Information Technology Reference
In-Depth Information
refer, or both access and refer to, a thing . The problem at hand for Berners-Lee was
how to provide a name for his distributed hypertext system that could get “over the
problem of documents being physically moved” (1991). Using simple IP addresses
or any scheme that was tied to a single server would be a mistake, as the thing that
was identified on the Web should be able to move from server to server without
having to change identifier.
For at least the first generation of the Web, the way to overcome this problem was
to provide a translation mechanism for the Web that could provide a methodology
for transforming “unique identifiers into addresses” (Berners-Lee 1991). Mecha-
nisms for translating unique identifiers into addresses already existed in the form
of the domain name system that was instituted by the IETF in the early days of the
expansion of ARPANet (Mockapetris Novemeber 1983). Before the advent of the
domain name system, the ARPANet contained one large mapping of identifiers to
IP addresses that was accessed through the Network Information Centre, created
and maintained by Engelbart (Hafner and Lyons 1996). However, this centralized
table of identifier-to-address mappings became too unwieldy for a single machine as
ARPANet grew, so a decentralized version was conceived based on domain names ,
where each domain name is a specification for a tree structured name space, where
each component of the domain name (part of the name separated by a period) could
direct the user-agent to a more specific “domain name server” until the translation
from an identifier to an IP address was complete.
Many participants in the IETF felt like the blurring of this distinction that
Berners-Lee made was incorrect, so URIs were bifurcated into two distinct spec-
ifications. A scheme for locations that allowed user-agents via an Internet protocol
to access information was called Uniform Resource Locations (URLs) (Berners-
Lee et al. 1994) while a scheme whose names could refer to things outside of the
causal reach of the Internet was called Uniform Resource Names (URNs) (Sollins
and Masinter 1994). Analogue things like concepts and entities naturally had to
be given URNs, and digital information that can be transmitted over the Internet,
like web-pages, were given URLs. Interestingly enough, URNs count only as a
naming scheme, as opposed to a protocol like HTTP, because they cannot access any
information. While one could imagine a particular Web-accessible realization, like a
web-page, disappearing from the Web, it was felt that identifiers for things that were
not accessible over the Web should “be globally unique forever, and may well be
used as a reference to a resource well beyond the lifetime of the resource it identifies
or of any naming authority involved in the assignment of its name” (Mealling and
Daniel 1999).
Precisely because of their lack of ability to access information, URNs never
gained much traction, while URLs to access web-pages became the norm. Building
on this observation about the “blurring of identifiers,” the notion of URIs implodes
the distinction between identifiers used only for access (URLs) and the identifiers
used for reference (URNs). A Uniform Resource Identifier is a unique identifier
whose syntax is given by its latest IETF RFC that may be used to either or both
refer to or access a resource (Berners-Lee et al. 2005). URIs subsume both URLs
and URNs, as shown in Fig. 2.8 . Berners-Lee and others were only able to push
Search WWH ::




Custom Search