Re: [openpgp] First 4880bis drafts

[Aaron Zauner <[e-mail]>]

Hi,

Apparently my message came of the wrong way (and I'm to blame for that,
because of my wording):

Daniel Kahn Gillmor wrote:
> On Thu 2015-11-05 04:14:08 +0900, Werner Koch <[e-mail]> wrote:
>> On Wed,  4 Nov 2015 18:34, [e-mail] said:
>>
>>> Hrm. I'm against this. CAMELLIA is going to be deprecated in e.g.
>> You may be against it but it is a matter of fact that CAMELLIA is an
>> officially assigned OpenPGP cipher algorithm for 6 years.
> 
> As discussed in the meeting tuesday, deprecation is a tricky subject for
> formats with stored data (as distinguished from on-the-wire network
> traffic).  people have archives of encrypted data that may still use
> this cipher.

Totally agree there. And PGP implementations will support these ciphers
for years because of stored data that might have been encrypted with one
of these ciphers.

> 
>> We may latter decide to deprecate certain algorithms but that is not a
>> question right now.
> 
> The sense of the room in Yokohama was to deprecate as much as possible,
> and encourage a limited, sensible set of algorithms for message creation
> and signing.  But sensible implementations will likely continue to allow
> decryption of these ciphers for years to come.
> 

Yes. But we should discourage further use. I'm not sure if the right
place is the updated RFC or another document entirely. My concern is
that we'll end up with a unmanageable 'cipher-zoo'. I'm happy to help
with such a document and am _not_ trying to get in the way of updating
the current OpenPGP spec.

Hope that clears things up a bit,
Aaron