The World Wide Web - Enabling Semantic Web Services: The Web Service Modeling Ontology - page 9

Here, scheme denotes, for instance (but not necessarily), a protocol over which
the respective resource is accessible. Nowadays, the http:// scheme is most
commonly used. The owning entity of the resource is denoted by authority .
In most cases this is a server or domain name which has the practical side
effect that domains indeed belong to legal entities. Explaining the details
of the query and fragid (fragment identifiers) parts would go beyond the
scope of this topic, but most readers will have experienced them already in
the common use of HTTP. For example, when a search engine is accessed,
the query for a keyword is usually coded in the URI, such as http://www.
google.com/search?q=wsmo ; this URI encodes a query for the search string
wsmo at Google. Fragment identifiers denote a secondary resource with no
direct path where the interpretation of fragment IDs depends on the type
of medium. For instance, the usual use in HTML denotes an anchor within
a hypertext document. URIs without scheme parts are called relative URIs,
where “relative” means that these URIs are understood relative to a base URI.
What this base URI is or defaults to depends on the particular application of
the URI; in HTML, for instance, it is the document URI, where the document
is accessible on the Web.
Note that we learn from the now-deprecated distinction between URNs
and URLs that although URIs/IRIs identify resources, it does not follow that
any URI is a deferencable, web-accessible resource. As we shall see in later
chapters, the principle of unique identification via globally unique identifiers
can and is used in ways other than dereferencing Web resources accessible
over a particular protocol on a particular server. Still, it is often useful to take
advantage of some of the useful characteristics of URIs (such as ownership
by an authority by the use of domains) for unique identification of resources
other than documents on the Web. Usual uses of URIs include:

• addresses on the Web (which include documents, service endpoints, etc.);

• namespaces in XML QNames or other languages (see Section 2.3);

• identifiers of things and concepts (e.g. RDF, see Section 3.2);

• unique keys (e.g. MIME message ID).

The specification of URIs reached its current state in RFC 3986 in January
2005. IRIs (International Resource Identifiers) are a recent extension of URIs
which allow a richer character set, i.e. full Unicode, as defined in RFC 3987.

2.2.2 A Protocol for Hypertext Transfer - HTTP

The Hypertext Transfer Protocol (HTTP) is a simple protocol that provides
all necessary means for simple client/server interaction in a stateless request-
response manner. Such a manner allows a client to open the connection with
a request, whereafter the server responds and closes the connection again.
HTTP is the protocol of choice for most Web-based communication for de-
livering files and other data (collectively called resources) on the World Wide