Bug#318587: gnupg: should encrypt to all subkeys

2005-07-18 Thread Steinar H. Gunderson
On Mon, Jul 18, 2005 at 07:55:57AM +0200, Werner Koch wrote:
 Well, does OpenPGP specify at all which subkeys to encrypt to? Is there a
 good reason why GnuPG simply can't encrypt to both by default?
 No.  Why only to both ot them? There are often more than just 2
 non-expired encryption keys. 

Well, s/both/all/. What is the disadvantage, if any?

 For sure that is not a bug.  Using the latest valid encryption subkey
 is what almost everyone would expect.  Anything elese does not make
 much sense.

Does it make much more sense having multiple subkeys, but in reality only use
one of them? I'm not sure if I catch the logic here :-)

 Whether something is a card key or a gpg-agent controlled
 key or a plain disk stored key or a PGP 8 key or ... is not visible to
 a someone going to encrypt to a key.

Yes, that is _exactly_ my point, and which is why it should encrypt to all
available subkeys by default :-)

/* Steinar */
-- 
Homepage: http://www.sesse.net/


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#318587: gnupg: should encrypt to all subkeys

2005-07-18 Thread Werner Koch
On Sun, 17 Jul 2005 16:54:44 +0200, Steinar H Gunderson said:

 Well, does OpenPGP specify at all which subkeys to encrypt to? Is there a
 good reason why GnuPG simply can't encrypt to both by default?

No.  Why only to both ot them? There are often more than just 2
non-expired encryption keys. 

 Mm, but then I'd have to revoke the old encryption subkey to work around what
 I consider is a bug in GnuPG, and I'd hate accumulating cruft for such
 reasons :-/

For sure that is not a bug.  Using the latest valid encryption subkey
is what almost everyone would expect.  Anything elese does not make
much sense.  Whether something is a card key or a gpg-agent controlled
key or a plain disk stored key or a PGP 8 key or ... is not visible to
a someone going to encrypt to a key.


Shalom-Salam,

   Werner




-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#318587: gnupg: should encrypt to all subkeys

2005-07-18 Thread Werner Koch
On Mon, 18 Jul 2005 11:36:20 +0200, Steinar H Gunderson said:

 Well, s/both/all/. What is the disadvantage, if any?

Overhead in size and performance.

 Does it make much more sense having multiple subkeys, but in reality only use
 one of them? I'm not sure if I catch the logic here :-)

The subkeys are used for different operations (sign, authenticate,
encrypt) and for key-rollover (to achieve a certain amount of PFS
(perfect forward secrecy)).  The latter actually requires that one
does not encrypt to any older subkeys even if they are still valid -
the owner of the key might have already deleted that key.

Please continue the discussion on [EMAIL PROTECTED]  I don't think it is
appropriate for a BTS.


Salam-Shalom,

   Werner



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#318587: gnupg: should encrypt to all subkeys

2005-07-17 Thread Werner Koch
On Sat, 16 Jul 2005 12:45:35 +0200, Steinar H Gunderson said:

 When encrypting to a master key with multiple encryption subkeys, GPG
 currently signs to only the newest one. In my case, one is available on
 my home computer (which does not always have a smart card reader
 attached), and the other one is available on a smart card only -- in
 other words, I would really like all messages encrypted to both by default.

Add:

encrypt-to 12345678!
encrypt-to 9abcdef0!

to your gpg.conf.  The two keys are the keyIDs of the respective
subkeys.  Don't forget the exclamation mark to force gpg to use
excactly these subkeys.

You can't however force others to encrypt to a specific key; this is
not defined by OpenPGP and we don't implement the highly questionable
ARR PGP provides.

Another way to solve this is by generating the key on the host and
transferring a copy to the smartcard.  Off-card generation is actually
the default for smartcard encryption keys.


Salam-Shalom,

   Werner



-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#318587: gnupg: should encrypt to all subkeys

2005-07-17 Thread Steinar H. Gunderson
On Sun, Jul 17, 2005 at 04:00:59PM +0200, Werner Koch wrote:
 Add:
 
 encrypt-to 12345678!
 encrypt-to 9abcdef0!
 
 to your gpg.conf.  The two keys are the keyIDs of the respective
 subkeys.  Don't forget the exclamation mark to force gpg to use
 excactly these subkeys.

That doesn't help me at all, of course -- I very rarely encrypt stuff to
myself.

 You can't however force others to encrypt to a specific key; this is
 not defined by OpenPGP and we don't implement the highly questionable
 ARR PGP provides.

Well, does OpenPGP specify at all which subkeys to encrypt to? Is there a
good reason why GnuPG simply can't encrypt to both by default?

 Another way to solve this is by generating the key on the host and
 transferring a copy to the smartcard.  Off-card generation is actually
 the default for smartcard encryption keys.

Mm, but then I'd have to revoke the old encryption subkey to work around what
I consider is a bug in GnuPG, and I'd hate accumulating cruft for such
reasons :-/

/* Steinar */
-- 
Homepage: http://www.sesse.net/


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]



Bug#318587: gnupg: should encrypt to all subkeys

2005-07-16 Thread Steinar H. Gunderson
Package: gnupg
Version: 1.4.1-1
Severity: wishlist

When encrypting to a master key with multiple encryption subkeys, GPG
currently signs to only the newest one. In my case, one is available on
my home computer (which does not always have a smart card reader
attached), and the other one is available on a smart card only -- in
other words, I would really like all messages encrypted to both by default.

-- System Information:
Debian Release: testing/unstable
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: i386 (i686)
Shell:  /bin/sh linked to /bin/bash
Kernel: Linux 2.6.11.8
Locale: LANG=en_DK.UTF-8, LC_CTYPE=en_DK.UTF-8 (charmap=UTF-8)

Versions of packages gnupg depends on:
ii  devfsd  1.3.25-23Daemon for the device file system
ii  libbz2-1.0  1.0.2-7  high-quality block-sorting file co
ii  libc6   2.3.5-1  GNU C Library: Shared libraries an
ii  libldap22.1.30-11OpenLDAP libraries
ii  libreadline55.0-10   GNU readline and history libraries
ii  libusb-0.1-42:0.1.10a-16 userspace USB programming library
ii  makedev 2.3.1-78 creates device files in /dev
ii  zlib1g  1:1.2.2-8compression library - runtime

gnupg recommends no packages.

-- no debconf information


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]