-users&m=147432705018043&w=2
) that I now know myself how to verify mails from web. And I want to
share it with less advanced than I am.
Untar that file ClausAssmann.tar. The result must be:
$ ls -l ClausAssmann.d
total 8
-rw------- 1 miro miro 1422 2016-09-21 21:31 m4
-rw------- 1 miro miro 824 2016-09-21 21:31 m4.sig
$
)
With the mail we just found opened in Mutt, type:
C
and save it to (the touched) ClausAssmann.eml .
Now you can run this command:
grep -A80 'Content-Type: text\/plain; charset=utf-8' ClausAssmann.eml \
| head -36 > ClausAssmann.eml.1
You're close to verifing that email. But not there yet, as I purposefully left
the incriminating header (culprit my_hdr of Mutt being buggy as Claus Assmann
stated in this mail:
http://marc.info/?l=mutt-users&m=147458978829311&w=2
)
If you now view the diff of the two files with, they're entirely different.
But if you view them with vimdiff there's only the X-Clacks-Overhead line the
difference...
Rerun the slightly modified command:
grep -A80 'Content-Type: text\/plain; charset=utf-8' ClausAssmann.eml \
| grep -v 'X-Clacks-Overhead' \
| head -35 > ClausAssmann.eml.2
Now still entirely different with diff, and completely same with vimdiff.
Because it's the end of line is the difference! Solely!
I can convert, the one that I saved, with Vim:
vi ClausAssmann.eml.2
:se ff=dos
:wq
Now I can verify that email, either way, because both the incomplete parts of
the raw email (the part that PGP verifies) are exactly the same:
$ ls -l ClausAssmann.eml.2 ClausAssmann.d/m4
-rw------- 1 miro miro 1422 2016-09-21 21:31 ClausAssmann.d/m4
-rw-r--r-- 1 miro miro 1422 2016-09-23 15:00 ClausAssmann.eml.2
$ sha256sum ClausAssmann.eml.2 ClausAssmann.d/m4
f86f0c710b7b7feaaca11511e66c1335b35bb7c8e18c47208702fb603c393c02
ClausAssmann.eml.2
f86f0c710b7b7feaaca11511e66c1335b35bb7c8e18c47208702fb603c393c02
ClausAssmann.d/m4
$
Any of these two verify correctly:
gpg --verify ClausAssmann.d/m4.sig ClausAssmann.eml.2
gpg --verify ClausAssmann.d/m4.sig ClausAssmann.d/m4
with signature good.
---
Anyway, folks, it really would be time for Lurker (if only I didn't work
at turtle speed... Worse, I have been sick now for almost two days,
still recovering.)... Whatever the interface of marc.info, it is pretty
poorly creating and following the threads.
The emails that I needed to post the addresses of, would be a few
seconds and not minutes to find (like I spent searhing to find the links
for this email), if the Mutt archive was deployed with Lurker...
Only saying. And wishing.
Allow errata after I post this.
Regards!
--
Miroslav Rovis
Zagreb, Croatia
http://www.CroatiaFidelis.hr
signature.asc
Description: Digital signature
