Composite Capability/Preference Profiles (CC/PP): Structure and
Vocabularies 1.0
W3C Recommendation 15 January 2004
This version:
Latest version:
Previous version:
Editors:
Graham Klyne,
GK@acm.org
, Nine by
Nine
Franklin Reynolds,
franklin.reynolds@nokia.com
Nokia Research Center
Chris Woodrow,
woodroc@metaphoria.net
Information Architects
Hidetaka Ohto,
ohto@w3.org
, W3C
(through March 2002) / Panasonic
Johan Hjelm,
Johan.hjelm@ericsson.com
Ericsson
Mark H. Butler,
mark-h_butler@hp.com
Hewlett-Packard
Luu Tran,
luu.tran@sun.com
, Sun
Microsystems
Please refer to the
errata
for this
document, which may include some normative corrections.
The English version of this specification is the only normative version.
Non-normative
translations
may
also be available.
W3C
MIT
ERCIM
Keio
), All Rights Reserved. W3C
liability
trademark
document
use
, and
software
licensing
rules apply.
Abstract
This document describes CC/PP (Composite Capabilities/Preference Profiles)
structure and vocabularies. A CC/PP profile is a description of device
capabilities and user preferences. This is often referred to as a device's
delivery context and can be used to guide the adaptation of content presented
to that device.
The Resource Description Framework (RDF) is used to create profiles that
describe user agent capabilities and preferences. The structure of a profile
is discussed. Topics include:
structure of client capability and preference descriptions, AND
use of RDF classes to distinguish different elements of a profile, so
that a schema-aware RDF processor can handle CC/PP profiles embedded in
other XML document types.
CC/PP vocabulary is identifiers (URIs) used to refer to specific
capabilities and preferences, and covers:
the types of values to which CC/PP attributes may refer,
an appendix describing how to introduce new vocabularies,
an appendix giving an example small client vocabulary covering print
and display capabilities, and
an appendix providing a survey of existing work from which new
vocabularies may be derived.
Status of this document
This section describes the status of this document at the time of its
publication. Other documents may supersede this document. A list of current
W3C publications and the latest revision of this technical report can be
found in the
W3C technical reports index
at http://www.w3.org/TR/.
This document is a Recommendation of the W3C. It has been reviewed by W3C
Members and other interested parties, and has been endorsed by the Director
as a W3C Recommendation. It is a stable document and may be used as reference
material or cited as a normative reference from another document.
W3C's role in making the Recommendation is to draw attention to the
specification and to promote its widespread deployment. This enhances the
functionality and interoperability of the Web. A
test
suite
for has been developed, along with an
implementation
report
This document has been produced by the W3C Device Independence Working
Group as part of the
Device Independence
Activity
within the
W3C
Interaction Domain
. Continued status of the work is reported on the
Device Independence Working Group Home Page
Member-only link
).
The public is invited to send comments or reports errors to the editors at
www-mobile@w3.org
, the public forum
for discussion of W3C's work on Mobile Web Access. An archive is available at
Patent disclosures relevant to this specification may be found on the
CC/PP Working Group's
patent
disclosure page
in conformance with W3C policy.
Table of contents
1. Introduction
This document describes CC/PP (Composite Capabilities/Preference Profiles)
structure and vocabularies. A CC/PP profile is a description of device
capabilities and user preferences that can be used to guide the adaptation of
content presented to that device. Here profile does not refer to a subset of
a particular specification, for example the CSS Mobile profile, but refers to
the document(s) exchanged between devices that describe the capabilities of a
device.
As the number and variety of devices connected to the Internet grows,
there is a corresponding increase in the need to deliver content that is
tailored to the capabilities of different devices. Some limited techniques,
such as HTTP '
accept
' headers and HTML '
alt=
attributes, already exist. As part of a framework for content adaptation and
contextualization, a general purpose profile format is required that can
describe the capabilities of a user agent and preferences of its user. CC/PP
is designed to be such a format.
CC/PP is based on RDF, the Resource Description Framework, which was
designed by the W3C as a general purpose metadata description language. RDF
provides the framework with the basic tools for both vocabulary
extensibility, via XML namespaces
[XMLNAMESPACES]
, and
interoperability. There is a specification that describes how to encode RDF
using XML
[RDF]
, and another that defines an RDF schema
description language using RDF
[RDFSCHEMA]
. RDF was designed
to describe the metadata or machine understandable properties of the Web. RDF
is a natural choice for the CC/PP framework since user agent profiles are
metadata intended primarily for communication between user agents and
resource data providers. For an introduction to RDF, see
[RDFPRIMER]
. Note that the
[RDFPRIMER]
document describes a more recent revision
of the RDF specifications than the ones on which this specification is
based.
A CC/PP profile contains a number of CC/PP attribute names and associated
values that are used by a server to determine the most appropriate form of a
resource to deliver to a client. It is structured to allow a client to
describe its capabilities by reference to a standard profile, accessible to
an origin server or other sender of resource data, and a smaller set of
features that are in addition to or different than the standard profile. A
set of CC/PP attribute names, permissible values and associated meanings
constitute a CC/PP vocabulary.
Some information contained in a profile may be sensitive, and adequate
trust and security mechanisms must be deployed to protect users' privacy. As
a part of a wider application, CC/PP cannot fully cover such issues, but is
intended to be used in conjunction with appropriate mechanisms. This topic is
covered in
Appendix F
, (CC/PP applications).
It is anticipated that different applications will use different
vocabularies; indeed this is needed if application-specific properties are to
be represented within the CC/PP framework. But for different applications to
work together, some common vocabulary, or a method to convert between
different vocabularies, is needed. (XML namespaces can ensure that different
applications' names do not clash, but does not provide a common basis for
exchanging information between different applications.) Any vocabulary that
relates to the structure of a CC/PP profile must follow this specification.
The appendices introduce a simple CC/PP attribute vocabulary that may be used
to improve cross-application exchange of capability information, partly based
on some earlier IETF work.
CC/PP is designed to be broadly compatible with the earlier UAProf
specification
[UAPROF]
from the WAP Forum. That is, we have
attempted to accomodate existing UAProf profiles.
CC/PP is compatible with IETF media feature sets (CONNEG)
[RFC2533]
in the sense that all media feature tags and values
can be expressed in CC/PP. However, not all CC/PP profiles can be expressed
as media feature tags and values, and CC/PP does not attempt to express
relationships between attributes.
Although the examples and use to date have been focused on device
capabilities, CC/PP can also convey information about user preferences that,
used sensibly, should be allow web servers to improve the accessibility of
web sites. A fuller discussion of web site accessibility can be found in the
Web Content Accessibility Guidelines
[WAI]
1.1 Scope and normative elements
CC/PP Structure and Vocabularies (abbreviated to CC/PP in the rest of this
document) defines a client profile data format, and a framework for
incorporating application- and operating environment-specific features. It
does not define how the profile is transferred, nor does it specify what
CC/PP attributes must be generated or recognized. CC/PP is designed for use
as part of a wider application framework. As such, the specification of CC/PP
elements that must be supported and those which may be omitted is a matter
for a specific application.
There are few protocol assumptions built into the design of CC/PP.
Although it is intended to be largely protocol independent, particular
consideration has been given to use of CC/PP with HTTP for retrieving Web
resources.
Appendix F
contains some further
discussion of CC/PP applications.
This document describes a number of features of CC/PP. Some features form
part of the essential structure of CC/PP, for which conformance is REQUIRED
(see section
). Others are features whose use is
RECOMMENDED or OPTIONAL. There is also discussion of how new vocabularies
should be introduced, directed to CC/PP application designers rather than
implementers.
The architecture
section
does not describe
specific features, but indicates general principles that underlie the design
of CC/PP. It is not normative but does contain information that should be
understood for proper implementation of CC/PP.
The section on
CC/PP structure
covers two
main areas:
CC/PP profile components: support for these is REQUIRED.
CC/PP profile defaults: support for these is REQUIRED.
The
section
on CC/PP attribute
vocabularies describes some general features of CC/PP attributes and their
values. Support for the described formats for simple attribute values is
RECOMMENDED -- the actual syntax for any simple CC/PP value is defined by the
corresponding attribute specification; such specifications may reference the
information provided here. Support for the structured CC/PP attribute formats
described, where relevant, is REQUIRED.
Support is not required for any specific vocabulary, but application
designers are strongly encouraged to re-use existing vocabularies where
possible.
CC/PP applications are not required to support features described in the
appendices, but any new attribute vocabularies defined MUST be based on RDF
classes and properties defined by the RDF schema in
Appendix B
(new CC/PP attributes are instances of
ccpp
:Attribute
, new component classes are subclasses of
ccpp
:Component
, etc.).
NOTE:
The reason for requiring new vocabularies to be
based on the CC/PP schema is so that schema-aware applications can include
CC/PP profile data along with other RDF data. Having new vocabulary terms
based on the CC/PP schema means that they are clearly identifiable as part
of a CC/PP profile when RDF data from multiple sources is combined. This
requirement does not affect stand-alone CC/PP profile processors, but the
real value of using RDF here will be in the longer term, allowing data from
multiple sources (e.g. document, security and privacy related information)
to be combined and processed by more general purpose handlers.
1.2 Structure of this document
The remainder of this section covers terminology, conventions and
notations used in this document.
Section 2,
CC/PP
architecture
, provides an overview of the CC/PP profile
structure and use of XML namespaces.
Section 3,
CC/PP structure
describes the structure of a CC/PP profile, and introduces the RDF elements
that are used to create the essential CC/PP elements.
Section 4,
Attribute
vocabularies
, describes how attributes are used in a CC/PP
profile, and presents the recommended structure of CC/PP elements used to
describe specific features.
The appendices contain additional supporting material that is not
essential to construct a valid CC/PP profile, but which provides additional
background information useful for understanding CC/PP, its relationship with
RDF, or defining attribute vocabularies for specific applications.
1.3 Document conventions
1.3.1 Terminology
See
CC/PP terminology and abbreviations
in
Appendix A
of this document.
The term "
CC/PP attribute
" is used here to refer to a specific
capability or characteristic of a client (or other system) that appears in a
CC/PP profile. The term "
feature
" refers to a client capability or
characteristic that may or may not be the basis of a CC/PP attribute. The
term "
attribute name
" is used to indicate an RDF property name
used to identify a CC/PP attribute.
The key words "MUST", "MUST NOT", "SHOULD", "SHOULD NOT", "MAY", "MAY
NOT", "REQUIRED", "RECOMMENDED" and "OPTIONAL" in this document are to be
interpreted as described in RFC 2119
[RFC2119]
1.3.2 RDF graph notation
The underlying structure of RDF is a directed labeled graph. For
communication between computer systems, RDF uses a serialization in XML to
represent these graphs. This XML notation is rather bulky and difficult for
human discourse, so a more visual notation is used here for describing RDF
graph structures:
Figure 1-1: RDF graph notation
Subject-resource
] --
propertyName
--> [
Object-resource
Indicates a graph edge labeled '
propertyName
' from an RDF
resource named 'Subject-resource' to another RDF resource named
Object-resource
'.
Subject-resource
] --
propertyName
--> "
Property value
Indicates a graph edge labeled 'propertyName' from an RDF resource
named 'Subject-resource' to a literal string containing the
indicated value.
Subject-resource
] --
propertyName
--> { "
Val1
", "
Val2
", ... }
This is a shorthand for a property whose value is an
rdf:Bag
resource containing the indicated values (see
section
4.1.2.1
).
[<
Subject-type
>] --
propertyName
--> [<
Object-type
>]
Names in angle brackets are used to indicate an RDF resource of the
indicated type (i.e. having the indicated
rdf:Type
property value), without indicating a specific name for the
resource. This is useful for showing the RDF classes that may be
linked by a property.
Subject-resource
] --
propertyName
--> [
Object-resource
-------------------------------
+--
property1
-->
(val1)
+--
property2
-->
(val2)
(etc.)
Property arcs can be chained, and multiple arcs drawn from a
subject resource.
Here are some XML examples of the RDF graph structures described above:
Figure 1-2: RDF graph example in XML
">
propertyName
" />
propertyName
">
propertyName
rdf:resource="
" />
">
propertyName
property value
propertyName
">
propertyName
Val1
Val1
propertyName
">
propertyName
Val1
Val1
propertyName
" />
propertyName
" />
propertyName
">
propertyName
" >
property1
val1
property1
property2
val2
property2
propertyName
2. CC/PP architecture
This section is not normative, but provides an overview of the features of
CC/PP.
