Re: For 1.5.9 - fixed smime-encrypt-self patch
On Thu, Feb 17, 2005 at 11:48:01PM -0500, Derek Martin wrote:
> On Thu, Feb 17, 2005 at 05:13:29PM +0100, Christoph Ludwig wrote:
> > If nobody sees problems with my patch then I request that it is included in
> > 1.5.9. If you are used to keep an fcc of your messages then
> > smime-encrypt-self
> > patch is a "must have". If you are afraid someone might force you (e.g.,
> > with
> > a subpoena) to decrypt and hand over copies of your communications
> > then the encrypt self feature can be disabled in .muttrc.
>
> Just curious... Why support this in mutt at all, when it can be
> controlled directly via gpg options?
There are two points to consider:
a) The gpgme code can be enabled and disabled in .muttrc so it makes sense to
build mutt with support for both the classic and the gpgme smime backend.
If you use the classic smime backend then you need the enrypt-self
patch. (See
<url:http://www.mail-archive.com/mutt-users%40mutt.org/msg31164.html> for
the origin of Omen Wild's original patch.) But the old encrypt-self patch
conflicted with the gpgme code. If both were enabled then you could not
send encrypted mails anymore.
Let's assume you are right and gpgme supports the encrypt-self feature
without help from mutt. Then you still have a combination of configuration
variables that break mutt. That casts the quality of mutt's code into
doubts because you cannot expect a user fiddling with .muttrc to know about
the internals of mutt's crypto implementation.
b) How do you think gpgme supports encrypt-self? (That is *not* a rhetorical
question. Most of what I know about mutt's crypto support I learned by
reverse engineering due to the terse documentation. So it is very well
possible that I missed something.) I am looking at the muttrc(5)
man page: with the exception of crypt_use_gpgme all options dealing with
gpg are marked as "PGP only". But I am talking about S/MIME support.
BTW: muttrc(5) mentions many configuration variables marked as "S/MIME
only" Some of them make only sense with the classic smime backend, others
may or may not be used by the gpgme backend as well. (I'd need to inspect
the code in order to tell for each variable.) The documentation should
explicitly state which backends honor the respective variables.
Regards
Christoph
--
http://www.informatik.tu-darmstadt.de/TI/Mitarbeiter/cludwig.html
LiDIA: http://www.informatik.tu-darmstadt.de/TI/LiDIA/Welcome.html