Access experimental subpackets of 'User Attribute Packet'

David Srbecky dsrbecky at
Fri Aug 12 10:17:10 CEST 2005

Joe Smith wrote:
> Be careful with this. First of all both the second and third packet can 
> be implemented as the first packet type.
That's the point of the first packet - to be able to express anything 
the user can think of. However, the other two packets are there to save 
space by not using string identifier. They are also easier to implement.

> Second encouraging users to 
> place extra data into theie keys is just asking for problem. Imaginge 
> when somebody decides to attaches a 50 MiB mpeg to their key, and 
> uploads it to the keyservers.
Unfortunately I am not asking for problem, the problem is already here. 
Users already can upload their photo albums to keyserver and no-one can 
stop them. One vCard and a few preferences really won't make it any 
worse, these are small compared to images.

> Third and most importantly, Anything for specifying pgpmime vs inline 
> mime should be avoided. Let the experts (The openpgp working group) hash 
> these things out.
I have already contacted the openpgp working group, but they are in last 
call and so it is unlikely that they will implement similar mechanism.

> For example there is the pgp corp implementation: 
This solves only just one little bit.

Not considering the feasibility, is it possible to achieve this with GnuPG?

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 254 bytes
Desc: OpenPGP digital signature
Url : /pipermail/attachments/20050812/4664b3e2/signature.pgp

More information about the Gnupg-users mailing list