Languages of the Web: Background

1
Languages of the Web Background

• David Cornforth
• School of ITEE
• University of NSW,
• Australian Defence Force Academy

2
Summary

• History of typesetting and markup shows
• Technology becomes redundant - we need something
  that can grow, or be extended
• Documents contain data - how can we make it
  accessible?
• Data and format are different things
• XML could be the answer?

3
Why XML?

• Its extensible - create your own tags
• Incorporates meta-data (self-describing)
• It separates data from formatting information
  (data can be portable)

4
Separate data, formatting
Source data Customer Smith bought 56 units at
45.80 each Customer Patel bought 34 units at
39.50 each Customer Wicks bought 23 units at
47.50 each
5
How did we get here document processing, a brief
history

• Writing invented around 4000 BCE - pictographic
  (Egyptian, Chinese)
• Alphabet invented E Mediterranean and Mesopotamia
  around 2000-1500 BCE
• Printing with moveable type invented around 1000
  CE in China
• Made really practicable by Gutenberg, Germany,
  around 1450 CE

6
Markup languages brief history

• Pre-computer
• Markup conventions, for denoting the formatting
  of documents have existed since at least the
  start of the printing industry
• Early computer markup
• The first computer markup conventions were based
  on hand markup systems
• Commands could have parameters - ltpointsize,10gt -
  but users could not define macros
• Why do we need macros?

7
Markup languages brief history

• ROFF/NROFF/TROFF (1970s)
• A series of DEC-based lineprinter/ typesetter
  formatting systems
• Often still used in UNIX documentation
• .\" _at_()Golem1.oalpha.1 1.25 89/03/26 SMI
• .TH GOLEM1.0Alpha 1 "October 26, 1990"
• .SH NAME
• golem \- first order induction algorithm based on
  relative least general
• generalisation
• Supported limited macroing

8
Markup languages brief history

• COMTEXT / XICS (1970s)
• Early CSIRO / Xerox markup language
• ltbd/cm/pt,18gtGOLEM1.0Alphaltlt/ju/pt,12gt
• supported limited macroing, but without
  parameters
• lthdbd/cm/pt,18gt

9
Markup languages brief history

• SCRIBE
• The first concerted attempt at high-level rather
  than layout markup
• AImed at academic documents papers, theses
• High level description of document, combined with
  a separate style definition defining how the
  document is to be layed out

10
Markup languages brief history

• TeX / AMSTeX
• TeX highly flexible programming language built
  on top of a markup language
• Basic operations low level, but macro packages
  can be defined
• AMSTeX (for the American Mathematical Society)
  the first major macro package highly
  sophisticated, but relatively layout-dependent

11

• \hrule
• \vskip 1in
• \centerline\bf A SHORT STORY
• \vskip 6pt
• \centerline\sl by A. U. Thor !??!
  (modified)
• \vskip .5cm
• Once upon a time, in a distant galaxy called
  \"O\"o\c c, there lived a computer named R.J.
  Drofnats.
• Mr.Drofnats---or R. J.,'' as he preferred to
  be called--- error has been fixed!
• was happiest when he was at work typesetting
  beautiful documents.
• \vskip 1in
• \hrule
• \vfill\eject

12
Markup languages brief history

• LaTeX
• Combined many of the ideas of TeX and Scribe a
  generic macro package, permitting high-level
  document descriptions to be built on top of TeX
• Still widely used in scientific literature

13

• \documentstyle12ptarticle
• \chapterThe Formation of Hierarchical Structures
  in a Pseudo-Spatial Co-Evolutionary Artificial
  Life Environment
• \author David Cornforth
• \date8 March 2005
• \sectionIntroduction
• One of the key questions in the study of
  Artificial Life is to understand "open-ended
  complexity". That is, how do increasingly complex
  structures and behaviour arise in natural
  systems? In particular, is it possible to capture
  this phenomenon within a simulation model 1?

