<?xml version='1.0'?>
<!DOCTYPE slideshow SYSTEM "slide.dtd">
<slideshow>
 
<showtitleslide>
  <title>III. XML Evolution</title>
  <occasion>University of California Extension</occasion>
  <location>Sunnyvale, June 10, 1999</location>
  <speaker>Jon Bosak</speaker>
  <org>Sun Microsystems</org>
</showtitleslide>
 
<slidemodule>
 
  <moduletitleslide>
	 <title>The XML Standards Infrastructure</title>
  </moduletitleslide>
 
  <slide>
	 <head>Key XML standards organizations</head>
	 <bulletlist>
		<item>
		  <p>ISO/IEC JTC1/SC34</p>
		  <bulletlist>
			 <item>
				<p>Maintains SGML, from which XML is derived</p>
			 </item>
			 <item>
				<p>Also DSSSL, HyTime, Topic Maps</p>
			 </item>
		  </bulletlist>
		</item>
		<item>
		  <p>World Wide Web Consortium (W3C)</p>
		  <bulletlist>
			 <item>
				<p>Hosts the "XML Activity"</p>
			 </item>
			 <item>
				<p>Develops other standards related to XML</p>
			 </item>
			 <item>
				<p>Develops other web technologies not directly related to
				  XML</p>
			 </item>
		  </bulletlist>
		</item>
		<item>
		  <p>OASIS</p>
		  <bulletlist>
			 <item>
				<p>Organization for the Advancement of Structured
				  Information Standards</p>
			 </item>
			 <item>
				<p>Formerly SGML Open</p>
			 </item>
			 <item>
				<p>Dedicated to XML interoperability and conformance</p>
			 </item>
			 <item>
				<p>Rapidly expanding role in XML implementation</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>Other significant XML organizations</head>
	 <bulletlist>
		<item>
		  <p>Graphic Communications Association (GCA)</p>
		  <bulletlist>
			 <item>
				<p>Old-time printing standards association</p>
			 </item>
			 <item>
				<p>Historic home of SGML events</p>
			 </item>
			 <item>
				<p>Now hosts XML events such as XML Europe</p>
			 </item>
		  </bulletlist>
		</item>
		<item>
		  <p>eCo Commerce Framework</p>
		  <bulletlist>
			 <item>
				<p>CommerceNet initiative</p>
			 </item>
			 <item>
				<p>Developing methods for the electronic negotiation of XML
				  commercial protocols</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>The W3C</head>
	 <bulletlist>
		<item>
		  <p>Close to 300 members</p>
		</item>
		<item>
		  <p>Working group decisions are only advisory; all final W3C
			 decisions are made by the Director</p>
		</item>
		<item>
		  <p>To protect against premature disclosure of product plans,
			 discussions are mostly visible only to members</p>
		</item>
		<item>
		  <p>XML activity has opened this up significantly through the
			 publication of Requirements Documents</p>
		  <bulletlist>
			 <item>
				<p>Show technical goals that each working group is trying
				  to accomplish</p>
			 </item>
			 <item>
				<p>Publicly visible (http://www.w3.org/TR/)</p>
			 </item>
			 <item>
				<p>Open to public comment</p>
			 </item>
			 <item>
				<p>Process includes periodic "checkpointing" and revision
				  in light of comments</p>
			 </item>
			 <item>
				<p>Process is currently experimental -- XML activity only 
				</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>W3C XML Working Groups</head>
	 <p>The original XML work has been divided up among six W3C working
		groups.</p>
	 <bulletlist>
		<item>
		  <p>In the W3C Architecture Domain:</p>
		  <bulletlist>
			 <item>
				<p><emph>XML Linking WG: </emph>Next-generation hypertext 
