W3C in a Changing World

W3C goals

sounds good!
here is how we do it

the typical compute power, memory, storage and bandwidth is increased by factor of some 1000
some 100 Web servers --> countless
new applications, multimedia, ...
impacts our work and leisure time

from
- the triple
  - HTTP to communicate
  - URL to identify resources
  - HTML as a document language
we went to

teaching and evangelism
research on our subjects
volunteer translators of our documents
promotion of our technologies by other organisations
public funding of projects
- helped us to set-up the Offices program
- support for technology demonstrators
- also WAI and SW received a lot of support
use of our technologies in legislation (WAI)

it's all true!

(but sounds a bit like a fairy-tale)

ideally it would work like this
- proposals for new technologies and standards are brought into W3C by members or public
- W3C discusses with members and public
- consensus on a solution that all most consider the best and all can live with
but we don't live in an ideal world
- we see friendly groups or even members not following our main stream
- this gives us hard times some time
- we often don't like it - at least not in the beginning
- but sometimes it is helpful

this is more juicy than the perfect harmony stuff

so here are a few selected examples

WAI and especially WCAG 1 (1999) was a great success
- picked-up by many organizations of disabled people
- governments used it for accessibility legislation
- often slightly modified - problem to accept W3C as an expert on this
WCAG 2 (Dec 2008) took quite some time to mature
- ambitious in goals
- hard to find consensus
- work started in Jan 2001
- 15 working drafts, or candidate recommendations before rec
- people started to make jokes at us (April Fool 2008: "WCAG 2 is rec")
- but when it finally was done they loved it, lot of work in translating it
- many authorized translation on the way

1997 W3C presented RDF as standard for metadata on the web
- seen as base for the semantic web
- initiated by some work from Netscape
1999 topic maps for ontologies ISO standard 13250
- language for knowledge representation
- predecessor hi-time 1993 (SGML world)
RDF and Topic maps compete for semantic Web, lot of noise
note from a W3C group on "A Survey of RDF/Topic Maps Interoperability Proposals" (2006)
"An OWL DL construction for the ISO Topic Map Data Model" (Anne Cregan) - is there a resolution?

May 2001 W3C presents XML Schema
- basis for defining XML languages
- validate XML against a schema
may 2001 OASIS presents RELAX NG
- basis for defining XML languages
- validate XML against a schema
RELAX NG is simpler
XML Schema allows more sophisticated constructs
XML Schema has a wide range of datatypes
RELAX NG has only two datatypes: token and string
- but can attach type libraries
- or borrow from XML Schema
XML Schema offers a Post Schema Validation Infoset (PSV)
- for good (additional information e.g. about datatypes or defaults)
- or bad (infoset schould not depend on validation) - no defaults in RELAX NG
questions
- is there a perfect (formal e.g. programming) language for a given purpose?
- aren't formal metalanguages always incomplete?
- is the existance of two languages a threat to interoperability?
- meanwhile RELAX NG has friends also within W3C
- coexistance seems possible

while W3C was preaching Semantic Web(SW) there suddenly was this Web 2.0 hype
- Web 2.0 claimed to keep some of the SW promises
- APIs allowed the exchange of data between Web sites
- Microformats provided a means to represent data on the Web
- tagging as easy metadata generation
- lot of new Web (killer) applications popped up
- new paradigms: social networks, long tail, mass intelligence ...
sad for the Semantic Web and RDF? not if we look closer
- it showed the power of data in the web
- proprietary APIs will not scale in the long range
- neither will the centralized rather unflexible approach of microformats
- it raised the interest for doing the homework ... specifying ontologies etc.
Web 2.0 amongst others generated a tailwind for Semantic Web
- RDF(S) and OWL used to be popular in the scientific arena before
- Linking Open Data gave an impressive example of how data can be linked

there was HTML 4 and there was XML
- a huge success for XML after released in 1997
- W3C decides to reformulate of HTML in XML
  - use of name spaces allows easily compound documents
  - XML cornerstone of W3C architectue
- XHTML 1 mostly backward compatible with HTML4
- XHTML 2 should tidy up issues with HTML 4, not backward compatible
- goal: every browser that supports XML, XFORMS and CSS can speak XHTML 2
- XHTML 2 an XML language
  - open for other XML formats
  - obeying accessability requirements
  - supporting better document structure and other design virtues
there was HTML 4 and the real Web
- concerns inside W3C (2002)
  - Mark Pilgrim: How liberal is too liberal?
  - Daniel Glazman: I strongly suggest dropping all "XHTML 2.0" efforts in favor of a new "xHTML 5.0 ... successor to HTML 4, feature-oriented, made for the web"
- 2004 David Baron: If the W3C does not act, the problem will have to be solved either by some other standardization body, WHATWG launched
2006 W3C Tim Berners-Lee announces the plan for a a completely new working HTML group to standardize HTML 5
- open for (nearly everybody)
- work with WHATWG (Ian Hickson Editor)
What exactly is different with HTML 5?
- open for (nearly) everybody
- work with WHATWG (Ian Hickson Editor)
- HTML 5 design
  - Support existing content
  - Ensure interoperability
  - Handle errors (non-draconically)
  - Evolution not revolution
For W3C a revolution in many ways
http://esw.w3.org/topic/HTML/history