#6 allow and do use key-id's instead?

[Tomas Pospisek]

Currently DIBS is using the email adresses, that are
associated with the gpg keys as IDs. This is
problematic, since this behaveour is being assumed as
being known in the docu.

If one is using gpg's key-IDs things "just don't" work.
In case the state of the art will be maintained, I
suggest to:
1) document this thoroughly (make this really clear)
2) improve error reporting a lot. As is if one is
using keys that DIBS will for whatever reason have
problems with, there is absolutely no feedback. Things
just don't happen.

Using email addresses as keys is quite ambiguous, since
email addresses are also used for emailing messages in
DIBS, not only for identification, I suggest switching
to gpg's own key-IDs for identification. For one using
Key-IDs (as gpg uses them) consequently and only
throughout the documentation would disambiguate
functional entities of the interface a lot. Enabling
using Key-IDs in addition or parallel to email
addresses as identifiers inside DIBS would allow to
keep backwards compatibility and to continue with an
improved interface.

I hope my idea comes accross, thanks,
*t

[Christian Stork]

Logged In: YES
user_id=23468

I just wanted to say that I agree with Tomas and I was about to suggest
renaming the --peer option to --peer-id or --peer-key-id.

Anecdotal evidence that this is a good idea: A friend of mine had
problems with the GPG generation and he send me two public keys with
the same email belonging to both keys.

Key IDs are unique and eliminate all such problems with the exception of
readability.

[Emin Martinian]

Logged In: YES
user_id=665772

OK, thanks for the comments. I'll make this a high priority
item.

-Emin