Re: [saag] IETF 93 Agenda Request - Key Discovery

[Richard Barnes <[e-mail]>]

On Wed, Jul 22, 2015 at 10:28 PM, Viktor Dukhovni
<[e-mail]> wrote:
> On Wed, Jul 22, 2015 at 09:18:15AM +0200, ? Matt Miller wrote:
>
>> However, do not confuse RFC 7565 "acct:" URIs for email addresses.
>> Yes, they look a lot like an email address, as a "xmpp:" or "sip:" URI
>> can look a lot like an email address.  The "acct:" URI is for a
>> generic account identifier; it could be a placeholder to email, IM,
>> VoIP, filesharing, etc.
>>
>> My draft is intended for more than email.
>
> A major difficulty is that email addresses and "accounts", are not
> necessarily in one to one correspondence or even "few to one"
> correspondence.

This doesn't actually matter.  WebFinger lets you put any URI in the
"resource" field.  So just use a "mailto" URI if you're sad about
"acct".  The document should probably say that, and probably say you
should return the same thing in either case.

--Richard

>
> How is one to know which "acct" URI is applicable to finding the
> keys for a particular email recipient?  I probably should not even
> be able to easily find out the login account behind a particular
> email address.
>
> Is the proposal intended to support key discovery for first contact?
> If so how is the sender (of an email say), going to the find the
> associated account to query?
>
> Or is the sender expected to have previously received an "acct"
> URI out-of-band and is only using it to maintain key freshness?
>
> It still seems like this proposal is not a natural fit for email.
> Perhaps it is better applicable to other protocols...
>
> --
>         Viktor.
>
> _______________________________________________
> saag mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/saag