2.1 CC/PP profile structure
A CC/PP profile is broadly constructed as a 2-level hierarchy:
a profile having at least one or more
components
, and
each component having at least one or more
attributes
The initial branches of the CC/PP profile tree describe major components
of the client. Examples of major components are:
the hardware platform upon which software is executing,
the software platform upon which all applications are hosted, or
an individual application, such as a browser.
A simple, graphical representation of the bottom of a CC/PP tree based on
three components (
TerminalHardware
TerminalSoftware
and
TerminalBrowser
) would be:
Figure 2-1a: CC/PP profile components
example
:MyProfile]
+--
ccpp
:component-->[
example
:TerminalHardware]
+--
ccpp
:component-->[
example
:TerminalSoftware]
+--
ccpp
:component-->[
example
:TerminalBrowser]
The corresponding XML might look like this:
Figure 2-1b: CC/PP profile components in
XML
xmlns:example="http://www.example.com/schema#">
A CC/PP profile describes client capabilities and preferences in terms of
a number of "CC/PP attributes" for each component.
The description of each component is a sub-tree whose branches are the
capabilities or preferences associated with that component. Though RDF makes
modeling a wide range of data structures possible, including arbitrary
graphs, complex data models are usually best avoided for profile attribute
values. A capability can often be described using a small number of CC/PP
attributes, each having a simple, atomic value. Where more complex values are
needed, these can be constructed as RDF subgraphs. One useful case for
complex attribute values is to represent alternative values; e.g. a browser
may support multiple versions of HTML. A hypothetical profile might look like
this:
Figure 2-2a: Complete CC/PP profile
example
ex
:MyProfile]
+--
ccpp
:component-->[
ex
:TerminalHardware]
| |
| +--
rdf
:type----> [
ex
:HardwarePlatform]
| +--
ex
:displayWidth--> "320"
| +--
ex
:displayHeight--> "200"
+--
ccpp
:component-->[
ex
:TerminalSoftware]
| |
| +--
rdf
:type----> [
ex
:SoftwarePlatform]
| +--
ex
:name-----> "EPOC"
| +--
ex
:version--> "2.0"
| +--
ex
:vendor---> "Symbian"
+--
ccpp
:component-->[
ex
:TerminalBrowser]
+--
rdf
:type----> [
ex
:BrowserUA]
+--
ex
:name-----> "Mozilla"
+--
ex
:version--> "5.0"
+--
ex
:vendor---> "Symbian"
+--
ex
:htmlVersionsSupported--> [ ]
----------------------------
+--
rdf
:type---> [
rdf
:Bag]
+--
rdf
:_1-----> "3.2"
+--
rdf
:_2-----> "4.0"
The corresponding XML might look like this:
Figure 2-2b: Complete CC/PP profile example in
XML
xmlns:ex="http://www.example.com/schema#">
The attributes of a component can be included directly, as in the previous
example, or may be specified by reference to a default profile, which may be
stored separately and accessed using its specified URI.
This use of an externally defined default profile is somewhat similar to
the idea of dynamic inheritance. It makes possible some important
optimizations. As a separate document, it can reside at a separate location
and it can be separately cached. This is particularly useful in wireless
environments such as cellular networks, where the profiles may be large and
the client link slow and expensive. Using default values, only a small part
of the overall profile is sent over the wireless network.
Default values for a component of a CC/PP profile are indicated by a
ccpp
:defaults
arc from the component concerned to a
component that describes the default values.
Figure 2-3a: CC/PP profile using defaults
ex
:MyProfile]
+--
ccpp
:component--> [
ex
:TerminalHardware]
| |
| +--
rdf
:type-------> [
ex
:HardwarePlatform]
| +--
ccpp
:defaults--> [
ex
:HWDefault]
+--
ccpp
:component--> [
ex
:TerminalSoftware]
| |
| +--
rdf
:type-------> [
ex
:SoftwarePlatform]
| +--
ccpp
:defaults--> [
ex
:SWDefault]
+--
ccpp
:component--> [
ex
:TerminalBrowser]
+--
rdf
:type-------> [
ex
:BrowserUA]
+--
ccpp
:defaults--> [
ex
:UADefault]
ex
:HWDefault]
+--
rdf
:type----> [
ex
:HardwarePlatform]
+--
ex
:displayWidth--> "320"
+--
ex
:displayHeight--> "200"
ex
:SWDefault]
+--
rdf
:type----> [
ex
:SoftwarePlatform]
+--
ex
:name-----> "EPOC"
+--
ex
:version--> "2.0"
+--
ex
:vendor---> "Symbian"
ex
:UADefault]
+--
rdf
:type----> [
ex
:BrowserUA]
+--
ex
:name-----> "Mozilla"
+--
ex
:version--> "5.0"
+--
ex
:vendor---> "Symbian"
+--
ex
:htmlVersionsSupported--> [ ]
+--
rdf
:type---> [
rdf
:Bag]
+--
rdf
:_1-----> "3.2"
+--
rdf
:_2-----> "4.0"
The corresponding XML might look like this:
Figure 2-3b: CC/PP profile using defaults in
XML
Device profile referencing defaults:
xmlns:ex="http://www.example.com/schema#">
Defaults for HardwarePlatform:
Defaults for SoftwarePlatform:
Defaults for BrowserUA:
If a given attribute value is applied directly to a component resource,
and also appears on a resource referenced by the
ccpp
:defaults
property, the directly applied
value takes precedence:
Figure 2-4a: Overriding a default value
ex
:MyProfile]
+--
ccpp
:component--> [
ex
:TerminalHardware]
+--
rdf
:type--------> [
ex
:HardwarePlatform]
+--
ccpp
:defaults---> [
ex
:HWDefault]
+--
ex
:memoryMb-------> "32"
ex
:HWDefault]
+--
rdf
:type----> [
ex
:HardwarePlatform]
+--ex:displayWidth--> "320"
+--ex:displayHeight--> "200"
+--
ex
:memoryMb---> "16"
In this example, the default component indicates 16 Mb of memory, but this
value is overridden by the
memoryMb
property applied directly to
the profile component. Thus, in this profile, the
memoryMb
attribute has a value of 32.
The corresponding XML might look like this:
Figure 2-4b: Overriding a default value in
XML
Device profile referencing defaults:
xmlns:ex="http://www.example.com/schema#">
Defaults for HardwarePlatform:
A resource indicated by a default property may appear in a separate
document, in which case an absolute URI reference should be specified for the
default resource. In such cases, the URI part of the default resource
identifier (i.e. not including the fragment identifier part) is used to
retrieve an RDF document containing the default resource description. Thus,
if the default resource is named
, the URI
is used to retrieve an RDF
document, and a resource within that document having the local identifier
#HardwarePlatform
is taken as the default resource. (Such a
resource might be defined within the target document using
about='http://example.com/DeviceProfile#HardwarePlatform'
" or
ID='HardwarePlatform'
". See also section
3.5.
NOTE
: Individual applications may allow relative URIs to
be used. Those that do should specify exactly how the corresponding RDF
document is located.
2.2 Extensibility and
namespaces
CC/PP is extended primarily through the introduction of new attribute
vocabularies.
Any application or operational environment that uses CC/PP may define its
own vocabulary, but wider interoperability is enhanced if vocabularies are
defined that can be used more generally; e.g. a standard extension vocabulary
for imaging devices, or voice messaging devices, or wireless access devices,
etc. Accordingly, this specification defines a small core vocabulary of
features that are applicable to range of print and display agents whose use,
where appropriate, is strongly recommended. This core vocabulary is based on
IETF specification RFC2534
[RFC2534]
, and serves as an
example of how CC/PP attribute vocabularies may be defined. Another such
example is the WAP Forum UAProf specification
[UAPROF]
Any CC/PP expression can use terms drawn from an arbitrary number of
different vocabularies, so there is no restriction caused by re-using terms
from an existing vocabulary rather then defining new names to identify the
same information. Each vocabulary is associated with an XML namespace, as are
the names that describe the underlying RDF and CC/PP structures.
XML namespaces
[XMLNAMESPACES]
define a notation for
associating convenient name forms with arbitrary URIs. The RDF graph syntax
does not specifically employ namespaces, but XML serializations of an RDF
graph do. We also use namespace prefixes when presenting RDF in the graph
notation described above.
The CC/PP framework uses the XML namespace mechanism to create identifying
URIs for RDF core elements, CC/PP structural elements and CC/PP attribute
vocabularies. Consider the following namespace declaration example:
Figure 2-7: Example namespace declarations
xmlns:prf="http://www.wapforum.org/profiles/UAPROF/ccppschema-20010430#">
The first namespace declaration is for RDF usage. The second declaration
names the CC/PP core structural vocabulary, which includes
component
", "
defaults
" and other properties that
are intrinsic to the CC/PP framework. The third namespace declaration names a
component CC/PP properties vocabulary.
NOTE:
Remember that the namespace prefixes are quite
arbitrary: applications MUST NOT assume that the prefix
rdf:
refers to the RDF vocabulary, or that
ccpp:
refers to the
intrinsic CC/PP vocabulary, etc. It is the URI to which a namespace prefix
is bound that matters.
NOTE:
Although namespace names are identified by URI
references, there is no requirement that a schema be available at that URI.
In the above example, the UAProf namespace name is
" yet
there is no schema at that URI. It is generally preferred practice that a
corresponding schema exists at the URL used to identify a namespace, but
this is not a requirement and CC/PP applications MUST NOT assume that such
a schema will exist.
The use of multiple component property vocabularies is allowed and
encouraged. Different user communities and application domains (WAP Forum,
ETSI, MExE, IETF CONNEG, etc.) may define their own property vocabularies.
This is an important mechanism for providing support for the needs of those
communities.
The following namespaces are introduced by the CC/PP framework:
Normative RDF schema defining class declarations for CC/PP, and core
structural properties (listed in
Appendix
B.3
).
Example but non-normative vocabulary for describing simple client
capabilities, with particular relevance to print and display clients
(listed in
Appendix C
).
NOTE:
To retrieve these schemas it is necessary for your
browser to add the header
Accept:text/xml
in the request.
Browsers that do not add this accept header or use the header
Accept:*/*
or variants thereof will receive a HTML page that
notes these are namespaces reserved for the CC/PP Schemas.
3. CC/PP structure
The general structure of a CC/PP client profile is a two-level tree:
components and attributes, with provision for each component to reference an
externally defined set of default attribute values.
3.1 Components
A CC/PP profile contains one or more
components
, and each
component contains one or more attributes. Each component is represented by a
resource of type
ccpp
:Component
(or some RDFS subclass
thereof), and related to the client profile resource by a
ccpp
:component
property. Here, the
ccpp
namespace is
namespace used to qualify
component
MAY be a UAProf
namespace.
The object of a
ccpp
:Component
resource MAY have an
rdf
:type
property (or equivalent RDF structure)
indicating what kind of client component it describes. The example in figure
2-2b is of a profile with an explicit indication of component subtype.
However, CC/PP processors MUST be able to handle profiles that do not contain
component type indicators. As long as the CC/PP attributes used are all
specific to a given component type, a processor will have sufficient
information to interpret them properly. No more than one instance of a
component type should be present for any given profile resource.
If a CC/PP profile uses any attribute that can appear on different
component types, then the type of any component on which such an attribute
appears MUST be indicated by an
rdf
:type
property, or
equivalent RDF. A CC/PP processor MUST be able to use this type information
to disambiguate application of any attribute used.
3.2 Attributes
CC/PP profiles are constructed using RDF
[RDF]
. The RDF
data model represents CC/PP attributes as named
properties
linking a
subject
resource to an associated
object
resource or RDF
literal value.
To describe client capabilities and preferences, the client being
described is a resource whose features are described by labeled graph edges
from that resource to corresponding object values. The graph edge labels
identify the client feature (CC/PP attribute) being described, and the
corresponding object values are the feature values.
Figure 3-1: RDF statement describing a client
attribute
[Client component resource] --attributeName--> (Attribute-value)
CC/PP attribute labels are represented by XML name values (per XML
specification
[XML]
, section 2.3), which may include a
namespace prefix (i.e. a
qualified name
, per XML namespaces
[XMLNAMESPACES]
, section 3). When combined with the
corresponding namespace or default namespace declaration, each label MUST be
mapped to a URI. Thus, CC/PP attribute names are URIs, with XML namespace
syntax used to avoid some of the RDF expressions becoming too cumbersome.
