XRI

Draft

Introduction

I-names represent a unique name for a person or an organization in the same way that domain names represent unique names for machine/software identities like web services. I-names are resistered by a central instance, but can be re-assigned, e.g. if the owner sells and identity or if he stops paying the registration fee. I-numbers are machine readable i-names (i.e. the equivalent of IP addresses for humans), but in addition, these numbers cannot be re-assigned.

“XRIs are a new kind of identifier on the Internet, similar to URLs or e-mail addresses. However, a single XRI can be used for different services, such as a website, e-mail, skype, icq or any other. They are therefore neither website nor e-mail addresses alone; they can be both at the same time, and more.” (@fullXRI, retrieved 22:19, 23 February 2010).

The practical advertized advantage of i-name is that they are device independent and that a user can control what kind of information what kind of service or agent can access. e.g. one may give or not give permission to tranlate an i-name into an email-address. XDI (retrieved 21:35, 23 February 2010) explains the advantage of I-Names in the following way: “Conventional addresses such as postal addresses, phone numbers, and email addresses are tied to a specific location, device, or service. By contrast, i-names are abstract—they are not tied to any specific location or device. Instead they are a way to ask permission to contact an individual or organization—and for the i-name owner to control to whom this permission is granted. [...] An i-name is simply unspammable — you can’t send it email, call it, or send it a fax directly unless the owner has given you permission.”. I-names/XRI therefore is a technical solution for personal identity management (PIM).

The XRI initiative has a somewhat controversial status. W3C strongly opposed revision 2 of this standard and as result (25% opposition votes) did not make it a standard. Version 3 three is currently (March 2010) under preparation and may be accepted by OASIS. According to Wikipedia, the core of the dispute is whether the widely interoperable HTTP URIs are capable of fulfilling the role of abstract, structured identifiers, as the TAG believes.

See also: OpenID. XRI and i-names are (being) integrated into the OpenID framework.

The XRI standard

Let's examine Version 3 of XRI standard (proposal as of March 2010): “XRI (Extensible Resource Identifier) provides a common language for structured identifiers that may be used to share semantics across protocols, domains, systems, and applications. XRI builds directly on the structure and capabilities of URI (Uniform Resource Identifier) [URI] and IRI (Internationalized Resource Identifier) [IRI]. XRI is a profile of URI and IRI syntax and normalization rules for producing URIs or IRIs that contain additional structure and semantics beyond those specified by [URI] or [IRI].”

This specification under the headings Introduction / Motivations then presents a few commonly cited motivations for for needing a common language for structured identifiers (XRI):

• To unambiguously assert that the same resource is being identified across different protocols, e.g., HTTP, HTTPS, FTP, SMTP, XMPP.
• To unambiguously identify the same resource in different contexts, i.e., within different domains, systems, applications, namespaces, etc.
• To assign, resolve, and determine the equivalence of different synonymous identifiers for the same resource, e.g., persistent vs. reassignable synonyms, human-readable vs. machine-friendly synonyms, localized vs. non-localized synonyms.
• To identify different versions of the same resource in a manner that is consistent across multiple domains, systems, and applications.
• To create structured identifiers to address, navigate, and share structured data, such as RDF graphs.

XRI is a standard that defines a fairly abstract concept for defining various identity schemes like i-cards, i-names, i-numbers and OpenID. XRI stands for EXtensible Resource Identifier and has been developed by OASIS as “a standard for a high-level naming/identification system for individuals, businesses, communities, services and data on the Internet. XRI, along with XDI, a general-purpose data interchange protocol based on XRI, were developed to create the "Dataweb," which enables the Web to operate like a global database.” (ZDNet, retrieved 22:19, 23 February 2010). XRI's are also an option for OpenID user names. I-names are unique humain readable names, but they may change over time for a given subject. I-numbers are machine readable identifiers and should remain persistent. I.e. an application would both remember the i-name and the i-number. The latter should always point to the same person, even when the i-name changes.

• The XRI Identifiers (I-Names and I-numbers) are administered by XDI.org. I.e. XDI.org accredits I-Brokers. You can find these on the i-broker page page of inames.net

The XRI proposal also refers to related work. URNs (identified by the urn: scheme) are persistant. XRIs have part that is persistent, but like URIs can be reassigned. In the XRI Syntax v2.0 Submitted for OASIS Standard (retrieved 17:34, 24 February 2010 (UTC)) mail we can read that “XRIs build on the foundation of interoperable Web identifiers established by URIs (Uniform Resource Identifiers, RFC 3986) and IRIs (Internationalized Resource Identifiers, RFC 3987). Just as the IRI specification created a new identifier by extending the unreserved character set allowed in generic URIs, and defined rules for transforming IRIs into valid URIs, the XRI Syntax 2.0 specification creates a new identifier by extending the syntax of IRIs and defining transformations of XRIs into valid IRIs (which can then be transformed into valid URIs. [...] XRI Syntax 2.0 extends IRI/URI syntax by: (a) Allowing the internal components of an XRI to be explicitly tagged as either persistent or reassignable. (b) Enabling XRIs to contain other XRIs (or IRIs or URIs), a syntactic structure called "cross-referencing" that allows sharing of identifiers, such as generic identifiers or "tags", across multiple authorities and namespaces. (c) Supporting new types of identifier authorities including global context symbols and cross-references.”