				</p>
			 </item>
			 <item>
				<p><emph>XML Schema WG: </emph>Next-generation DTDs</p>
			 </item>
			 <item>
				<p><emph>XML Fragment WG: </emph>Partial XML documents</p>
			 </item>
			 <item>
				<p><emph>XML Infoset WG: </emph>XML object model <i>per
				  se</i></p>
			 </item>
			 <item>
				<p><emph>XML Syntax WG: </emph>XML profiles, stylesheet
				  linking...</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
	 <bulletlist>
		<item>
		  <p>In the W3C User Interface Domain:</p>
		  <bulletlist>
			 <item>
				<p><emph>XSL (Extensible Stylesheet Language) WG:
				  </emph>Industrial-strength layout and formatting language for XML uses
				  on and off the Web</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>Closely related W3C working groups</head>
	 <bulletlist>
		<item>
		  <p>Document Object Model (DOM) WG</p>
		  <bulletlist>
			 <item>
				<p>Not actually an object model (though one is implied)</p>
			 </item>
			 <item>
				<p>Provides a generalized tree-oriented API for HTML and
				  XML document structures</p>
			 </item>
			 <item>
				<p>(There is also a widely used stream-oriented XML API
				  called SAX that was produced outside of W3C.)</p>
			 </item>
			 <item>
				<p>DOM Level 1 done, Level 2 in progress</p>
			 </item>
		  </bulletlist>
		</item>
		<item>
		  <p>I18N (Internationalization) WG</p>
		  <bulletlist>
			 <item>
				<p>Coordinates I18N issues among W3C WGs and between W3C
				  and outside standards organizations (in particular, the Unicode
				  Consortium)</p>
			 </item>
			 <item>
				<p>Responsible for the W3C character model</p>
			 </item>
			 <item>
				<p>Example: is "&#234;tre" &#234; + tre or e + ^ + tre?</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>Other related W3C working groups</head>
	 <bulletlist>
		<item>
		  <p>HTML WG</p>
		  <bulletlist>
			 <item>
				<p>Creating the XML version of HTML: XHTML</p>
			 </item>
			 <item>
				<p>Maximum compatibility with old web browsers (complete
				  compatibility probably not possible)</p>
			 </item>
			 <item>
				<p>Backward compatibility will require users to follow
				  specified guidelines</p>
			 </item>
		  </bulletlist>
		</item>
		<item>
		  <p>Cascading Style Sheets and Formatting Properties
			 (CSS&amp;FP) WG</p>
		  <bulletlist>
			 <item>
				<p>Stylesheets for HTML and simple XML documents</p>
			 </item>
			 <item>
				<p>Lacks ability to transform document structure</p>
			 </item>
			 <item>
				<p>Insufficient for current print formatting or future
				  online display formatting</p>
			 </item>
			 <item>
				<p>Will probably be superseded by XSL for print and
				  industrial-strength web applications</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>XML activity coordination</head>
	 <bulletlist>
		<item>
		  <p>XML Coordination Group</p>
		  <bulletlist>
			 <item>
				<p>Schedules the work of the XML WGs</p>
			 </item>
			 <item>
				<p>Coordinates dependencies between XML specifications</p>
			 </item>
			 <item>
				<p>Interacts with standards efforts outside of W3C</p>
			 </item>
		  </bulletlist>
		</item>
		<item>
		  <p>XML Plenary</p>
		  <bulletlist>
			 <item>
				<p>Union set of participants in the XML working groups</p>
			 </item>
			 <item>
				<p>Decides major XML policy issues</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

