ietf-openpgp
[Top] [All Lists]

Re: "Yes, I can handle PGP/MIME"

2004-04-24 09:59:32

On Thu, 22 Apr 2004, Jon Callas wrote:

I disagree with where this is going.

The point of having notations is so that someone can put data into a
signature without having to have it be part of the standard -- without
having to get a consensus on it.

Agreed. This was my reason for bringing it up in the first place -- I
think that, without a means in OpenPGP for specifying the MIME preference,
this is what PGP Universal *should have* done.

If we put this in the standard, but define it to be a notation of
"rfc3156":{y|n} (or some other syntax) then we've done exactly what
David proposed, but improved his proposal by making it bigger and
harder to parse.

[snip]

I don't hear a consensus for putting this in. If other people want to
use notations for an ad-hoc implementation, great.

Okay. Then in the interest of achieving a consensus on this matter, I
hereby retract my suggestion that the OpenPGP/MIME preference be expressed
in the notation data field, and endorse David's proposal that we add an
additional subpacket flag. Are other backers of the notation-data proposal
willing to do the same? Please speak up.

The primary use of OpenPGP is email. While that may change in the future,
hypothetical concerns about instant messaging foo aren't presently an
issue. We have MUAs claiming "inline messages" are "old-format" and
deprecated, and we have users generating PGP/MIME messages which cannot be
processed by their recipients. Obviously there's a need for a means of
expressing this preference -- the GnuPG authors are asking for it, and the
PGP authors have already gone ahead with their own hack.

This additional flag adds minimal bloat to the standard, and addresses a
major usability problem with OpenPGP. I think it is a mistake to dismiss
this issue or require that yet another RFC be created to address it.


--Len.


<Prev in Thread] Current Thread [Next in Thread>