Re: Use of message disposition notification

[Russ Allbery <[e-mail]>]

"Charles Lindsey" <[e-mail]> writes:
> Julien ÉLIE <[e-mail]> writes:

>> By the way, as you speak about security considerations, I see that
>> USEPRO references both pgpmoose and pgpverify.  Wouldn't a note on
>> Cancel-Lock be added in 6.1?  ("Currently, many sites are ignoring
>> all cancel control messages and Supersedes header fields due to the
>> difficulty of authenticating them and their widespread abuse.")  It
>> could be said that Cancel-Lock might be used
>> (draft-ietf-usefor-cancel-lock-01) and also NoCeM, which is more and
>> more spread nowadays.

> Cancel-Lock is not formally defined anywhere, hence I know of no
> news-server that ever looks at it, although a few people add it in
> hope :-).

> Our original plan of work was that, when we were done with our present
> drafts, the next jobs would be UTF-8 in newsgroup-names and a document
> covering security matters - notably pgpmoose and pgpverify (which
> really need to be unified) plus Cancel-Locks and, I would hope,
> NoCeM. But those would rechartering, and the probability of that
> happening is not, shall we say, too promising :-( .

Unless someone feels very strongly otherwise, I'm going to bail on
cancel lock for this document on those same "requires thought" grounds.
The logical next step for Cancel-Lock would be to complete the work of
publishing an RFC, which could then update this one and generate the
right references in RFC listings.

-- 
Russ Allbery ([e-mail])             <http://www.eyrie.org/~eagle/>