Attribute values may be of simple or structured data types.
Simple data types are discussed in the section
4.1.1
. Each basic data type may support a range
of tests that can be used in the process of determining the suitability of
different resource variants for presentation by a client; e.g. equality,
compatibility, less-than, greater-than, etc.
Structured data types are supported through the use of specific RDF
properties that join simple RDF literal values into composites. Specific
CC/PP semantics for RDF properties used in this way are discussed in the
section
4.1.2
3.3 Defaults
Each component of a client profile may indicate a single separate resource
that in turn indicates a subordinate collection of default attribute values.
This collection of default values can be a separate RDF document that is
named via a URI, or can appear in the same document as the client profile
(though, in practice, there is probably little value in defaults in the same
document). If an attribute in the collection of defaults is also present in
the main part of the client profile, the non-default value takes precedence.
The intent is that a hardware vendor or system supplier may provide default
values that are common to a number of systems in a place easily accessible to
an origin server, and then use the client profile to specify variations from
the common profile. The owner of the product or system operator may be able
to add or change options, such as additional memory, that add new
capabilities or change the values of some original capabilities.
Default values are referenced by the property
ccpp
:defaults
. This name conforms to the
name format recommendations of the RDF model and syntax specification
[RDF]
, Appendix C.1. However, for compatibility with earlier
versions of CC/PP used with UAProf, CC/PP processors SHOULD recognize the
property name
ccpp
:Defaults
(i.e. with capital "D") as
equivalent. Here, the
ccpp
namespace is
namespace used to qualify
defaults
or
Defaults
MAY
be a UAProf namespace.
Defaults can be encoded inline or as separate documents referred to via
URI. Defaults can not be encoded both inline and as a separate document. It
is the responsibility of any server interpreting a CC/PP to combine profiles
with any externally referenced defaults in such a way as to be able to
correctly interpret the profile. A profile with defaults in the same document
is logically equivalent to a profile with the same non-default data and
referenced external document(s) containing the default values. Here is a
simple profile graph using default values:
Figure 3-2a: CC/PP profile using defaults
ex
:MyProfile]
+--
ccpp
:component--> [
ex
:TerminalHardware]
| |
| +--
rdf
:type-------> [
ex
:HardwarePlatform]
| +--
ccpp
:defaults--> [
ex
:HWDefault]
| +--
ex
:displayWidth--> "640"
| +--
ex
:displayHeight-> "400"
+--
ccpp
:component--> [
ex
:TerminalSoftware]
| |
| +--
rdf
:type-------> [
ex
:SoftwarePlatform]
| +--
ccpp
:defaults--> [
ex
:SWDefault]
+--
ccpp
:component--> [
ex
:TerminalBrowser]
------------
+--
rdf
:type-------> [
ex
:BrowserUA]
+--
ccpp
:defaults--> [
ex
:UADefault]
+--
ex
:htmlVersionsSupported--> { "2.0", "3.2", "4.0" }
ex
:HWDefault]
+--
rdf
:type----> [
ex
:HardwarePlatform]
+--
ex
:cpu------> "PPC"
+--
ex
:displayWidth--> "320"
+--
ex
:displayHeight--> "200"
ex
:SWDefault]
+--
rdf
:type----> [
ex
:SoftwarePlatform]
+--
ex
:name-----> "EPOC"
+--
ex
:version--> "2.0"
+--
ex
:vendor---> "Symbian"
ex
:UADefault]
+--
rdf
:type----> [
ex
:BrowserUA]
+--
ex
:name-----> "Mozilla"
+--
ex
:version--> "5.0"
+--
ex
:vendor---> "Symbian"
+--
ex
:htmlVersionsSupported--> { "3.2", "4.0" }
If a component referenced by
ccpp
:defaults
contains an
attribute that is not present on the referencing profile component, then the
effect is as if the attribute value in the default component is applied
directly to the profile component. For example the profile in Figure 3-2a
should be interpreted as describing the same capabilities as shown in Figure
3-2b.
Figure 3-2b: Resolving a CC/PP profile using
defaults
ex
:MyProfile]
+--
ccpp
:component--> [
ex
:TerminalHardware]
| |
| +--
rdf
:type-------> [
ex
:HardwarePlatform]
| +--
ex
:displayWidth--> "640"
| +--
ex
:displayHeight-> "400"
| +--
ex
:cpu------> "PPC"
+--
ccpp
:component--> [
ex
:TerminalSoftware]
| |
| +--
rdf
:type-------> [
ex
:SoftwarePlatform]
| +--
ex
:name-----> "EPOC"
| +--
ex
:version--> "2.0"
| +--
ex
:vendor---> "Symbian"
+--
ccpp
:component--> [
ex
:TerminalBrowser]
------------
+--
rdf
:type-------> [
ex
:BrowserUA]
+--
ex
:htmlVersionsSupported--> { "2.0", "3.2", "4.0" }
+--
ex
:name-----> "Mozilla"
+--
ex
:version--> "5.0"
+--
ex
:vendor---> "Symbian"
And here is the corresponding XML serialization, with the default resource
descriptions coded inline in the client profile description. Note that this
example uses a default namespace for RDF elements, but still must use
explicit namespace prefixes for RDF attributes.
Figure 3-2c: CC/PP profile using inline defaults,
in XML
xmlns:prf="http://example.com/Schema#">
Inline defaults are logically equivalent to defaults contained in an
external referenced document, and such external documents would be a normal
way of providing default values. The following is the XML serialization of
the same profile with references to externally defined defaults:
Figure 3-3: CC/PP profile referencing externally
defined defaults, in XML
xmlns:prf="http://example.com/Schema#">
Each external defaults resource is a separate RDF document referenced by a
URI.
NOTE
A default document uses a
rdf
:Description>
element as its root node. The
rdf
:Description>
is named using an
rdf
:about
whose value is a URI. This URI MUST
correspond to the value of the
rdf
:resource
XML
attribute in the
ccpp
:defaults>
element in the
referencing document. (The default component does not need to be identified
when it occurs inline, as in the first example above.) In the examples of
default documents below, the URLs of the external default values documents
are used. However the default resource URI does not have to be the document
URL, as long as the URI is uniquely identified, the same URI is used in
both the source document and the external default values document, and
there is some way for the processing software to locate and retrieve the
document containing the default resource.
Examples of default documents referenced by the previous example are as
follows:
Figure 3-4: External HardwarePlatform default
values
Document: http://example.com/HWDefault
Figure 3-5: External SoftwarePlatform default
values
Document: http://example.com/SWDefault
Figure 3-6: External BrowseUA default
values
Document: http://example.com/UADefault
3.4 Distinguishing profile
structure from attributes
CC/PP uses namespaces to distinguish the vocabulary associated with the
structure (e.g.
ccpp
:component
) from vocabularies
associated with applications (e.g. TerminalHardware, display).
In this example we use the namespace
",
associated with prefix "
prf:
", to describe properties that are
not defined in the CC/PP or RDF namespaces:
Figure 3-7: XML serialization of CC/PP profile,
with namespaces
xmlns:prf="http://www.wapforum.org/profiles/UAPROF/ccppschema-20010430#">
All RDF resources that relate to the overall structure of CC/PP are
defined in the
ccpp
namespace, and have
associated schema properties that allow them to be distinguished from
attribute vocabulary or other RDF statements by a schema-aware processor.
3.5 Notes on RDF usage
This specification uses "
rdf:about
" to specify the URIs of
resources in examples. This was a deliberate choice to ensure that such URIs
are absolutely and unambiguously specified. This is also different to UAProf,
which uses both "
rdf:about
" and "
rdf:ID
".
CC/PP allows "
rdf:ID
" attributes or "
rdf:about
attributes. However, the values of "
rdf:ID
" attributes represent
URIs which are relative to the base URI of the document
[RDFFRAGMENT]
. When a document is moved to another location on
the web the meaning of the value of an "
rdf:ID
" attribute
changes. The meaning is undefined when the RDF is contained in a document
with no base URI, e.g. when encapsulated in a message. The RDFCore WG have a
Working Draft
[RDFXML]
that proposes that RDF should
support "
xml:base
" attributes. If this addition to RDF achieves
recommendation status, then it would be appropriate to use
rdf:ID
" attributes in conjunction with an
xml:base
" attribute instead of "
rdf:about
attributes. For now we recommend that CC/PP profiles SHOULD use
rdf:about
" and that the URIs of resources are fully
specified.
The component resources in a profile are instances of components
identified in the corresponding schema, which in turn MUST be subclasses of
ccpp
:Component
. They may usefully be
identified as such, by means of the
rdf
:type
property whose value matches the name of the component type in the schema.
(Sometimes this type indication MUST be present: see
section 3.1, Components
.)
The RDF statements that make up an RDF graph do not necessarily occur in a
single document. For CC/PP, the profile delivered may contain references to
RDF subgraphs that are transferred separately, or are retrieved from
designated Web resources.
When an external sub-graph is referenced in this way, the effect is
equivalent to taking the sets of RDF statement "triples" described by the
referencing document and the referenced document, and constructing a new
document that describes the union of these sets. (
NOTE:
implementations are
not
required to actually construct such
a document, just to interpret the RDF statements as they would from a single
document.)
This composition of multiple RDF documents presumes that the content of
the referenced document is trusted to accurately represent the capabilities
that are presented to the sender of some resource data. Accordingly, such
composition is restricted to documents describing resources referenced by
properties whose intended interpretation embodies such a notion of trust;
viz.
ccpp
:defaults
4. Attribute vocabularies
4.1 Attribute data
This section describes the basic data types and data structuring options
that are available for the values associated with a CCPP attribute.
All CC/PP attributes should be defined with values that can be treated as
one of the simple or complex data types discussed later. Support for the
described formats for attribute values is RECOMMENDED; this specification
does not prohibit the use of other valid RDF forms, but provides no guidance
for their interpretation. (See also section
1.1
and
Appendix F
.)
All simple CC/PP attribute values are represented as RDF plain literal
values. In RDF/XML these may appear as character sequences either in XML
elements or as XML attributes.
The acceptable plain literal values for an attribute may be constrained to
the lexical space associated with a specific application data type. This
section introduces some specific data types that may be associated with
simple CC/PP attributes.
Base CC/PP usage defined here leaves any further interpretation of the
values used to the processing application. Future versions of CC/PP may
introduce additional structures that provide for standardized matching of
client profiles with other resource metadata. To allow such developments, and
to ease interworking with IETF media feature descriptions, it is RECOMMENDED
that any simple attribute values should be defined in terms of one of the
data types described below.
All attribute values are ultimately sequences of UCS (Unicode) characters.
It is assumed that character coding issues in specific serializations of the
RDF data are defined by the enclosing XML representation.
NOTE
: Attribute comparison is beyond the scope of this
document, as are specific mechanisms for determining the simple type
corresponding to a given attribute value. Applications are presumed to know
how to deal with any CC/PP attribute that they handle.
Where given, formal syntax expressions use the notation presented in
Section 6 of the XML specification
[XML]
The data type of a CC/PP attribute value may be defined to be a case
sensitive text string.
The RDF literal value is constrained to the lexical space defined in the
"string" XML schema datatype
[XMLSCHEMA-2]
. Any 'lang' tag
is ignored.
In general, such values may be compared for equality or inequality. When
comparing text values, every character must match exactly for equality to be
declared.
Some examples:
Browser name: "Mozilla"
Browser version: "5.0"
The data type of a CC/PP attribute value may be defined to be an integer
number.
The RDF literal value is constrained to the lexical space defined in the
"int" XML schema datatype
[XMLSCHEMA-2]
. Any 'lang' tag is
ignored.
Integer numbers may be positive, zero or negative. They are represented by
a string containing a sequence of decimal digits, optionally preceded by a
' or '
' sign. Leading zeros are permitted and are
ignored. The number value is always interpreted as decimal (radix 10). It is
RECOMMENDED that implementations generate and support integer values in the
range -2147483648 to +2147483647, or -(2^31) to (2^31-1); i.e. integers whose
absolute value can be expressed as a 31-bit unsigned binary number.
