W3C home > Mailing lists > Public > public-xg-socialweb@w3.org > January 2010

Re: Kim Cameron's Laws of Identity

From: Melvin Carvalho <melvincarvalho@gmail.com>
Date: Sat, 23 Jan 2010 13:44:56 +0100
Message-ID: <9178f78c1001230444v41916909j8ba3bb47e7b1aa21@mail.gmail.com>
To: Kaliya <kaliya@mac.com>
Cc: public-xg-socialweb@w3.org
On Fri, Jan 22, 2010 at 6:53 PM, Kaliya <kaliya@mac.com> wrote:

> It isn't a "digital identity"
>
> see our lexicon...http://wiki.idcommons.net/Digital_Identity
>
> Digital Identity
> Definition: A digital representation of a set of Claims <http://Claim> made
> by one Party about itself or another Digital Subject<http://Digital_Subject>.
> [originally from Kim's Laws, ScottL, PaulT, BobWyman]
> Comment1: A Digital Identity is just one set of Claims <http://Claim> about
> a Digital Subject <http://Digital_Subject>. For any given Digital Subject<http://Digital_Subject> there
> will typically exist many *Digital Identities* . [PaulT]
> Comment2: A Digital Identity can be created on the fly when a particular
> identity transaction is desired, or persisted in a data store to provide a
> referenceable representation [ScottL, Drummond, MaryRuddy]
> Comment3: A Digital Identity may contain Claims <http://Claim> made by
> multiple Claimants. [DickH]
> Comment4: A Digital Identity may be signed by a Digital Identity Provider<http://Digital_Identity_Provider> to
> provide assurance to a Relying Party <http://Relying_Party> [ConorC]
>
>
As I mentioned in the telecon, TimBL wrote a great essay describing aspects
of the "Social Graph".  IMHO, it's an excellent "big picture" piece, which I
always get something from (re) reading

http://dig.csail.mit.edu/breadcrumbs/node/215

The same author, also came up with the concept of a "Web Identity" (Web ID)
[1].  This is an identity in the Linked Data world, which has the property
that it is defined to be a (FOAF) Person (the FOAF naming system, is machine
readable, dates back about 10 years, and includes terms like Agent, which is
a super class of Person).  It is periodically debated whether someone should
have 'one true Web ID' or a collection thereof.  The current thinking
(fashion) is that someone should be able to have a number of Web ID's that
can be linked together, as desired.  One of the motivations of the "Linked
Data" movement, is to provide an intuitive method for doing this (linking
via hyperlinks).

Kaliya made a good point on the telecon, that whatever people choose to name
things, it would be helpful to create columns that map one set of
terminologies to another, so that each group can understand how terms
translate into other contexts.

The good news is that this process has already begun, partly as a
consequence of the conversations that have been taking place in the SWXG.
For example, a mapping has been made [2], between the FOAF vocablulary (the
biggest in the sem web) and Portable Contacts / OpenSocial, and alignment
changes have already made their way into latest version of the FOAF
vocabulary.  I hope that continuing effort, to keep communication channels
open, will facilitate a greater understanding between groups, and ability to
leverage the work contributed to, by disperate communities.

[1] http://esw.w3.org/topic/WebID
[2]
http://spreadsheets.google.com/ccc?key=0AqtcXb4fZGyCdEtNSVJ0LXdvT3dBTHllOTZEbmZFaFE&hl=en


