Re: the possible impact of future changes in webfinger (was Re: Anonymity and multiple identities)

Hi Kingsley,

On Sat, Jul 7, 2012 at 8:52 PM, Kingsley Idehen <kidehen@openlinksw.com> wrote:
> To conclude, neither http: nor acct: based agent URIs monopolize federated
> identity.

that's a valid standpoint, and if there are IdP's (identity providers)
who prefer to offer http: URLs instead of implementing webfinger, then
we can allow both.

What we do have to decide IMO is what data format the document found
there should have. For webfinger we have XRD and JRD, and for
http-based identities i think it's mainly foaf, right?

examples:

http://docs.oasis-open.org/xri/xrd/v1.0/xrd-1.0.html#examples
http://tools.ietf.org/html/rfc6415#page-14

http://wiki.foaf-project.org/w/DescribingAPerson

there's also a microformat to describe a vCard:

http://microformats.org/wiki/jcard

and an rdf vcard option:

http://www.w3.org/TR/vcard-rdf/

i think ideally we want to support all common formats, especially at
the beginning, to get as much federation as we can, and then later we
can deprecate some of them?

> We need to stop preoccupying ourselves with implementation details.

i agree we should try to look beyond syntactical issues, and just use
anything that works, although if we want interoperability then we will
have to make sure implementation details of server A are compatible
with those of server B, of course.

anyway, i think with this i could write up a wiki page about identity.
we can then make other wiki pages about how to deal with activity
streams, with private messages, and with i don't know, groups, and
that way create a bit of a knowledge base describing what people are
doing. once we've described, we can analyse, cull, and in the long
run, prescribe.

is that useful?

Received on Saturday, 7 July 2012 19:08:59 UTC