Figure 4-2: Syntax for integer numbers
Signed-integer ::= ( '+' | '-' )? Unsigned-integer
Unsigned-integer ::= Digit (Digit)*
Some examples:
The data type of a CC/PP attribute value may be defined to be a rational
number.
In other words, the RDF literal value is constrained to the lexical space
defined below. Any 'lang' tag is ignored.
A rational number is expressed as a ratio of two integer numbers. Two
positive integers are separated by a '
', and optionally preceded by
a '
' or '
' sign.
It is RECOMMENDED that implementations generate and support numerators of
a rational number (the first number, before the '
') in the range 0
to 2147483647 (2^31-1), and denominators (after the '
') in the
range 1 to 2147483647.
Figure 4-3: Syntax for rational numbers
Rational-number ::= Signed-integer ( '/' Unsigned-integer )?
If the denominator is omitted, a value '
' is assumed; i.e. treat
value as an Integer.
Some examples:
1/2
768/1024
-254/100
+2000/65536
NOTE
: The rational number schema described above may be
defined in XML-Schema
[XMLSCHEMA-0]
as follows:
Figure 4-4: Possible XML-Schema for rational
numbers
The canonical lexical representation of any value
will be the form of the value reduced to its lowest
common denominator, and with '1' in the denominator
if applicable.
Note that while the pattern above provides a lexical definition, it does
so imperfectly: it strictly disallows any whitespace at all. Further, the
simple type definition above does not define a numeric value space;
ordering, equality, and implied support for arithmetic operations are not
defined as some users of the type might expect -- processors need only
recognize the definition as a string. Because of these deficiencies, use of
rational numbers as defined here may be harmful to interoperability. (The
XML-Schema Working Group may define a workable rational data type in the
future.)
4.1.2 Complex CC/PP attribute data
In addition to the simple values described above, a CC/PP attribute may have
a complex value expressed in the form of a resource with its own collection
of RDF properties and associated values. Specific data types represented in
this way are:
Set of values
Sequence of values
A profile MUST NOT have multiple occurrences of a single attribute within
a single component. CC/PP attributes that need to have multiple values should
use sets or sequences. Other complex CC/PP attribute values may be
represented by arbitrary RDF resources. A definition of the interpretation of
such values is beyond the scope of this specification.
A set consists of zero, one or more values, all different and whose order is
not significant.
Set values are useful for representing certain types of device
characteristics; e.g. the range of typefaces that can be supported by a
client, or the HTML versions supported by a browser.
A set is represented as an '
rdf
:Bag
', with each member of
the set corresponding to a property of that resource named
rdf
:_1
', '
rdf
:_2
', etc. This construct is
described in section 3 of the RDF Model and Syntax specification
[RDF]
Figure 4-4: RDF representation of set values in
CC/PP
[(Client-resource)]
+--(attributeName)--> [
+--rdf:_1--> (set-member-value-1)
+--rdf:_2--> (set-member-value-2)
+--rdf:_n--> (set-member-value-n)
NOTE
The '
rdf:Bag
' construct
does not require that every contained value be unique. A set cannot contain
duplicate values, so every property of an '
rdf:Bag
' used to
represent a set must have a distinct value.
There is a clear distinction drawn between an attribute that has a single
value, and an attribute whose value is a set with zero, one or more
elements:
Figure 4-5: Attribute with set value containing a
single member
[(Client-resource)]
+--(attributeName)--> [
Compare the above attribute value, which is a set containing one element,
with the following, which is a simple value:
Figure 4-6: Attribute with a simple value
[(Client-resource)]
+--(attributeName)--> (attribute-value)
A sequence consists of zero, one or more values, whose order is significant
in some way.
Sequence values are useful for a range of client features that may be
ordered or ranked in some way; e.g. a list of preferences in some order of
preference. This specification does not define the significance of the
ordering of values. A vocabulary that defines a sequence-valued CC/PP
attribute should also define the significance of the ordering of within the
sequence.
A sequence is represented as an '
rdf
:Seq
', with each
member of the set corresponding to a property of that resource named
rdf
:_1
', '
rdf
:_2
', etc. This construct is
described in section 3 of the RDF Model and Syntax specification
[RDF]
Figure 4-7: RDF representation of sequence values
in CC/PP
[(Client-resource)]
+--(attributeName)--> [
+--rdf:_1--> (sequence-value-1)
+--rdf:_2--> (sequence-value-2)
+--rdf:_n--> (sequence-value-n)
There is a clear distinction drawn between an attribute that has a single
value, and an attribute whose value is a sequence with zero, one or more
elements:
Figure 4-8: Attribute with sequence value
containing a single member
[(Client-resource)]
+--(attributeName)--> [
Compare the above attribute value, which is a sequence containing one
element, with the simple value as shown in figure 4-6 above.
CC/PP attribute names are in the form of a URI. Any CC/PP vocabulary is
associated with an XML namespace, which combines a base URI with a local XML
element name (or XML attribute name) to yield a URI corresponding to an
attribute name. E.g. the namespace URI:
and the core vocabulary name:
type
are combined to yield the attribute name URI reference:
Anyone can define and publish a CC/PP vocabulary extension (assuming
administrative control or allocation of a URI for an XML namespace). For such
a vocabulary to be useful, it must be interpreted in the same way by
communicating entities. Thus, use of an existing extension vocabulary is
encouraged wherever possible; failing this, publication of a new vocabulary
definition containing detailed descriptions of the new CC/PP attributes.
Many extension vocabularies will be drawn from existing applications and
protocols; e.g. WAP UAProf, IETF media feature registrations, etc.
Appendix E
surveys some possible sources of additional
CC/PP vocabularies.
Attribute names are defined, and associated with an XML namespace, using
an RDF schema.
Appendix B
to this document contains an RDF
schema describing terms for use in CC/PP profiles.
Appendix C
contains an example Schema describing a
CC/PP vocabulary.
Appendix D
contains
recommendations for creating a new vocabulary.
A CC/PP processor is not required to understand and process RDF Schema
definitions; it merely needs to understand enough about the CC/PP profile
structure and vocabulary used to perform its job. (A schema-aware processor
may be able to handle CC/PP profiles in other ways, or in combination with
other RDF information, but such behavior is beyond the scope of this
specification.)
5. Conformance
This section explains how to make a
valid
claim
that a product conforms to this specification. Anyone may make a
claim (e.g., vendors about their own products, third parties about those
products, journalists about products, etc.). Claims may be published anywhere
(e.g., on the Web or in product documentation). Claimants are solely
responsible for their claims. If the subject of the claim (e.g., the
software) changes after the date of the claim, the claimant is responsible
for updating the claim. Claimants are expected to modify or retract a claim
if it may be demonstrated that the claim is not valid. Claimants are
encouraged to conform to the most recent specification available.
There are three classes of products of CC/PP:
documents (e.g. a web resource)
producers (e.g. a web client)
consumers (e.g. a web server)
5.1 CC/PP Document Conformance
Documents may exist as resources accessible via a URL, or may be
transmitted as data in a message. A document is CC/PP conformant when it
meets the following criteria:
The document MUST be
valid RDF
serialized
in XML, and be based on one or more vocabularies derived from the RDF
Schema in
Appendix B
. See
section 1
The document MUST use valid syntax for namespace declarations. See
section 2.2
The profile element MUST contain one or more components. See
section 2.1
Each component in the profile MUST contain one or more attributes. See
section 2.1
The component names MAY be in rdf:about or rdf:ID attributes. See
section 3.1
Components MUST be indicated using a ccpp:component property where the
namespace used to qualify component is the CC/PP namespace or a UAProf
namespace. See
section 3.1
Component names, component types, and attribute names must all refer to
different URIs within a profile. See
section
If a component type is given as an element name and as an rdf:type
element, they MUST refer to the same URI. See
section 3.1
Default references MUST be valid URLs. See
section 3.3
Defaults MAY be written as ccpp:defaults or ccpp:Defaults. See
section 3.3
Defaults MUST be indicated using a ccpp:defaults or ccpp:Defaults
property where the namespace used to qualify defaults or Defaults is the
CC/PP namespace or a UAProf namespace. See
section 3.3
Component attributes MAY contain both a default value and a directly
applied value, with the directly applied value taking precedence. See
section 3.3
Components MAY contain inline defaults. See
section 3.3
Components MUST NOT contain both inline and referenced defaults. See
section 3.3
Components MAY reference a default document which does not have an
rdf:type. See
section 3.3
Attributes MAY have sets of values (Bags). See
section 4.1.2.1
Attributes MAY have sequences of values (Seq). See
section 4.1.2.2
Attributes MAY have string values. See
section 4.1.1.2
Attributes MAY have Integer number values. See
section 4.1.1.3
Attributes MAY have Rational number values. See
section 4.1.1.4
A component MUST NOT contain more than one attribute with the same
name. See
section 3.2
Attributes of the same name MAY be in different components. See
section 3.1
Profiles MAY use multiple namespaces for attributes. See
section 2.2
5.2 CC/PP Producer Conformance
A producer is CC/PP conformant when any CC/PP profile document generated
by the producer is a CC/PP conformant document.
5.3 CC/PP Consumer Conformance
A consumer is CC/PP conformant when the consumer accepts any CC/PP
conformant document and extracts CC/PP information. Schema-aware processing
is not required, and therefore, support for the RDF Schema in
Appendix B
by CC/PP consumers is OPTIONAL (see
section 4.3
).
There are two categories of conformance for CC/PP consumers:
Conformant
: a CC/PP consumer can claim to be a
"CC/PP 1.0
conformant consumer"
if it accepts any valid CC/PP profile and
extracts information from it.
Validating
: a CC/PP consumer can claim to be a
"CC/PP 1.0
conformant validating consumer"
if it is conformant and if it rejects
all invalid CC/PP profiles.
NOTE
: A
consumer implementation may be configurable to act as either a conformant
consumer or a conformant validating consumer at different times.
5.4 Conformance Claims
5.4.1 Validity
A conformance claim is valid if it is
well formed
and meets the appropriate
conformance criteria for the applicable product class as given above.
5.4.2 Well-formed
A conformance claim is well-formed if it includes the following
information:
the date of the claim
the product class (document, producer, or consumer)
the consumer category (conformant or conformant validating) if
applicable
the title and dated URI of this document
the product name (identity), including a version, date, or other
identifier that uniquely identifies the product
6. Acknowledgments
This document is a distillation of many discussions of the W3C CC/PP
Working Group, with final amendments introduced by the W3C Device
Independence Working Group. The following were CC/PP Working Group members
for some or most of the period of preparation of this specification, and its
predecessors:
Mikael Nilsson, Ericsson Infotech
Ulrich Kauschke, T-Mobil
Ann Navarro, HTML Writers Guild
Brad Topol, IBM
Franklin Reynolds, Nokia
Graham Klyne, Baltimore Technologies
Noboru Iwayama, Fujitsu Laboratories LTD
Takashi Nishigaya, Fujitsu Laboratories LTD
Lalitha Surayanrayana, SBC Technology Resources
Hidetaka Ohto, W3C (through March 2002) / Panasonic
Simon McBride, DSTC Pty Ltd
Varuni Witana, DSTC Pty Ltd
Chris Woodrow, Information Architects
Johan Hjelm, Ericsson
Barry Briggs, Interleaf
Gerd Hoelzing, SAP
Ted Hardie, Equinix
Serge Rigori, Sun
Ted Wugofski, Phone.com
Kynn Bartlett, HTML Writers Guild
Sandeep Singhal, IBM
Thorsten Kassing, T-Mobil
Larry Masinter, Adobe
During the period when the CC/PP WG was developing the specification,
useful revisions and clarifications were suggested by Yuichi Koike, Stuart
Williams, Sean Palmer and Toni Penttinen. Special thanks are due to Aaron
Swartz for a very thorough and revealing review of the first last call
draft.
Following the handing over of the work to the DI WG, special thanks are
also due to David Ezell (XML Schema WG), Brian McBride (RDF Core WG),
Masayasu Ishikawa (HTML WG), and Lynne Rosenthal (QA WG) for their help in
completing the specification.
The following members of the DI WG also provided assistance in completing
the specification: Stephane Boyera, Roger Gimson, Kazuhiro Kitagawa, Andreas
Schade.
7.1.
