Datum: 13.03.2015 08:05
Sender: Bernhard Reiter
thanks for your feedback and the suggestion.
I think for the discussion part, the forum or one of the mailinglists would be the best place.
E.g. gnupg-users@. The “mail transfer agent” solution has been discussed many times.
What I have taken from these discussion are two challenges with it:
a) If you use this relay on a server, it is not end-to-end encryption anymore, so you have to trust this server.
So far the Gpg4win Initiative has focussed on end-to-end crypto, so it would be out of scope for us (currently).
b) If you use in on the desktop it could be a personal solution, but in order to make a number of choices
that are almost unavoidable in some situations, it would need to have an interface. The natural place for this
interface is within the application that already handles your email. So from the user interface point of view,
the end-to-end crypto should be build into the client you are already using.
Note that Werner (principal author of GnuPG) also started an MTA solution product in 2001, interest was low,
ps.: Flattr Gpg4win at https://flattr.com/thing/2053326,
if you appreciate this answer and my work within the Gpg4win Initiative.
Datum: 12.03.2015 09:50
Sender: Esh L’Orac
I found an interesting project named GPGRelay (http://sourceforge.net/projects/gpgrelay/). It seems to be abandoned since 2005, nevertheless has yet today a userbase. It appears to solve the feature request mostly.
Of course there has to be fixed some bugs with imap relay and added some features mostly to notify and ask the user whether he likes to send the email unencrypted.
It is open source, so why not integrate it into the gpg4win project? It has several advantages, it’s easy to setup, runs even on windows 8.1 and last, but not least, it lets you use your favorite email client without a plugin (so no maintenance for plugins for further versions is needed). What speaks generally against this proposal?