What does XML have to do with Immanuel Kant?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

• first off the block
• improved reliability
• innovation
• simplicity

The Web succeeded because it was simple.

Why was it made simple?

Why the World Wide Web, and not (e.g.) Guide or InterMedia or System-G?

• first off the block
• improved reliability
• innovation
• simplicity

The Web succeeded because it was simple.

Why was it made simple? In order to scale.

Simplicity is not enough.

Consider the story of Gopher. And remember Albert Einstein.

Our mission: to lead the Web to its full potential.

Our goals:

• universal access
• semantic Web
• trust
• interoperability
• evolvability (through simplicity, modularity, compatibility, extensibility)
• decentralization
• cooler multimedia

Because of our goals, we care about

A decentralized world-wide organization.

• members from Europe, North America, and Asia
• Working Groups, Advisory Committee
• technical staff
• host organizations
  • ERCIM (European Research Consortium for Informatics and Mathematics), Sophia-Antipolis
  • Keio University, Tokyo
  • Massachusetts Institute of Technology, Cambridge, Mass.
• offices in many regions: Australia, Benelux, Finland, Germany/Austria, Greece, Hong Kong, Hungary, Israel, Italy, Korea, Morocco, Spain, Sweden, United Kingdom and Ireland.
  • outreach, publicity, workshops, etc.
  • translations (e.g. the books XML & Co and XHTML, CSS & Co, http://www.edition-w3c.de/#Buecher)

As customers, demand that your vendors support open standards!

As users, read and comment on our draft specifications!

As developers, read and comment on our draft specifications, and then implement them!

As institutions, join the W3C to give users and developers a stronger voice within the organization!

First, the heritage of SGML (Standard Generalized Markup Language, ISO 8879):

• generic / generalized / descriptive markup (vs. application-specific)
  • for data reuse
  • for data longevity
• (therefore) declarative markup (vs. imperative)
• information has validatable structure
• markup semantics are defined by application, not centrally
• (ideally) markup semantics fit the application

Note the social pattern: user owns data, not application software owns data.

Also:

• simplicity

XML is SGML without some of the cruft.

Some dates: Has XML lost its simplicity?

The syntax of XML is simple:

<doc>
  <rule number="1">Everything is delimited.
    <elucidation>
      <subrule>Elements are delimited by <term>start-</term> 
        and <term>end-tags</term>.</subrule>
      <subrule>Tags are delimited by angle brackets.</subrule>
      <subrule>Attribute values are delimited by 
        quotation marks.</subrule>
    </elucidation>
  </rule>
  <rule number="2">Everything nests.
    <elucidation>
      <subrule>Elements occur within elements.</subrule>
      <subrule>Attributes are specified within start-tags.</subrule>
    </elucidation>
  </rule>
</doc>

Three layers of rules governing data.

Distinguish logical and physical structure.

To every thing there is a season. And a spec.

The original plan:

• XML (a lightweight SGML) 1998
  • generic markup
  • validation
• XLink / XPointer (lightweight HyTime) 2002, 2003
  • stand-off links
  • n-way linking
  • structure-based addressing
• XSL (lightweight DSSSL) 1999, 2001
  • structure-based formatting and rendering
  • tree transformations
  • flow objects
• XML applications / XML-based languages (SMIL, XHTML, MathML, ...)

An early addition: XML Namespaces 1.0.

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

A conceptual stack for markup (or data representation generally).

XML Information Set: what is information, what is insignificant variation

XML 1.1: alignment with Unicode 3

XML Namespaces 1.1: improve processing on abstract model, re-serialization

A typical data flow diagram.

When you have a good hammer ...

(Possible analogies with Jenny, die Seeräuberbraut, or with Joh. Faustus.)

If XML is a good way to represent complex data, then we'll also need:

• APIs for XML
  • Document Object Model (DOM)
  • Simple API for XML (SAX)
• XML querying
  • XPath
  • XQuery
  • XSLT
• XML-to-XML transformations
  • XQuery
  • XSLT
• XML messages for distributed applications
  • SOAP
  • Web Services description, choreography
  • Digital signatures, encryption
  • ontologies for distributed processes, services, clients, ...
• XML applications (commercial, academic, ...)

And what about

• semantics

Every XML vocabulary entails (implicitly or explicitly) an ontology.

Can we connect them to each other?

• ownership of information
  • freedom from vendor lock-in
  • freedom from pre-defined semantics
  • application independence

• ownership of information
  • freedom from vendor lock-in
  • freedom from pre-defined semantics
  • application independence
• responsibility for information
  • You must think about the processing you need.
  • You must decide what information you wish to capture, and what information you wish to discard.
  Sometimes painful, but worthwhile.

For more information: http://www.w3.org/XML