Normative References
[XML]
Extensible Markup Language (XML) 1.0 (Second Edition); Tim Bray, Jean
Paoli, C. M. Sperberg-McQueen, Eve Maler; World Wide Web Consortium
Recommendation 6 October 2000:
As amended by: XML 1.0 Second Edition Specification Errata;
specifically
[XMLNAMESPACES]
Namespaces in XML; Tim Bray, Dave Hollander, Andrew Layman; World
Wide Web Consortium Recommendation 14 January 1999:
[RDF]
Resource Description Framework (RDF) Model and Syntax Specification;
Ora Lassila, Ralph Swick; World Wide Web Consortium Recommendation 22
February 1999:
[RDFSCHEMA]
Resource Description Framework (RDF) Schema Specification; Dan
Brickley, R. V. Guha; World Wide Web Consortium Candidate
Recommendation 27 March 2000:
[RDFXML]
RDF/XML Syntax Specification; Dave Beckett; World Wide Web Consortium
Working Draft:
7.2.
Informative References
[RFC2506]
RFC 2506: Media Feature Tag Registration Procedure; K. Holtman, A.
Mutz, T. Hardie; IETF Request for Comments:
ftp://ftp.isi.edu/in-notes/rfc2506.txt
[RFC2533]
RFC 2533: A Syntax for Describing Media Feature Sets; G. Klyne; IETF
Request for Comments:
ftp://ftp.isi.edu/in-notes/rfc2533.txt
[CONNEGMATCH]
A revised media feature set matching algorithm; G. Klyne;
Internet-Draft, work in progress:
draft-klyne-conneg-feature-match-02.txt
[RFC2534]
RFC 2534: Media Features for Display, Print, and Fax; L. Masinter, D.
Wing, A. Mutz, K. Holtman; IETF Request for Comments:
ftp://ftp.isi.edu/in-notes/rfc2534.txt
[UAPROF]
WAP-174:
UAProf
User Agent Profiling Specification (1999)
as amended by WAP-174_100
User Agent Profiling Specification Information Note (2001) Wireless
Application Protocol Forum available at
Also see WAP-248-UAProf Version 20-Oct-2001 available at
[DATASTRUCTURE]
Notes on Data Structuring; C. A. R. Hoare; in Structured Programming,
Academic Press, 1972. ISBN 0-12-2000556-2.
[XMLSCHEMA-0]
XML Schema. Part 0: Primer; David C. Fallside; World Wide Web
Consortium Recommendation 2 May 2001:
[XMLSCHEMA-1]
XML Schema. Part 1: Structures; Henry S. Thompson, David Beech,
Murray Maloney, Noah Mendelsohn; World Wide Web Consortium
Recommendation 2 May 2001:
[XMLSCHEMA-2]
XML Schema. Part 2: Datatypes; Paul V. Biron, Ashok Malhotra; World
Wide Web Consortium Recommendation 2 May 2001:
[SEMANTICTOOLBOX]
The Semantic Toolbox: Building Semantics on top of XML-RDF; Tim
Berners-Lee;
[RFC2531]
RFC 2531: Content Feature Schema for Internet Fax; G. Klyne, L.
McIntyre; IETF Request for Comments:
ftp://ftp.isi.edu/in-notes/rfc2531.txt
[TIFF]
TIFF (Tagged Image File Format) 6.0 Specification; Adobe Systems
Inc.;
[RFC2301]
RFC 2301: File Format for Internet Fax; L. McIntyre, S. Zilles, R.
Buckley, D. Venable, G. Parsons, J. Rafferty; IETF Request for
Comments:
ftp://ftp.isi.edu/in-notes/rfc2301.txt
[MULTIMEDIA]
Multimedia Programming Interface and Data Specifications 1.0
(contains WAVE file format); IBM Corporation and Microsoft Corporation;
[RFC2361]
RFC 2361: WAVE and AVI Codec Registries; E. Fleischman; IETF Request
for Comments:
ftp://ftp.isi.edu/in-notes/rfc2361.txt
[MPEG]
MPEG-4 Overview - (V.14 - Geneva Version), ISO/IEC JTC1/SC29/WG11
N3444 Rob Koenen Overview of the MPEG-4 Standard:
[PWG]
Printer Working Group;
[RFC2566]
RFC 2566: Internet Printing Protocol/1.0: Model and Semantics; R.
deBry, T. Hastings, R. Herriot, S. Isaacson, P. Powell; IETF Request
for Comments:
ftp://ftp.isi.edu/in-notes/rfc2566.txt
[SALUTATION]
Salutation Consortium Specification;
[RFC2119]
RFC 2119: Key words for use in RFCs to Indicate Requirement Levels;
S. Bradner; IETF Request for Comments:
ftp://ftp.isi.edu/in-notes/rfc2119.txt
[MPEG-7]
MPEG-7 Overview (version 8.0), ISO/IEC JTC1/SC29/WG11 N3445 Jos矍.
Mart쭥z (UPM-GTI, ES) Overview of the MPEG-7 Standard:
[RFC2277]
RFC 2277: IETF Policy on Character Sets and Languages; H. Alvestrand;
IETF Request for Comments:
ftp://ftp.isi.edu/in-notes/rfc2277.txt
[RFC2396]
RFC 2396: Uniform Resource Identifiers (URI): Generic Syntax; T.
Berners-Lee, R. Fielding, L. Masinter; IETF Request for Comments:
ftp://ftp.isi.edu/in-notes/rfc2396.txt
[RFC2278]
RFC 2278: IANA Charset Registration Procedures; N. Freed, J. Postel;
IETF Request for Comments:
ftp://ftp.isi.edu/in-notes/rfc2278.txt
[CCPPARCH]
Composite Capabilities/Preference Profiles: Requirements and
Architecture; Mikael Nilsson, Johan Hjelm, Hidetaka Ohto; World Wide
Web Consortium Working Draft 21 July 2000:
[RFC2616]
RFC 2616: Hypertext Transfer Protocol -- HTTP/1.1; R. Fielding, J.
Gettys, J. Mogul, H. Frystyk, L. Masinter, P. Leach, T. Berners-Lee;
IETF Request for Comments:
ftp://ftp.isi.edu/in-notes/rfc2616.txt
[CONCEPTUAL]
Conceptual Structures: Information Processing in Mind and Machine;
John F. Sowa; Addison Wesley, Reading MA, 1984.
[KNOWLEDGE]
Knowledge Representation; John F. Sowa; Brooks/Cole, 2000. ISBN:
0-534-94965-7
[RDFFRAGMENT]
Re: How to address RDF fragment; Ralph R Swick; Message to World Wide
Web Consortium RDF-comments mailing list:
[CCPPEX]
CC/PP exchange protocol;Hidetaka Ohto, Johan Hjelm; World Wide Web
Consortium Note 24 June 1999:
[WAI]
Web Content Accessibility Guidelines 2.0; Wendy Chisholm, Jason
White, Gregg Vanderheiden; World Wide Web Consortium Working Draft 22
August 2002:
[RDFPRIMER]
RDF Primer; Frank Manola, Eric Miller; World Wide Web Consortium
Working Draft 23 January 2003:
Appendix A: Terminology and abbreviations
A.1 Terminology
This appendix is INFORMATIVE.
Attribute, or CC/PP attribute
A CC/PP attribute refers to the data elements describing the profile
and is denoted as an RDF property. Each CC/PP attribute is associated
with a value or a list of values or am RDF resource.
NOTE
: this is quite distinct from an XML attribute;
except where the meaning obvious in context, the term "CC/PP attribute"
is generally used to emphasize this usage.
CC/PP Processor
A CC/PP processor transforms a CC/PP document from its RDF format
into some other format. A CC/PP processor understands CC/PP syntax and
structure, including "defaults", but it does not understand application
semantics associated with CC/PP attributes of CC/PP components.
CC/PP Repository
A server that stores the user agent profile or profile segments
persistently in a form that may be referenced by and incorporated into
a profile. A CC/PP repository is typically a Web server that provides
CC/PP profiles or profile segments in response to HTTP requests.
Cacheable
A data resource is said to be "cacheable" if the data resource
contains a property that allows a server to determine whether the
cached resource matches a request for a similar resource.
Cache
A storage area used by a server or proxy to store data resources that
have been retrieved or created in response to a request. When a new
request for a "cached" data resource is received, the server or proxy
can respond with the cached version instead of retrieving or creating a
new copy.
Capability
An attribute of a sender or receiver (often the receiver) which
indicates an ability to generate or process a particular type of
message content. See also "CC/PP Attributes".
Client
An entity that is the original compositor of a CC/PP profile.
Confidentiality
Protecting the content of a message from unauthorized disclosure.
Content Generation
For the purpose of this specification, "content generation" refers to
generating content appropriate to the user agent profile of the request
by using the user agent profile as input to a dynamic content
generation engine. The XSL and style sheets of the document are used to
tailor the document to the user agent profile of the request.
Content Negotiation
The mechanism for selecting the appropriate representation when
servicing a request. The representation of entities in any response can
be negotiated (including error responses).
Content Selection
For the purpose of this specification, "content selection" refers to
selecting an appropriate document from a list of possible choices or
variants by matching the document profile with the user agent profile
of the request.
Content Provider
A server that originates content in response to a request.
Data Resource
A data object that can be transferred across a network. Data
resources may be available in multiple representations (e.g. multiple
languages, data formats, size, resolutions) or vary in other ways.
Document
For the purpose of this specification, "document" refers to content
supplied in response to a request. Using this definition, a "document"
may be a collection of smaller "documents", which in turn is a part of
a greater "document".
Document Profile
Document profiles offer a means to characterize the features
appropriate to given categories of user agents. For instance, one
profile might include support for style sheets, vector graphics and
scripting, while another might be restricted to the tags in HTML 3.2.
Document profiles can be used by servers to select between document
variants developed for different user agent categories. They can be
used to determine what transformations to apply when such variants are
not available. Content developers can use document profiles to ensure
that their Web sites will be rendered as intended.
Dynamic Content
Content that is generated in response to a request. This may be used
for content that depends on changing environmental factors such as time
(e.g., stock quotes) or place (e.g., nearby gas stations)
Feature
Functional property of a device or entity.
Gateway
Software that is capable of bridging disparate network protocols. For
the purposes of this specification, "gateway" refers to protocol
bridging functionality, which may exist in a stand-alone gateway or may
be co-located with a proxy or origin server.
Hint
A suggestion or preference for a particular option. While this option
is strongly recommended, its use is not required.
Machine Understandable
Data that is described with tags that associate a meaning to the data
(i.e., an "author" tag would describe the author of the document),
allowing data to be searched or combined and not just displayed.
Namespace
A qualifier added to an XML tag to ensure uniqueness among XML
elements.
Negotiate Content
Message content that has been selected by content negotiation.
Negotiation Metadata
Information which is exchanged between the sender and the receiver of
a message by content negotiation in order to determine the variant
which should be transferred.
Non-variant Content
When the form/format of the content being sent does not depend on
receiver's capabilities and/or preferences
Origin Server
Software that can respond to requests by delivering appropriate
content or error messages. The origin server may receive requests via
either WSP or HTTP. Application programs executing on the origin server
deliver content that is tailored in accordance with the CC/PP that can
be found within the provided Profile. For the purpose of this
specification, "origin server" refers to content generation
capabilities, which may physically exist in a stand-alone Web server or
may be co-located with a proxy or gateway.
Preference
An attribute of a sender or receiver (often the receiver) which
indicates a preference to generate or process one particular type of
message content over another, even if both are possible.
Preventing the unintended or unauthorized disclosure of information
about a person. Such information may be contained within a message, but
may also be inferred from patterns of communication; e.g. when
communications happen, the types of resource accessed, the parties with
whom communication occurs, etc.
Profile
An instance of the schema that describe capabilities for a specific
device and network. A profile need not have all the attributes
identified in the vocabulary/schema.
Proxy
Software that receives HTTP requests and forwards that request toward
the origin server (possibly by way of an upstream proxy) using HTTP.
The proxy receives the response from the origin server and forwards it
to the requesting client. In providing its forwarding functions, the
proxy may modify either the request or response or provide other
value-added functions. For the purposes of this specification, "proxy"
refers to request/response forwarding functionality, which may exist in
a stand-alone HTTP proxy or may be co-located with a gateway or origin
server.
RDF Resource
An object or element being described by RDF expressions is a
resource. An RDF resource is typically identified by a URI.
Receiver
A system component (device or program) which receives a message.