</slidemodule>

<slidemodule>
 
  <moduletitleslide>
	 <title>Continuing Work: Schemas and Linking</title>
  </moduletitleslide>
 
  <slide>
	 <head>XML schemas</head>
	 <numberedlist>
		<item>
		  <p>Provide existing DTD functionality in XML syntax</p>
		</item>
		<item>
		  <p>Enhance existing DTD functionality</p>
		  <bulletlist>
			 <item>
				<p>Data typing</p>
			 </item>
			 <item>
				<p>Inheritance</p>
			 </item>
			 <item>
				<p>Namespace validation</p>
			 </item>
		  </bulletlist>
		</item>
	 </numberedlist>
	 <bulletlist>
		<item>
		  <p>The need to improve DTDs has been known for years</p>
		</item>
		<item>
		  <p>Two submissions to start from: DCD and SOX</p>
		</item>
		<item>
		  <p>Also relates to XML query languages</p>
		</item>
		<item>
		  <p>Massive involvement from major software companies,
			 especially those with database or electronic commerce concerns</p>
		</item>
	 </bulletlist>
	 <p>Doing a good job with schemas will take a while!</p>
  </slide>

  <slide>
	 <head>XML linking (XLink and XPointer)</head>
	 <p>XML linking builds on 20 years of hypertext research.</p>
	 <bulletlist>
		<item>
		  <p>Fully extensible -- any element can be a link</p>
		</item>
		<item>
		  <p>Links to collections</p>
		</item>
		<item>
		  <p>Automatic traversal</p>
		</item>
		<item>
		  <p>Transclusion</p>
		</item>
		<item>
		  <p>Addressing by structure</p>
		</item>
		<item>
		  <p>Links outside of documents</p>
		  <bulletlist>
			 <item>
				<p>Links can be managed separately from objects</p>
			 </item>
			 <item>
				<p>Links can be applied to read-only objects</p>
			 </item>
			 <item>
				<p>Links can be made long after the things linked have been
				  released</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>The importance of XML linking</head>
	 <bulletlist>
		<item>
		  <p>Doesn't get much publicity but has deep implications</p>
		</item>
		<item>
		  <p>Allows new ways of associating information</p>
		</item>
		<item>
		  <p>Promotes the creation of advanced information structures and
			 site management</p>
		</item>
		<item>
		  <p>Makes possible an industry devoted to knowledge management 
		  </p>
		</item>
		<item>
		  <p>Keep your eye on XLink and XPointer: http://www.w3.org/TR 
		  </p>
		</item>
	 </bulletlist>
  </slide>

</slidemodule>

<slidemodule>
 
  <moduletitleslide>
	 <title>Continuing Work: Infoset, Fragments, and Profiles</title>
  </moduletitleslide>
 
  <slide>
	 <head>Infoset WG</head>
	 <p>Basic problem: What's in a document after it's parsed</p>
	 <bulletlist>
		<item>
		  <p>Developing a data model for XML</p>
		</item>
		<item>
		  <p>Defining the objects that an XML application can expect to
			 find in a document</p>
		  <bulletlist>
			 <item>
				<p>Elements and attributes? <i>Of course.</i></p>
			 </item>
			 <item>
				<p>Comments? <i>Maybe.</i></p>
			 </item>
			 <item>
				<p>Individual characters? <i>Maybe not.</i></p>
			 </item>
		  </bulletlist>
		</item>
		<item>
		  <p>Other key groups are heavily dependent on the data model</p>
		  <bulletlist>
			 <item>
				<p>DOM</p>
			 </item>
			 <item>
				<p>XSL</p>
			 </item>
			 <item>
				<p>Linking</p>
			 </item>
			 <item>
				<p>Schemas</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>Fragment WG</head>
	 <p>Basic problem: How to work with part of a document in isolation 
	 </p>
	 <bulletlist>
		<item>
		  <p>Classic example: autonumbering</p>
		</item>
		<item>
		  <p>Document editing</p>
		</item>
		<item>
		  <p>Partial document transmission</p>
		</item>
		<item>
		  <p>W3C work builds on an existing solution developed over
			 several years by OASIS (SGML Open) for SGML documents</p>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>Profiling (Syntax WG)</head>
	 <bulletlist>
		<item>
		  <p>Originally, XML subsets (simpler versions)</p>
		</item>
		<item>
		  <p>Turning into an attempt to define classes of XML processors 
		  </p>
		</item>
		<item>
		  <p>Requirements are hard to agree on</p>
		</item>
		<item>
		  <p>Growing ubiquity of full XML 1.0 processors has taken a lot
			 of the urgency out of this work item</p>
		</item>
		<item>
		  <p>But the discussion has exposed weaknesses in the way the the
			 specification defines XML processors</p>
		</item>
		<item>
		  <p>My guess is that we will get cleaner definitions of the
			 categories implied in the 1.0 Recommendation</p>
		</item>
	 </bulletlist>
  </slide>

</slidemodule>

<slidemodule>
 
  <moduletitleslide>
	 <title>Continuing Work: Layout and Formatting</title>
  </moduletitleslide>
 