>
> So "digital Identity" doesn't mean..."This one site can contain the
> aggregated of all the users information from multiple personae.  The person
> (or user) has control over what aspects are shared in what locations. "
>
> Find a different word or phrase that describes what you are suggesting. I
> have put several forward that I think would do.  "master IdP" or "uber-IdP"
>  "complete identity host"
>
> We don't want a web where everyone says "your ID please"
> it is more like what identifier do you want to use today (and the user has
> many options)
>
>
>
>
>
> On Jan 22, 2010, at 9:31 AM, Alexander Korth wrote:
>
> Call it »Digital Identity«, then for instance. Luckily everything comes
> with a context. Thus, the industry may very well understand the specific
> concept.
> »Your ID, please.« ;)
>
> - Alex
>
> On 22.01.2010, at 18:14, Kaliya wrote:
>
>
> On Jan 22, 2010, at 2:14 AM, Alexander Korth wrote:
>
>
> Hi all,
>
>
> -1 (minus) need to come up with new term. The term is already very widely
> accepted in both, industry and science. There won't be any benefits from
> renaming the term.
>
>
> the term "identity" means many things and it is accepted in industry when
> not used alone for a specific thing you decide to define (when other people
> have a different definition.
>
>
>
> The specific way that "identity" is defined in the paper as
>
> "This one site can contain the aggregated of all the users information from
> multiple personae.  The person (or user) has control over what aspects are
> shared in what locations"
>
>
> is NOT what is meant by the term IdP - that is the provider/host of AN
> (one) identifier that belongs to a person not the complete total of all
> identities/identifiers + claims (profile fields) that they have on the web.
>
>
> You need to be clear what you mean by "identity" notice the other term you
> mention is "Identity PROVIDER" it is used in conduction with another term
> that helps it be at least somewhat specific in what it is referring to.
>
>
> You could go with something like "master IdP" or "uber-IdP" or some other
> "complete identity host" (cause shoudln't the user/person be able to move
> this to different places) - what ever you do please don't use the word
> "identity" by itself - that will cause problems.
>
>
> Just thoughts.
>
> -Kaliya
>
>
>
>
> The problem of having several profiles will not diminish but change
> significantly over the next few years:
>
> People will choose one main profile which is hosted by an identity
> provider. That provider most of the times will be one that they trust most
> and/or that does already have loads of profile information about the user
> anyways. There will be a tipping point where the opening to these profile
> data from the provider side will be so significant that service providers
> will choose to not any more implement an autarkic user management - this is
> the separation of service and identity management. There are many visions
> towards this, including my own [1,2,3].
>
> When this happens, there won't be many equivalent and proprietary profile
> providers any more but *the* one main identity provider per user. These
> providers will have an agreed set of interfaces to read (and write) data.
> This is the enabler for a new generation of user-centric profile management
> features, incl. control over information reach, deletion of information etc.
>
> The terms identity and identity provider (IdP) are widely accepted and
> used. To me, it emphasizes a service's focus on identities as such which I
> understand as a much higher sophisticated profile from both, the user model
> quality and the management feature set point of view.
>
>
> Cheers,
>
> Alexander Korth
>
>
> Managing Director of Make Customers Friends (
> http://www.makecustomersfriends.com)
>
>
> [1] Marc Canter's Open Mesh
> http://blog.broadbandmechanics.com/how-to-build-the-open-mesh/
>
> [2] Dick Hardt's legendary and brilliant because 4 yrs old Identity 2.0
> talk http://www.youtube.com/watch?v=RrpajcAgR1E
>
> [3] My Web of Identities, which is kind of a LinkedData amongst IdPs to
> look up ppl data
> http://www.readwriteweb.com/archives/web_of_identities_making_machine-accessible_people_data.php
>
>
> On 22.01.2010, at 05:08, Renato Iannella wrote:
>
>
>
> On 21 Jan 2010, at 20:19, Yuk Hui wrote:
>
>
> if the identity is an aggregated profiles + selected + verified, then this
> seems to be much more complicated, for example I have three profiles (e.g.
> facebook, twitter, youtube) with different information, what will be this
> unique identity then?
>
>
>
> This is probably why "Identity" is the wrong term....
>
>
> Cheers...  Renato Iannella
>
> NICTA
>
>
>
>
>
> --
>
> Alexander Korth
>
> www.twitter.com/alexkorth
>
>
>
>
>
>
> VG,
> Alex
>
> --
> Alexander Korth
> www.twitter.com/alexkorth
>
>
>
>
Received on Saturday, 23 January 2010 12:45:28 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:22:08 UTC