Schema, RDF Schema
An RDF Schema denotes resources which constitute the particular
unchanging versions of an RDF vocabulary at any point in time. It is
used to provide information (such as organization and relationship)
about the interpretation of the statements in an RDF data model. It
does not include the values associated with the attributes.
Security
Describes a set of procedures applied to data communications to
ensure that information is transferred exactly as the sender and
receiver intend, and in no other way. Security generally breaks down
into Integrity, Authentication, Confidentiality and Privacy.
Sender
A system component (device or program) which transmits a message.
User
An individual or group of individuals acting as a single entity. The
user is further qualified as an entity who uses a device to request
content and/or resource from a server.
User agent
A program, such as a browser, running on the device that acts on a
user's behalf. Users may use different user agents at different
times.
User agent profile
Capabilities and preference information pertaining to the
capabilities of the device, the operating and network environment, and
users personal preferences for receiving content and/or resource.
Variant
One of several possible representations of a data resource.
Variant Content
When the form/format of the content being sent depends on receiver's
capabilities and/or preferences
Vocabulary
A collection of attributes that adequately describe the CC/PP. A
vocabulary is associated with a schema.
CC/PP
Composite Capabilities/Preferences Profile
CC/PPex
CC/PP Exchange Protocol
CONNEG
Content Negotiation Working Group in the IETF
ER
Entity-Relationship
HTML
HyperText Markup Language
HTTP
HyperText Transfer Protocol
HTTPex
HTTP Extension Framework
IANA
Internet Assigned Numbers Authority
IETF
Internet Engineering Task Force
IOTP
Internet Open Trading Protocol
LDAP
Lightweight Directory Access Protocol
OTA
Over The Air, i.e. in the radio network
RDF
Resource Description Framework
RFC
Request For Comments
TBD
To Be Determined
TCP/IP
Transmission Control Protocol/Internet Protocol
UAProf
WAP User Agent Profile
W3C
World Wide Web Consortium
WAP
Wireless Application Protocol
WBXML
WAP Binary XML
WML
Wireless Markup Language
WSP
Wireless Session Protocol
XHTML
Extensible HyperText Markup Language
XSL
Extensible Stylesheet Language
XML
Extensible Markup Language
Appendix B: RDF schema for structure
This appendix is NORMATIVE, but support by CC/PP processors is
OPTIONAL.
Figure B-1: CC/PP class hierarchy
rdfs:Resource
ccpp:Profile {Profile deliverable to origin server}
ccpp:Component
rdfs:Literal
ccpp:string {A text value of a CC/PP attribute}
ccpp:integer {An integer value of a CC/PP attribute}
ccpp:rational {A rational number CC/PP attribute value}
rdf:Bag {A set value for a CC/PP attribute}
rdf:Seq {A sequence value for a CC/PP attribute}
rdf:Property
ccpp:Property {A property applied to a CCPP:Resource}
ccpp:Structure {A structural property in a CC/PP profile}
ccpp:Attribute {A property denoting a CC/PP attribute}
Figure B-2: CC/PP structural properties
ccpp:component Domain=ccpp:Profile, Range=ccpp:Component
ccpp:defaults Domain=ccpp:Component, Range=ccpp:Component
(Schema URI:
Figure B-3: RDF schema for CC/PP classes and core
properties
]>
xmlns:rdfs = '&ns-rdfs;'
xmlns:ccpp = '&ns-ccpp;'>
This class is any complete profile that can be delivered to an
origin server or other system that generates content for a client.
A base class for any collection of CC/PP attribute values.
A CC/PP client profile consists of one or more components,
typically using a derived class that indicates the use of the
component (e.g. prf:HardwarePlatform, prf:SoftwarePlatform).
This is the class of RDF Literals that represent CC/PP
attribute string values.
This is the class of RDF Literals that represent CC/PP
attribute integer number values.
This is the class of RDF Literals that represent CC/PP
attribute rational number values.
ccpp:Property is the super-class for ccpp:Structure and
ccpp:Attribute. Therefore all property arcs that are not part
of the core RDF namespace and constitute parts of a CC/PP
profile are defined as subclasses of ccpp:Property. This
allows schema-validating environments with language mixing to
isolate the CC/PP elements of an RDF graph rooted in some
given resource from other attributes of that resource.
All properties that are structural elements of a CC/PP profile
are defined as instances of ccpp:Structure. This allows
structural combining elements of a profile to be distinguished
from attributes in a schema-aware environment.
All properties that describe client capabilities or preferences
in a CC/PP profile should be defined as instances of
ccpp:Attribute. This allows structural combining elements
of a profile to be distinguished from client features in a
schema-validating environment.
Indicates a component of a top-level client profile.
This property indicates a Component that contains default
properties for some other component. That is, any attributes
that are not found in the subject resource but are present in
the object resource may be incorporated from the object into
the resulting CC/PP profile.
Same as 'defaults'.
Defined as sub-property for backwards compatibility with UAProf
Use of this is deprecated: use 'defaults' instead.
Appendix C: CC/PP attribute vocabulary for print
and display
The appendix is OPTIONAL and INFORMATIVE.
Designers of CC/PP applications who need to describe such features are
encouraged to use this vocabulary rather than define new terms. This
vocabulary is based in part on work done in the IETF media feature
registration (CONNEG) Working Group
[RFC2534]
The client attribute names defined below may be used to identify some
common features associated with client devices that print or display visual
information, such as text or graphics. They are described using XML namespace
local parts, which are further qualified by the XML namespace identifier
(These attributes apply to presented capabilities of the client rather than
to a specific internal component or aspect of a client system.)
deviceIdentifier
(Value data type: String) A URI that serves as an identifier of the
client device or user agent type.
type
(Value data type: set of Strings) A MIME content type that can be
accepted and presented by a client. Similar in principle to the HTTP
'accept:' header, but specifying a single MIME content-type, without
associated content-type parameters. Multiple accepted content-types can
be described by a value that is a set of content-type string values.
Where needed, content-type parameters can be expressed by additional
CC/PP attributes.
schema
(Value data type: set of Strings) A URI that identifies a schema that
is recognized by the client. The schema may be an XML DTD
[XML]
, XML Schema
[XMLSCHEMA-1]
, RDF
Schema
[RDFSCHEMA]
or any other applicable document
structure that can be identified by a URI. A
Schema
value
refines any acceptable document type indicated by the
Type
attribute, but its meaning must not depend on the value of
Type
Typically, this will be used to indicate specific XML DTDs or schema
that are recognized within
text/xml
or
application/xml
document types.
cha
rW
idth
(Value data type: Integer) For a text display device (type="text/*"),
the width of the character display. For non-proportional font displays,
the number of display cells. For non-proportional font displays as
typically used in East Asia, the number of half-width display cells
(ideographic characters and other full-width characters typically
occupy two display cells). For proportional font displays, the width of
the display in ens (where an en is the typographical unit that is the
width of an en-dash/letter 'n').
charHeight
(Value data type: Integer) For a text display device
type="text/*"
), the number of lines of text that can be
displayed (i.e. the display height in characters).
charset
(Value data type: set of Strings, per
[RFC2278]
For a text handling device, a character encoding that can be processed
(values per MIME 'charset' parameter on content-type
"text/*"
).
NOTE:
the term "charset" is a
historical misnomer, and does not necessarily indicate a repertoire of
characters that can be displayed, just an encoding. In some cases,
though, the encoding may imply a repertoire.
pix-x
(Value data type: Integer) For an image display device
type="image/*"
), the number of horizontal pixels that can be
displayed.
pix-y
(Value data type: Integer) For an image display device
type="image/*"
), the number of vertical pixels that can be
displayed.
color
(Value data type: String, per
[RFC2534]
) For text
and image display devices, an indication of the color capabilities (per
RFC 2534
[RFC2534]
, possible values are
binary
", "
grey
", "
limited
",
mapped
" and "
full
").
NOTE
the
color
attribute
provides a very coarse indication of color capabilities, sufficient for
a range of simple applications, and may be refined by additional
attributes where capabilities need to be described in greater
detail.
Figure C-1: CC/PP client vocabulary
properties
ccpp-client:deviceIdentifier Domain=ccpp:Component, Range=ccpp:string
ccpp-client:type Domain=ccpp:Component, Range=rdf:Bag
ccpp-client:schema Domain=ccpp:Component, Range=ccpp:string
ccpp-client:charWidth Domain=ccpp:Component, Range=ccpp:integer
ccpp-client:charHeight Domain=ccpp:Component, Range=ccpp:integer
ccpp-client:charset Domain=ccpp:Component, Range=rdf:Bag
ccpp-client:pix-x Domain=ccpp:Component, Range=ccpp:integer
ccpp-client:pix-y Domain=ccpp:Component, Range=ccpp:integer
ccpp-client:color Domain=ccpp:Component, Range=ccpp:string
(Schema URI:
Figure C-2: RDF schema for client
vocabulary
]>
xmlns:rdfs = '&ns-rdfs;'
xmlns:ccpp = '&ns-ccpp;'
xmlns:ccpp-client = '&ns-ccpp-client;'>
A URI that identifies the type of client device or user agent.
A string containing a MIME content-type, or a set of such strings,
indicating the MIME content-types that can be handled.
A URI that identifies a language or DTD that is recognized by
the client, or a set of such URIs.
Specific values of this attribute may be applicable to certain
MIME content types. For example, a URI that is associated with
a resource containing an XML DTD will generally be applicable
only with text/xml or application/xml content types.
For character displays, the number of characters that can be
rendered across the display. For displays using a proportional
font, this is the display width in typographical 'em's.
For character displays, the number of rows of characters that
can be displayed.
For character displays, the MIME 'charset' values that
can be handled.
For raster displays, the width of the display in pixels.
For raster displays, the height of the display in pixels.
For display or print devices, an indication of the color
rendering capabilities:
binary - indicates bi-level color (black-and-white, or similar).
grey - indicates gray scale capability, capable of sufficient
distinct levels for a monochrome photograph.
limited - indicates a limited number of distinct colors, but
not with sufficient control for displaying a color
photograph (e.g. a pen plotter, high-light printer or
limited display).
mapped - indicates a palettized color display, with enough
levels and control for coarse display of color
photographs.
full - indicates full color display capability.
Appendix D: Recommendations for creating a
vocabulary
This appendix in INFORMATIVE.
Fundamental to the design of CC/PP is the idea that new client attributes
can be defined, as needed, through the introduction of new vocabularies.
Similarly, new relationships can be introduced through new vocabulary
items, though the introduction of these needs a great deal of care to ensure
their semantics are adequately and consistently defined. A general principle
is that application-neutral CC/PP processors should be able to understand and
manipulate CC/PP relationships without necessarily understanding the CC/PP
attributes to which they refer.
It is recommended that RDF Schema be used, in conjunction with supporting
documentation, to define any new CC/PP vocabulary. The rest of this section
assumes that RDF Schema is being used for defining any new vocabulary. The
previous appendix is an example of this approach.
New vocabularies are introduced through XML namespaces. Their relationship
to other CC/PP vocabulary items can be defined by new RDF schema statements
(which must necessarily augment the core RDF schema for the CC/PP vocabulary
given in
Appendix C
of this document).
D.1 Basic format for all vocabulary items
All vocabulary items used by CC/PP are URIs and optional fragment
identifiers, used as RDF property arc identifiers. Relative URI forms should
not be used. Vocabulary items used for different purposes are generally
associated with different XML namespaces. Some common RDF base classes are
defined so that a schema-aware RDF processor can perform improved analysis of
a CC/PP profile, and separate CC/PP profile elements from other statements
made about any resource that appear in the same RDF graph as a CC/PP profile.
All properties used as CC/PP attributes must be instances of the class
ccpp:Attribute
, which itself is a subclass of
rdf:Property
(That is, the schema defining CC/PP attribute properties should define them
as instances of
ccpp:Attribute
. Thus, a schema-aware processor can
distinguish between properties that are part of a CC/PP profile, and
properties which may be part of an attribute value.)
Each CC/PP attribute is associated with a component of a profile (e.g.
HardwarePlatform, SoftwarePlatform, etc.), and is used as a property of an
instance of the appropriate component resource. All such component resource
types are subclasses of
ccpp:Component
. New
ccpp:Component
based classes may be introduced for new types of
attribute vocabulary, but it is strongly recommended that an existing
ccpp:Component
type be used if one is applicable.