14
Markup languages brief history

• Xanadu
• Ted Nelson, 1960 - present
• A full-scale implementation of hypertext,
  including multi-way links, version control,
  licensing mechanisms
• One of the ancestors of html/www
• http//www.xanadu.net/

15
Markup languages brief history

• Dynabook
• Alan Kay, late 1960s - 1970s (Xerox PARC)
• First seriously GUI-based hypertext system
• Click on links and other metaphors now embedded
  in the www
• http//ftp.sheridanc.on.ca/randy/design.dir/softw
  are.dir/alan_kay.htm

16
Markup languages brief history

• Hypercard
• Bill Atkinson, 1987
• A graphic programming environment as much as a
  hypertext system
• Nevertheless the first widely-available GUI-based
  hypertext environment

17
Markup languages brief history

• Office Document Architecture (ODA)
• Initially an IBM internal effort. Initiated as
  part of the ISO OSI standards effort in the early
  1980s, and issued as ISO standard 8613 in 1987.
  It defines a language for describing the
  structure of office documents. ODIF specifies the
  exact format to be used for the interchange of
  documents between computer systems.
• Distinguishes between a document's LOGICAL
  structure and its LAYOUT structure. The logical
  structure associates the content of the document
  with a hierarchical tree of logical objects,
  whereas the layout structure associates the same
  content with a hierarchical tree of layout
  objects.
• (ISO International Organization for
  Standardization)

18
Markup languages brief history

• Generalised Markup Language (GML)
• Also initially an IBM internal effort, initiated
  by Charles Goldfarb. Concentrated on the
  definition of logical structure, with an emphasis
  on language extensibility
• Migrated out of IBM to become the ISO SGML
  standard

19
Markup languages brief history

• Standard Generalised Markup Language (SGML)
• Became an ISO standard (8879) in 1986
• Initially very slow to gain acceptance
• Nevertheless the root of almost all important
  markup developments since that date

20

• ltchaptergt
• ltchaphdgt SGML tags lt/chaphdgt
• ltchapabsgt In this chapter, the various properties
  of SGML tags are investigated in monotonous
  detail. lt/chapabsgt
• ltchaptxtgt ltparagt Tags are the basic markup
  elements of SGML, .....lt/paragt
• ltfig id"tagfig"gt
• ltfigbodygt
• ltartwork depth"24p"gt lt/artworkgt
• lt/figbodygt
• ltfigcapgt A Picture of Pure Monotony lt/figcapgt
• lt/figgt
• ltparagt Yet more monotonous detail about SGML tags
  lt/paragt

21
Markup languages brief history

• Hypertext Markup Language (HTML)
• Designed by Tim Berners-Lee At CERN in the late
  1980s
• Originally intended as a means of communication
  between physicists
• Intended to be an SGML instance language, though
  mistakes in the original design and a poor
  separation of logical and layout structures meant
  that it was originally non-compliant
• Led to a number of HTML-specific browsers (lynx,
  mosaic, Netscape, Internet Explorer etc - highly
  efficient, but unusable for other SGML)

22

• ltHTMLgtltHEADgt
• ltMETA HTTP-EQUIV"Content-Type"
  CONTENT"text/html charsetiso-8859-1"gt
• ltTITLEgtSGMLlt/TITLEgtlt/HEADgt
• ltBODYgt
• ltBgtltFONT FACE"Helvetica" SIZE7gtltP
  ALIGN"CENTER"gtFurther High Level Representation
  Standardslt/Pgt
• lt/FONTgtltFONT FACE"Helvetica" SIZE6gtltPgtDSSSLlt/Pgt
• lt/Bgtlt/FONTgtltFONT FACE"Helvetica" SIZE5gtltP
  ALIGN"JUSTIFY"gt(Document Style, Semantics and
  Specification Language)lt/Pgt

23
Markup languages brief history

• Document Style, Semantics and Specification
  Language (DSSSL)
• The other half of the SGML logic/layout
  separation
• A language for specifying the layout of documents
  based on their logical (SGML) structure also
  provides a mechanism for transforming from one
  SGML language to another
• Standardised in the late 1990s
• Syntax and semantics heavily based on Scheme
• But because of SGMLs flexibility, actually an
  SGML compliant language

24

• (element note
• (make sequence
• (make paragraph font-size 14pt
• font-weight bold
• (literal "Warning "))
• (make paragraph font-size 12pt
• font-weight medium
• (process-children))
• (make paragraph font-size 14pt
• font-weight bold
• (literal "You have been warned! "))))

