Did something break mailcrypt?

Brian Warner warner at lothar.com
Wed Sep 8 13:04:53 CEST 1999

Hash: SHA1

[sorry to take so long to respond to this.. I've been away at Burning Man for
the last week and a half]

dayton at overx.com (Soren Dayton) writes:

> I'm using XEmacs 21.1 (patch 6) "Big Bend" and mailcrypt 3.5b7, and I
> recently upgraded to gnupg 0.9.10 from 0.9.8 and I've found that
> mailcrypt just doesn't work anymore. (checked that it wasn't a
> mailcrypt/emacs problem by returning to the previous version of gpg,
> which worked.
> mailcrypt reports immediately that my passphrase is no good, before
> I've typed one in.  Glancing at the mailcrypt code, it seems that it
> determines this by the return value of gpg, which is not documented in 
> the manpage (hmmm.  Nice hidden dependency there...)
> Is this expected or should this have been expected?  If no, what sorts 
> of information would be useful for debugging?

Oops, sorry. A new status message appeared in 0.9.9, and my mc-gpg.el parser
didn't handle it correctly. The passphrase is reported as bad before it is
ever actually used.. this affects both signing and decryption.

In the process of fixing the problem, I completely revamped the parser. Now it
never looks at the stderr messages, and almost never uses the return code. It
should be much more robust in the face of new status-fd messages (at the
expense of some minor compatibility problems with older GPG versions, 0.9.4
and earlier).

The replacement mc-gpg.el is on my site <http://www.lothar.com/tech/crypto/> .
Now that I'm back from vacation, I'll be sending it to Len and will encourage
him to make a new release of mailcrypt soon.

Let me know if you have any problems with the new code.

   warner at lothar.com

Version: GnuPG v0.9.9 (GNU/Linux)
Comment: For info see http://www.gnupg.org


More information about the Gnupg-devel mailing list