NOTE
: A simple CC/PP parser is not required to be
schema-aware, and its implementation does not need to have knowledge of the
RDF class of any attribute or resource, nor is the profile required to
carry RDF type information. The discussion of classes and schema-aware
processing is in relation to possible future developments of generic RDF
processors that may deal with CC/PP and other RDF vocabularies and schemas,
possibly mixed in a single document. For such developments to be possible,
it is important to take account of class and schema issues in the design of
CC/PP, even though simple CC/PP processors need no such awareness.
D.2 Use of XML namespaces
All CC/PP attributes must be associated with a fully resolvable namespace
identifier URI. (Relative URIs, or URIs whose interpretation may depend on
the context in which they are used, should not be used.)
NOTE
: It is anticipated that a namespace URI
used for CC/PP attributes may also be used to identify an RDF or other
schema relating to those attributes. However, such usage is beyond the
scope of this specification.
Typically, new CC/PP attributes will be associated with a new namespace,
which serves (among other things) to distinguish between possible different
uses of the same attribute name local parts. For example,
a:foo
and
b:foo
name quite distinct attributes as long as the prefixes
a:
and
b:
are associated with different namespace
URIs.
D.3 Principles for defining new attributes
Re-using existing vocabularies, where applicable, leverages work that has
already been undertaken and reduces the opportunity for different attribute
names that mean almost but not quite the same thing.
Note that names using different namespaces may be freely mixed in a
profile, so requiring one additional feature is not a good reason to define a
complete new vocabulary.
D.3.2 Attribute value type and interpretation
Attribute definitions should indicate the type and interpretation of the
associated value. Ultimately it is a matter for agreement between generating
and receiving applications how any particular attribute value is to be
interpreted.
Where possible, for ease of processing and compatibility with other
frameworks, attribute values should be based on one of the data types
described in section
4.1
of this document.
Where attributes express a quantity associated with a client, the units of
that quantity should be clearly associated with the attribute definition.
There is no separate mechanism for indicating the units in which an attribute
value is expressed.
D.3.3 Interpretation not dependent on other attribute
values
The meaning of every attribute must be defined in isolation from other
attributes: no attribute may have a meaning that changes dependent on the
value of some other attribute. E.g. an attribute called, say, page-width must
always be expressed using the same units: it is not acceptable for this
attribute to be expressed in characters for some class of device, millimeters
for another, and inches for another. (Note that it is still allowable to
define an attribute that cannot be interpreted unless some other attribute is
also defined; the important principle here is that adding a new attribute
should not invalidate any knowledge of a client that can be deduced from
attributes previously defined.)
Attributes may be defined in "layers", so that simple capabilities (e.g.
ability to handle color photographic images) can be described by a simple
attribute, with additional attributes used to provide more detailed or arcane
capabilities (e.g. exact color matching capabilities).
D.3.4 Attribute naming conventions
Attributes are RDF properties. The RDF Model and Syntax document
[RDF]
, Appendix C, recommends the use of "interCap" name styles
for RDF property names (starting with a lowercase letter, and having 2nd and
subsequent words within a name started with a capital letter and no internal
punctuation). We recommend such style be used for CC/PP attribute names,
except where some other form is preferred for compatibility with other
systems (such as some CONNEG-compatible print and display attributes
described below).
RDF class names used in CC/PP profiles preferably begin with an uppercase
letter.
D.3.5 Attributes should have
specific applicability
If an attribute is defined with a broad range of applicability, problems
could arise if a user tries to apply a single attribute to different parts of
a profile.
An attribute defined very broadly might be subject to different privacy or
security concerns when applied in different circumstances. For example,
having a text-to-voice capability on a mobile phone type of device might be a
generally useful feature, but a similar feature in a PC might be indicative
of a personal disability. Thus a combination of text-to-voice capability and
using a PC-type platform might suggest private information not necessarily
associated with any of the features in isolation.
D.4 Protocol interactions
In some cases, there may be overlaps between CC/PP vocabularies and a
particular protocol with which CC/PP is used. E.g. the client vocabulary
charset
and the HTTP
accept-charset:
header. To some
extent, the protocol-independent nature of CC/PP makes this inevitable, to
the extent that existing protocols may have limited content negotiation
facilities.
When designing vocabularies, avoid defining features which would be
expected to be part of a particular protocol behavior; anything that
describes or relates to a transfer mechanism rather than what is transferred
should be avoided; e.g. support for a feature like HTTP persistent
connections should not be indicated in a CC/PP profile, as (a) it is a
protocol-specific feature, and (b) it doesn't really help the origin server
to select appropriate content for the client.
Similarly, when defining protocol bindings for using CC/PP, interaction
with existing negotiation mechanisms should be considered and specified. A
detailed treatment of this topic is beyond the scope of this
specification.
Appendix E: Review of applicable
vocabularies
This appendix is INFORMATIVE.
This section introduces some possible sources of properties to be
described by CC/PP attribute vocabularies. It is not normative, and is
included to give an idea of some kinds of client feature that CC/PP might be
used to convey.
E.1 IETF media feature registration (CONNEG)
The IETF has defined an IANA registry for media feature tags
[RFC2506]
and a syntax
[RFC2533]
for
relational-style expressions using these to describe client and server media
features. A small common vocabulary has been defined
[RFC2534]
, which has been used as a basis for the CC/PP client
common vocabulary. The IETF Internet fax Working Group has also created
additional registrations to describe the capabilities of fax machines
[RFC2531]
RFC 2506
[RFC2506]
defines three kinds of media feature
tags:
IETF tree
: registered feature tags that are simple names, which
are defined and assigned under the auspices of the IETF standards
process.
Global tree
: registered feature tags that are simple names
preceded by
'g.'
. These are defined by groups other than the
IETF, but are registered with IANA to ensure uniqueness of these
names.
Unregistered
: feature tags that consist of
'u.'
followed by a slightly restricted form of URI.
There is currently a proposal to create a URN namespace for IANA registries.
This would create a mechanism to allow IANA-registered feature tags to be
used directly as URIs in CC/PP expressions.
Unregistered feature tags may be used in CC/PP expressions by stripping
off the leading '
u.
' and taking the resulting URI.
All media feature tags and values can be expressed in CC/PP, but not all
CC/PP profiles can be expressed as media feature tags and values. In
particular, CC/PP text values are case sensitive whereas some media feature
values are case insensitive. Media feature values can be mapped to CC/PP text
values by applying a case-normalization convention (e.g. convert to
lowercase).
This version of CC/PP does not have mechanisms matching those in the IETF
media feature framework, which can be used within CC/PP to state capabilities
in terms of comparisons with fixed values (e.g. '
pix-x<=640
') and
attribute values that appear in certain combinations (e.g.
pix-x=640
' AND '
pix-y=480
' OR '
pix-x=800
' AND
pix-y=600
'). Future work may define such mechanisms.
E.2 WAP UAProf
UAProf
[UAPROF]
is a WAP Forum specification that is
designed to allow wireless mobile devices to declare their capabilities to
data servers and other network components.
The design of UAProf is already based on RDF. As such, its vocabulary
elements use the same basic format that is used for CC/PP.
The CC/PP model follows UAProf, in that each user agent property is
defined as belonging to one of a small number of components, each of which
corresponds to an aspect of a user agent device; e.g.
Hardware platform
Software platform
WAP characteristics
Browser user agent
Network characteristics
Although its RDF schema is more prescriptive regarding class and property
usage than UAProf, the design of CC/PP is backwards compatible. The goal is
that valid UAProf profiles are also valid CC/PP profiles; however not all
CC/PP profiles are necessarily valid UAProf profiles.
E.3 TIFF
TIFF is a raster image encapsulation file format developed and maintained by
Adobe Systems
[TIFF]
. It is also the basis for the standard
file format for Internet Fax
[RFC2301]
As well as pixel-based image data in a variety of coding and compression
formats, TIFF supports a wide range of options for different kinds of
image-related information. These options might be candidate CC/PP attributes.
Many of the TIFF properties relating to image handling capabilities have
already been defined as tags in the CONNEG space as part of the Internet Fax
work
[RFC2531]
; these might best be referenced using URIs
based on their CONNEG tag names.
E.4 WAVE
WAVE is an encapsulation format for audio data, developed and maintained by
Microsoft
[MULTIMEDIA]
There is a registry of WAVE-supported audio codecs that might be used as
CC/PP attributes
[RFC2361]
IETF work in progress for voice messaging (VPIM/IVM) could create IETF
media feature registry tags that are usable by CC/PP profiles through the
same mechanisms described in section
E.1
above.
E.5 MPEG-4
MPEG-4 is an encapsulation format for video data, possibly combined with
audio data, developed and maintained by the ISO MPEG Working Group
[MPEG]
E.6 MPEG-7
MPEG-7 is a metadata format for information associated with image, video,
audio and other data, currently in development by the ISO MPEG Working Group
[MPEG-7]
E.7 PWG
The printer Working Group defines attributes and capabilities applicable to
printing devices
[PWG]
. Some of this work is incorporated
into the IETF Internet Printing Protocol (IPP)
[RFC2566]
E.8 Salutation
Salutation is a protocol and identification scheme for communicating devices,
mainly in a LAN environment, developed and maintained by the Salutation
Consortium
[SALUTATION]
. The device capability
identification mechanisms probably include many items that might be used as
CC/PP attributes.
Appendix F: CC/PP applications
This appendix is INFORMATIVE.
CC/PP is a format framework designed to be used in the context of a wider
application or operational environment. This specification does not define
how to use CC/PP with any particular protocol or application.
This appendix highlights some other issues that application developers
must consider in their designs. Many of these issues may be covered by an
applicable protocol specification used to convey CC/PP profiles.
To make effective use of the CC/PP framework, the operating rules for the
wider environment must specify:
Capability exchange protocol
Trust model
Vocabulary
Security mechanisms
Constraints on allowable attribute value types
Attribute value processing and/or matching rules
Proxy vocabulary and processing
Rules for request profile identification
Additional information to be included with any transmitted resource
data
URI forms allowed for identifying referenced profile documents (e.g.
defaults)
Mechanisms for locating and retrieving referenced profile documents
Interactions with any existing negotiation mechanisms in the host
protocol
There are a few protocol assumptions built into the design of CC/PP.
Although it is intended to be largely protocol independent, some
consideration has been given to use of CC/PP with HTTP for retrieving Web
resources.
F.1 Outline of request processing in HTTP
CC/PP is envisaged to be used with HTTP in the following fashion.
(This is not a protocol specification, just an indication of the kind of
information flows envisaged. Defining a protocol to convey CC/PP information
is a separate effort
[CCPPEX]
).
Figure F-1: HTTP request processing
+------+ (5) (4) +-------+ +------+
|Client| <==response== | Proxy | <==response== |Origin| <====> (Resource)
| UA | ===request==> | | ===request==> |server| (3) ( data )
+------+ (1) | +-------+ (2) | +------+
| |
v v
(Client ) <--- (Client profile) <----- (Request profile)
(defaults) + local values |
(Proxy ) <--- (Proxy profile)
(defaults) + local values
The client sends an HTTP request, with an accompanying CC/PP client
profile. The client profile may contain references to default profiles
describing a range of common capabilities for the client concerned (e.g.
a particular computer/operating system/browser combination, or a
particular model of mobile device), and values that are variations from
the default profile.
The HTTP request may pass through a firewall/proxy that (a) imposes
constraints on the kinds of content that can be accessed, or (b) can
adapt other forms of content to the capabilities of the requesting
client. This proxy extends the CC/PP profile with a description of these
constraints and adaptations, and sends this with the HTTP request on to
the origin server. The request may pass through several such proxies.
The origin server receives the request and interprets the CC/PP
profile. It selects and/or generates content that matches the combined
proxy and client capabilities described in the profile. This is sent to
the last proxy in the request chain in an HTTP response.
If required, the proxy applies any content adaptations, and any other
functions it is designed to perform. The resulting response and content
is passed back toward the requesting client.
The client receives the HTTP response and presents the content it
contains.
NOTE
: There is some overlap between CC/PP and the various
HTTP accept-* headers. A protocol specification for using CC/PP with HTTP
must indicate how HTTP 'accept-*' headers may be used, and how they
interact with CC/PP profiles.
F.2 Protocol assumptions for proxy
behavior
The framework for describing proxy behaviors makes some assumptions about
the protocol used to convey a CC/PP profile:
Appendix G: RDF Compatibility
This appendix is INFORMATIVE.
