Re: Registration of acct: as a URI scheme has been requested

On 20 June 2012 19:21, Harry Halpin <hhalpin@ibiblio.org> wrote:

> On Wed, Jun 20, 2012 at 7:17 PM, Melvin Carvalho <melvincarvalho@gmail.com
> > wrote:
>
>>
>>
>> On 20 June 2012 19:02, Harry Halpin <hhalpin@ibiblio.org> wrote:
>>
>>> On Wed, Jun 20, 2012 at 5:49 PM, Henry S. Thompson <ht@inf.ed.ac.uk>wrote:
>>>
>>>> I think the TAG should look at this.  It raises the general question
>>>> of the architectural appropriateness of URI schemes intended only for
>>>> use internally within a particular protocol.  As I understand it acct:
>>>> URIs have no meaning outside messages in the proposed webfinger
>>>> protocol.
>>>>
>>>
>>> I strongly second this. There is no reason why the "name@domain" can
>>> not be just handled as a string to be parsed to get the domain and then do
>>> a lookup in SWD/WebFinger. An additional URI scheme in front does not
>>> really help the algorithm, as it doesn't need to distinguish between
>>> SMTP-enabled names@domains (mailto:) and those that don't/may not
>>> ("acct:").
>>>
>>> Thus, as "acct:" is not attached to any concrete functionality beyond
>>> SMTP (as it does not normatively state state that such as account should be
>>> SWD/Webfinger-enabled) AND may even overlap with mailto: names@domains,
>>> I don't really see the point. . If we go back to the world where every new
>>> application needs a new URI scheme, something has gone wrong.
>>>
>>
>> FWIW I 100% agree with you
>>
>> However this was not the gist of the argument over at IETF
>>
>> WebFinger should be general purpose discovery.
>>
>> General purpose discovery based on a URI.
>>
>
> *Any* URI. Then why get a new URI scheme?
>

Yes, I believe the current draft supports any URI.  Therefore, webfinger in
theory does not have a dependency on acct:, at least this was what was
articulated by the author on the IETF mailing list.  The practical
implementation may be slightly different.

It's unclear the extent to which the whole of the webfinger community
support creation of acct:, for sure there are some strong advocates, but
there are also people that are opposed, and I think most sit in the middle
(my personal impression)

To my mind there are merits and demertis.  Using acct: is an added
complexity to the spec, that IMHO is not needed, but possibly a nice to
have.  Essentially HTTP URIs for this purpose, seem not to have been
strongly considered, yet will certainly be used by some players (perhaps
facebook is one).  The inevitable conclusion will be that Webfinger
adopters (eg webmail providers, federated social web / ostatus, unhosted,
diaspora etc.) will in the longer term, move to a keyring of identifiers to
identify a user.  This will certainly cause confusion and implementation
challenges in web 2.0 (I know of no analogue of owl sameAs) and perhaps
will take time to learn a new paradigm.  In any case, it will take several
years for acct: to reach anything like the maturity of mailto: or http:,
and may even become abandoned for one of these two at some point.

Systems like Kingsley's ODS can take multiple identifiers it's stride, but
for many or all of the webfinger adopters, it will be more of a challenge,
IMHO.  But perhaps the education in itself will be worthhwhile, even if it
means taking slightly longer to build robust systems.

Just my 2 cents.


>
>> Now there's no mechanism in XRD to look something up based on the object
>> of <foo> <bar>  <mailto:user@host>
>>
>> Only to look things up based on ?resource=<foo>
>>
>> This has made an implicit translation from mailto:user@host (object) ->
>> acct:user@host (subject) desirable.
>>
>
> This seems to be an XRD issue, not a reason for a new URI scheme. I mean,
> XRD has legacy connections to XRI as sort of URI-ish scheme regardless.
>
>>
>> SWD looked up based on mailto:
>>
>> Webfinger based on the @subject of an XRD doc.  IIRC subject is mandated
>> in the xml schema.
>>
>> Now webfinger was designed for smtp, but after 3 years of feature creep
>> the folloing 2 schemes came into play:
>>
>> SIP
>>
>> XMPP - but xmpp already has its own discovery.
>>
>> Next the idea was that twitter might want to use  acct:harry@twitter.comas an account identifier (they didnt ask for this) but people argue it
>> could be useful to identify things in this way.
>>
>
>
> If Twitter didn't ask for this, I don't see why it should be needed for
> WebFinger/SWD to go forward.
>
>
>>
>> So the whole thing is a bit more all encompassing than it was before.
>> How this will mesh with HTTP URIs as identifiers is an interesting question
>> and I dont think has really been explored in as much depth as it could have
>> been yet.
>>
>>
>>>
>>>   cheers,
>>>      harry
>>>
>>>
>>>
>>>>
>>>> ht
>>>>
>>>>
>>>>
>>>> ---------- Forwarded message ----------
>>>> From: Bjoern Hoehrmann <derhoermi@gmx.net>
>>>> To: "Paul E. Jones" <paulej@packetizer.com>
>>>> Cc: apps-discuss@ietf.org
>>>> Date: Mon, 18 Jun 2012 07:51:54 +0200
>>>> Subject: Re: [apps-discuss] FW: I-D Action:
>>>> draft-jones-appsawg-webfinger-06.txt
>>>> * Paul E. Jones wrote:
>>>> >I have also sent a request to the email address specified for URI
>>>> scheme
>>>> >review to try to move registration of "acct" along.  I hope that will
>>>> not
>>>> >take long, and I would hope it would not considering the specific and
>>>> narrow
>>>> >scope of "acct".
>>>>
>>>> http://www.ietf.org/mail-archive/web/uri-review/current/msg01605.htmlif
>>>> anyone wants to follow along.
>>>> --
>>>> Björn Höhrmann · mailto:bjoern@hoehrmann.de ·
>>>> http://bjoern.hoehrmann.de
>>>> Am Badedeich 7 · Telefon: +49(0)160/4415681 ·
>>>> http://www.bjoernsworld.de
>>>> 25899 Dagebüll · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/
>>>> _______________________________________________
>>>> apps-discuss mailing list
>>>> apps-discuss@ietf.org
>>>> https://www.ietf.org/mailman/listinfo/apps-discuss
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> --
>>>>       Henry S. Thompson, School of Informatics, University of Edinburgh
>>>>      10 Crichton Street, Edinburgh EH8 9AB, SCOTLAND -- (44) 131
>>>> 650-4440
>>>>                Fax: (44) 131 650-4587, e-mail: ht@inf.ed.ac.uk
>>>>                       URL: http://www.ltg.ed.ac.uk/~ht/
>>>>  [mail from me _always_ has a .sig like this -- mail without it is
>>>> forged spam]
>>>>
>>>>
>>>
>>
>