Examples

What do XRIs look like?

Simple XRIs are understood by applications that can handle these. They start with either a = or a @ character, and after that can be made up of an arbitrary number of 'subsegments', which are usually separated by a * character.

• = identifiers refer to names of individuals
• @ identifiers refer to names of organizations

Examples of individuals:

• =daniel.k.schneider
• =danielkschneider (not recommended)
• =daniel-schneider (not recommended)

Examples of organizational names:

• @example.company.name

Examples of subentities in organizations (so-called community i-names)

• @example.company.name*division*sub-division
• @tecfa*daniel.schneider
• @blog*lucy

Full (absolute) XRIs are a kind of URI and may look like this:

• xri://=daniel.k.schneider

Such URLs are not yet understood by current web browsers. (22:19, 23 February 2010)

Since XRIs are a kind of URIs (i.e. more poweful than URLs, one can append extra arguments to an XRI (I don't know yet what would be standardized - Daniel K. Schneider). Example:

• =daniel.k.schneider(+blog)

Dealing with XRIs in todays webbrowsers

“Today, webbrowsers and operating systems do not yet natively support XRI resolution. In order to work with XRI technology, you either need to use special XRI-enabled software such as webbrowser plugins or our XRI Ping tool, or you can let public XRI proxies resolve your XRI.”(@fullXRI, retrieved 22:19, 23 February 2010).

“To use an XRI proxy, you need to form a so-called HXRI, which is an XRI prepended with the URL of a HTTP-based XRI proxy server. The proxy server will then perform the XRI resolution for you and redirect you to the appropriate target URI. XRI proxy servers usually begin with http://xri..”(@fullXRI, retrieved 22:19, 23 February 2010).

• http://xri.net/, operated by XDI.org - the organization behind the XRI standard.
• http://xri.fullxri.com/, proxy of a certified i-names registrar.

Life examples for DKS

I got the =daniel.k.schneider i-name with @fullXRI ($12/year) and now can use URLs like this:

• http://xri.net/=daniel.k.schneider (will resolve to the current home page of DKS)
• http://xri.fullxri.com/=daniel.k.schneider (same as above)

• http://xri.net/=daniel.k.schneider/(+contact) (will resolve to a short contact page, that doesn't give any URL or e-mail address)

• http://xri.net/=daniel.k.schneider/(+blog) (will resolve to the favorite blog of DKS)

Using the free @freeXRI:

• http://xri.net/@id*daniel.schneider (this one will redirect to my wiki user page)

Links

Standards (including propositions)

• XRI 1.0 (Jan 2004, deprecated now)
• XRI Syntax 2.0 (Committee Specification, 14 November 2005).

The current (March 2010) XRI initiative is based on series of proposed standard

• xri-syntax 3.0 wd 03 (PDF)
• XRD (Extensible Resource Descriptor), based on Cool URIs for the Semantic Web (W3C Interest Group Note 03 December 2008)
• XRI Resolution
• XDI (XRI Data Interchange)

Related standards

(there are more ... see the [xri-syntax 3.0 wd 03 (PDF) or more recent)

• URN RFC 2141, RFC 1737
• URI
• DOIs and the Handle System: Digital Object Identifiers are used for persistent and actionable identification and interoperable exchange of managed information on digital networks). I.e. most scientific articles online now have a DOI. DOI's are another kind of URN and uses for its name resolution the Handle System. (See RFC 3650, RFC 3651, and RFC 3652.)

• URNs, Namespaces and Registries A W3C W3C Technical Architecture Group (TAG) that has been used to critized XRI v2.0

i-name registrars

• @fullXRI
• 1id
• @freeXRI allows to create a free XRI of the kind: @free*yourname, @id*yourname, =web*yourname, etc.

XRI resolvers

• Use http://xri.net/ and just append the i-name, e.g. like http://xri.net/=daniel.k.schneider

Organizations

• OASIS is a major player for XML-related standards. With respect to digital identity: SAML, XDI, XRI etc. See also the XRI committee, Web Services Security (WSS) Technical Committee, eXtensible Access Control Markup Language (XACML) Technical Committee, etc.

• XDI.org (manages XRI spaces), i.e. through inames.net (the XDI.org portal for i-names)

Discussion (pro/against)

On 31 May 2008 OASIS Standard vote on XRI 2.0 specifications failed approval by 1 percentage point....

• Detailed technical reasons why I'm against XRIs by Dave Orchard.
• Talk:Extensible Resource Identifier (Wikipedia discussion page of the XRI page, raises problems with respect to patents and XRI actors)

• XriAsRelativeUri, OASIS page that summarizes a new proposal for how to integrate XRIs with URIs and WWW architecture that has arisen from discussions between the OASIS XRI TC and the W3C TAG starting in July 2008.

Technical information

• inames wiki, community wiki for XRI and i-names developers.
  • E.g. Why XRI ?

• OASIS XRI FAQ

• XRI and OpenID

Some Wikipedia articles

• Wikipedia Digital Identity
• i-name (Wikipedia)
• I-number (Wikipedia)
• I-broker (Wikipedia)
• XRI (Wikipedia)
• XRDS (Wikipedia)
• XDI (Wikipedia)