This CC/PP specification is based on Resource Description Framework (RDF)
Model and Syntax Specification
[RDF]
, a W3C Recommendation.
That version of RDF does not have explicit datatyping of literals. The RDF
specifications have been undergoing revision as this specification is being
written. The revised RDF (
RDF/XML Syntax
Specification (Revised)
), which at the time of writing has not reached
recommendation status, introduces support for specifying the XML Schema
datatype of a literal value. This appendix outlines the implications for
implementors in making their CC/PP implementations compatible with this
proposed addition to RDF. It is hoped that a future version of the CC/PP
specification will propose how explicit datatyping should be used in defining
CC/PP profiles.
In this specification, a CC/PP attribute value contained in a CC/PP
profile is, in RDF (Revised)
[RDFPRIMER]
terminology, an
RDF Plain
Literal
. A CC/PP vocabulary schema (such as the example in
Appendix C
), that is defined in terms of the simple
types introduced in the CC/PP schema (in
Appendix
), can provide additional type information for these attributes. A CC/PP
profile consumer application may use the vocabulary schema (either by
directly interpreting the schema data, or, for known vocabularies, by
embedding the equivalent information in the application) to check the
validity of the data provided in a profile, and to map the data into
programming language data types.
The revised RDF Working Draft supports explicit XSD (
XML Schema Datatypes
) datatyping. When explicit
datatyping is adopted, the CC/PP specification could be revised to permit an
attribute value in a CC/PP profile to be represented as an
RDF Typed Literal
. In the
XML
serialization of an RDF Typed Literal
, the type of the literal value is
specified as an attribute of the element containing the literal value. In
this case, a CC/PP profile consumer application could use this type
information to interpret the CC/PP attribute value, without requiring access
to additional vocabulary schema information.
In the future it may be that, for backward compatibility, a CC/PP profile
consumer should be able to handle CC/PP attributes that use either implicit
or explicit datatyping.
Appendix W: Revision history
20001218
Document created from merge of
architecture, structure and vocabulary documents.
20001219
Move some vocabulary and proxy material
from section to into sections 3 and 4. Various small edits.
20010109
Various editorial fixes. Merge
appendices dealing with print and display vocabulary. Remove some
vocabulary source references. Add XML examples to many of the RDF
graph examples. Reorganize material in sections 2 and 3, moving some
technical detail to section 3. Move discussion of CC/PP applications
to a new appendix. Assign figure numbers.
20010112
More group review editorial comments.
Fixed some schema errors. Moved client schema summary to appendix C.
Updated UAProf reference and namespace URI. Added Working Group
members to acknowledgments.
20010116
More group review editorial comments.
Added citation of RFC2119. Changed some instances of rdf:Bag to {...}
notation in graph descriptions. Use ccpp:defaults consistently in
examples; add note about allowing ccpp:Defaults for compatibility
with UAProf. Section 2.1.3: added some additional text about
references to external defaults. Added points for allowed URIs and
resolution mechanisms in appendix F. Figure 3-12 notation change.
Section 4.1: attempt to further clarify what is required behavior.
Section 4.3, add paragraph about support of RDF schema. Appendix D:
add text recommending use of RDF schema to define vocabularies.
Section 3.1.1, add text about use of rdf:type properties for
ccpp:Component resources. Appendix B: remove references to ccpp:Set
and ccpp:Value.
20010118
Validate RDF examples with SiRPAC (W3C
online facility) and correct errors detected. Fix up some internal
links.
20010129
Add hyperlinked ToC (courtesy of "htmltoc"
by Earl Hood)
20010223
Published as First Working Draft.
20010315
Added last-call section. Changed some
spelling. Published as Last-call working Working Draft.
20010425
Fold in review comments for editorial
matters. Explain use of names for URIs in examples. Remove unused
terms from glossary.
20010510
Remove P3P from glossary; P3P integration
will be discussed in a separate security document. Review all
examples to use full URIs rather than placeholder names, and qualify
all RDF attributes with namespace prefixes. Added cross-reference
from section 2.3 to 3.1.4 and 3.1.5 (CC/PP usage of RDF). Added NOTE
saying why proxy descriptions are not linked directly. Added sequence
value to data types. Remove requirement for an attribute to be unique
across all components of a profile. Added restriction that a single
"Proxy-behavior" must reference a single type of component, and that
the component type should be specified (section 3.2.2). Clarify that
'charset' feature really indicates a character encoding. Appendix F
re-worked to make it clearer that this document does not define an
HTTP binding for using CC/PP. Added 'xml:lang' attributes to schemas
in appendices B and C. Revise interpretation of charWidth to better
cater for international characters.
20010511
Reference XML schema data types, and fix
up some minor schema errors. Introduce new class ccpp:Structure to
clearly distinguish all CC/PP structural properties from attribute
properties. Editorial fix in relationship to UAProf (E.2).
20010522
Revise text relating to CC/PP attributes
that are represented by URI strings: preferred treatment is as RDF
resources rather than literal URI strings (section 4.1.1.1 and
various examples). Discussion of graph composition (new section
3.1.6). Deleted some gratuitous repetition (sections 3.1.1 and
3.1.2).
20010620
Revise all examples to use explicit
namespace prefixes for RDF elements and attributes. Remove unused
ccpp: namespace declaration from some examples.
20010704
Fix small errors in examples 2-3b and
3-2c. Editorial fixes.
20010906
Reword text in 3.1.5 describing use of
rdf:type on ccpp:Component instance. Revise description of
'charWidth' in appendix C to remove some errors with respect to
international characters. Section 4.1.1.1: Added some text indicating
that XML conventions must be used for encoding non-ASCII characters
in URI values (per RDF spec). Section 4.1.2.2: added note that
vocabularies that use sequence values should define the significance
of the ordering. Add text pointing about possible protocol
interactions to appendices D, F and F.1. Checked example of empty
description element with RDF validator.
20011102
Revise the wording in section 4.1.1.1
about encoding of non-ASCII characters in URI values, to reference
the XML specification errata
Updated reference [XML] to cite the errata document. Revision to
section 4.1.2 introduction text to clarify that arbitrary RDF
resources may be used for attribute values, though their meaning is
not defined here. Fix typo.
20020402
Add hyperlinks to schema documents.
20020501
Added the term 'XML attributes' where
necessary to distinguish between CC/PP attributes and XML attributes.
Fixed the local about reference in Figure 2-1b. Added 'section'
hyperlinks where necessary. Removed unnecessary angle brackets around
namespace identifiers. Removed extraneous space from date of Working
Group meeting. Removed the proxy hardware and software examples:
there is no need for a proxy to append its processor or operating
system to a CC/PP request and using this as an example will cause
confusion. Removed some remaining instances of the term 'URI string'.
Changed 'don't reuse existing vocabularies' heading to 'reuse
existing vocabularies'. Removed an unnecessarily capitalized
'Defaults' from Section 3.1.3. Fixed the references so they work:
there was a problem with the square brackets. Changed 'RDF schema' to
'RDF Schema' where it is used to refer to the W3C specification.
20020502
Fixed errors in RDF in Figure 1-2 and
Figure 3-3 identified with W3C RDF validation service.
20020507
Updated section 3.1.5 to resolve issue 16.
Updated section 2.2 to resolve issue 54. Added a paragraph about WAI
to resolve issue 115.
20020520
Updated CC/PP schema to create datatypes
for anyURI, string and integer in the CC/PP namespace that reference
the datatypes in the XML Schema. The previous version of the schema
just referenced the XML schema datatypes. Changed CC/PP namespace to
20020715
Updated the reference section to
distinguish between normative and informative references and use
[AAAA] not [nnnn] reference label format as per W3C Style Guide to
address issue 181. Removed examples using composite literals i.e.
changed all instances of
to
and
to address issue 152. Changed
example profiles so that rdf:about's point to a profile URI rather
than a schema URI as this means the component applies to this
particular profile instance, rather than any device using this
particular schema to address issue 153. Added some text to section
3.1.1 to address issue 175. Inserted Figure 3.2b to explain default
resolution to address issue 180.
20020717
Updated the UAProf URL, specifically in
Section 2.2, and added text proposed by Art Barstow to resolve issue
54. Also updated UAProf examples so they use legal UAProf property
names.
20020719
Removed a remaining instance of the
sentence "This is one of three properties to describe a proxy
behavior." to address issue 37.
20020812
Added a sentence to section 3.2 and three
additional paragraphs to section 3.2.2 to address issue 182.
20020812
Changed a paragraph in section 3.2 to
address issue 182.
20020924
Removed usage of ccpp:Resource to resolve
issue 31.
20021105
Altered text of abstract to indicate proxy
vocabulary is optional. Altered text of introduction to indicate
proxy vocabulary is optional. Added a preceding sentence in section
2.1.4 saying proxy support is optional. Ditto for section 2.2. Moved
Client-profile, Request-profile, Proxy-profile and Proxy-Behavior
from CCPP Schema to proxy Schema. Updated figures B.1 and B.2 and
added two new figures, B.3 and B.4 to reflect this. Due to the data
typing decision made by RDF-Core concerning interpreting literals,
changed all instances of XML Schema datatypes to CC/PP data types.
Removed the sentence "Note that, where available, XML schema
datatypes (xsdt:) are used for literal values [XMLSCHEMA-2]." from
section B.1. Updated URIs to 08 November 2002. Changed definition of
CC/PP profile in Section 2.1 to "A CC/PP profile is broadly
constructed as a 2-level hierarchy: a profile having a number of
components, and each component having at least one or more
attributes." to resolve the concern expressed by the UAProf drafting
committee that the CC/PP definition of a component is incompatible
with UAProf.
20021211
Fixed problem with Figure 2-3b described
in issue 183. Clarified meaning of CC/PP profile for issue 185.
Clarified the status of appendices to address issue 189. Removed the
term "CC/PP expression" from section 2.2 to address issue 191.
Updated the RDF Schema reference to point at the March 2000 Candidate
Recommendation version. Fixed UAProf reference to address issue 198.
Fixed schema in B.3 to address issue 200. Updated introductory
paragraph to Appendix C to address issue 201. Added a definition of a
"CC/PP processor" to appendix B to address issue 202 and replaced
instance of "CC/PP parser" with CC/PP processor.
20021213
Checked usage of optional and recommended
so that instances are capitalized where necessary to address issue
188.
20030203
Fixed the use of both
adopted the use of rdf:about in the schemas.
20030212
Fixed problem with section 2 in toc. Fixed
bold # in CC/PP namespace in section 2.2. Fixed grammatical error in
section 4.1.2.2.
20030320
Added mention of RDF data typing and
further DIWG work to status section. Added mention of compatibility
with CONNEG to section 1. Removed use of compound simple types in
examples in section 2.1.3. Removed section 2.1.4 (introduction to
proxies), section 2.3 (rdf primer), section 3.2 (proxies). Ensured
the document references the RDF specs (RDF M&S, RDF/XML Syntax,
RDF Primer, RDF Schema) in appropriate places. Promoted section 3.1
subsections to section 3 subsections. Removed case-insensitive text
from section 4.1.1.2. Changed minimum recommended value for integer
numbers and corrected the XML schema datatype in section 4.1.1.3.
Described difficulty of rational numbers in section 4.1.1.4.
Clarified expressive relationship with CONNEG in section E.1.
20030728
Removed remnants mentioning proxy. Added
note about RDF Primer referencing more recent versions of RDF specs
in status section. Changed schema URL from
namespace compatibility for component and defaults in section 3.1 and
3.3. Changed HTML 3.0 to HTML 3.2 in various examples. Removed Tokens
from section 4.1.1.2 and Appendix C. Fixed rational number schema in
section 4.1.1.4. Clarified that multiple occurences of a single
attribute within a single component are not allowed in section 4.1.2.
Added section 5 Conformance. Changed references to point to dated
copies in section 7. Removed anyURI from schema in Appendices B and
C. Changed Rational datatype to rational in Appendix B. Added note
about case normalization for CONNEG compatibility in section E.1.
Added Appendix G RDF Compatibility.
20030915
Various editorial changes and
clarifications. Added "delivery context" term to section 1. Removed
redundant paragraph regarding namespace URIs in section 2.2. Removed
section 4.1.1.1 "Values described by URIs" since it should have been
removed when anyURI was removed.