25
Markup languages brief history

• Extensible Markup Language (XML)
• An SGML sub-language
• Restrictions intended to
• retain most of the flexibility of SGML
• Approximate the processing cost of HTML
• Not require a document type definition for
  tokenisation (though obviously full parsing
  requires a DTD or schema)

26

• lt?XML version"1.0"?gt
• lt!DOCTYPE play PUBLIC "-//Free Text Project//DTD
  Play//EN"gt
• ltPLAYgt
• ltTITLEgtThe Tragedy of Hamlet, Prince of
  Denmarklt/TITLEgt
• ltfmgtltpgtText placed in the public domain by Moby
  Lexical Tools, 1992.lt/pgt
• ltpgtXML version by Jon Bosak, 1996-1997.lt/pgtltfmgt
• ltPERSONAEgt
• ltTITLEgtDramatis Personaelt/TITLEgt
• ltPERSONAgtCLAUDIUS, king of Denmark. lt/PERSONAgt
• ltPERSONAgtHAMLET, son to the late, and nephew to
  the present king.lt/PERSONAgt

27
Markup languages brief history

• Extensible Style Language (XSL)
• The XML analogue of DSSSL
• However XMLs inflexibility means that the syntax
  has to be standard XML syntax (ie fairly crude
  for a programming language)
• Provides both layout and transformation
  facilities of XML

28

• ltxslstylesheetgt
• lt!--Here is the pattern part --gt
• ltxsltemplate match greetinggt
• lt!--Here is the action part --gt
• ltfoblock color red font-size16ptgt
• ltprocess-children/gt
• lt/foblockgt
• lt/xsltemplategt
• lt/xslstylesheetgt

29
Why XML?

• SGML provides all the flexibility needed for the
  web, but
• Processing costs are unacceptable
• SGML documents cannot be processed without their
  defining DTDs, which may be many times larger
  than the documents themselves, so communication
  costs may also be unacceptable

30
Why XML?

• HTML provides the speed and DTD-independence
  needed for the web, but
• Too generic - syntax cannot match semantics
• Thus of limited usefulness in communicating
  structured information

31
Why XML?

• XML provides a compromise between SGML and HTML
• A generic language like SGML
• Restricted syntax (eg no omitted end tags, single
  character set) permits unambiguous tokenisation
  even without a DTD
• Restricted syntax also permits HTML-like
  processing speeds

32
What will XML do for us?

• Search and display of structured data
• Ie the obvious extension of the uses of HTML
• With better (ie structured) searching
  capabilities
• Standardisation of data within organisations
• The structuring capabilities of XML mean that
  organisations can standardise the structure of
  their information through schemas
• Standardisation of data between organisations
• Entire industries can standardise on
  data-exchange formats, enabling new levels of
  e-commerce service

33
What will XML do for us?

• Encapsulation of metadata
• XML has permitted the definition of extended
  languages to encode structured metadata
• The semantic web becomes a real possibility
• Standardisation of heirarchical databases
• The standardisation of SQL led to the recent
  dominance of relational databases
• The standardisation permitted by XML is leading
  to a major resurgence of heirarchical data storage

34
What will XML do for us?

• Reduction of Middleware
• XML solutions provide a real alternative to much
  of what is currently done in middleware
• We can expect to see a massive reduction in the
  use of programming solutions for middleware
  requirements
• Enabling Diverse Enterprise Architectures
• XML standards can provide the glue enabling
  diverse specialised systems to cooperate
• We can expect to see such solutions providing
  serious competition to monolithic enterprise
  resource systems

35
What will XML do for us?

• Integration of Legacy Systems
• XML solutions can provide standards for the
  integration of legacy systems, provided their
  interfaces can be modified to interact with XML
  streams
• However SGML/DSSSL may be useful to provide full
  functionality
• Because almost any structured data stream can be
  treated as an SGML stream
• DSSSL may be used to translate it into an XML
  stream without any requirement to alter the
  system itself

36
What will XML do for us?

• Plus all the things we havent thought of
• Like SGML, XML is already being used for purposes
  way beyond the wildest imaginings of its creators
• E.g. Slideware (replace PowerPoint?)

37
Questions?