  <slide>
	 <head>Completing global publishing</head>
	 <p>Large-scale cross-platform XML publishing demands that XML
		deliver on the <i>display-oriented </i>promises:</p>
	 <bulletlist>
		<item>
		  <p>User-configurable views</p>
		</item>
		<item>
		  <p>More powerful display-centric client-side applications</p>
		</item>
		<item>
		  <p>Asian-language rendering support</p>
		</item>
		<item>
		  <p>Media-independent publishing</p>
		</item>
		<item>
		  <p>In particular, printed and online deliverables (now and in
			 the future) from the same source</p>
		</item>
	 </bulletlist>
	 <p>XSL provides the foundation for professional publishing in the
		21st century.</p>
	 <bulletlist>
		<item>
		  <p>A <emph>single technology </emph>for online and print
			 publishing</p>
		</item>
		<item>
		  <p>A <emph>common infrastructure </emph>(training, tools, and
			 techniques) for online and print publishing</p>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>XSL transformational requirements</head>
	 <bulletlist>
		<item>
		  <p>Generated structures (e.g., TOCs)</p>
		</item>
		<item>
		  <p>Reordering (e.g., sort ordering)</p>
		</item>
		<item>
		  <p>Logical objects to formatting objects (FOs)</p>
		  <bulletlist>
			 <item>
				<p>In: <b>Book</b>- Chapter - Section - <b>Para</b></p>
			 </item>
			 <item>
				<p>Out: <b>Book</b>- Page sequence - Column set -
				  <b>Para</b></p>
			 </item>
		  </bulletlist>
		  <p>The logical tree and the formatting tree look similar at
			 first glance, but in reality, they are profoundly different.</p>
		</item>
		<item>
		  <p>Logical objects to HTML (to use existing web browsers)</p>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>XSL formatting and layout requirements</head>
	 <bulletlist>
		<item>
		  <p>Character alignment (top, middle, bottom)</p>
		</item>
		<item>
		  <p>Writing direction (left, right, vertical)</p>
		</item>
		<item>
		  <p>Mixed-width column sets</p>
		</item>
		<item>
		  <p>Multiple flows</p>
		</item>
		<item>
		  <p>Footnote zones</p>
		</item>
		<item>
		  <p>Callouts</p>
		</item>
		<item>
		  <p>Autonumbering</p>
		</item>
		<item>
		  <p>Automatic copyfitting</p>
		</item>
		<item>
		  <p>Content-driven headers and footers</p>
		</item>
		<item>
		  <p>Formatting changes at non-tag boundaries</p>
		</item>
		<item>
		  <p>Hyphenation controls</p>
		</item>
		<item>
		  <p>etc., etc., etc.:
			 <hlink
			 href="http://www.w3.org/TR/WD-XSLReq">http://www.w3.org/TR/WD-XSLReq</hlink>
		  </p>
		</item>
	 </bulletlist>
  </slide>

  <slide>
	 <head>Misconceptions about XSL</head>
	 <numberedlist>
		<item>
		  <p><emph>Misconception: </emph>XSL is competition for CSS</p>
		  <bulletlist>
			 <item>
				<p><emph>Fact: </emph>XSL is a parallel formatting language
				  designed to solve a set of problems unique to the page-oriented display
				  of structured documents</p>
			 </item>
		  </bulletlist>
		</item>
		<item>
		  <p><emph>Misconception: </emph>XSL is something new</p>
		  <bulletlist>
			 <item>
				<p><emph>Fact: </emph>A formatting language separate from
				  CSS and specially designed for the needs of XML was part of the XML
				  charter from the beginning (1996)</p>
			 </item>
		  </bulletlist>
		</item>
		<item>
		  <p><emph>Misconception: </emph>XSL is just a transformation
			 language (XML to HTML)</p>
		  <bulletlist>
			 <item>
				<p><emph>Fact: </emph>The transformation component of XSL
				  exists to satisfy the formatting requirements; the formatting
				  requirements of XSL are primary</p>
			 </item>
			 <item>
				<p>Rendering through HTML generation is a stopgap to get us
				  past the transition to better XML formatting; it is not and never has
				  been the end goal of XSL</p>
			 </item>
		  </bulletlist>
		</item>
	 </numberedlist>
  </slide>

  <slide>
	 <head>XSL scheduling</head>
	 <bulletlist>
		<item>
		  <p>New working draft, April 1999 (check it out!)</p>
		</item>
		<item>
		  <p>Support for the transformational component is catching on 
		  </p>
		</item>
		<item>
		  <p>Support for the formatting objects is much slower in coming 
		  </p>
		  <bulletlist>
			 <item>
				<p>Formatting is <emph>hard</emph></p>
			 </item>
			 <item>
				<p>Little incentive for magazine-quality design in current
				  web browser displays</p>
			 </item>
			 <item>
				<p>No instant gratification for low-end users</p>
			 </item>
			 <item>
				<p>No immediate reason for vendors of high-end publishing
				  systems to standardize on a completely interoperable page design
				  language</p>
			 </item>
			 <item>
				<p>In sum: not many people get the point yet (like XML 3-4
				  years ago)</p>
			 </item>
		  </bulletlist>
		</item>
	 </bulletlist>
  </slide>

</slidemodule>

</slideshow>
