File Access Services, Directory Services,

1
File Access Services, Directory Services, The
World Wide Web (WWW)

• 635.413.31 Summer 2007

2
File Access Services

• Definition of File Access Services
• The ability to transparently access information
  stored in files on other systems across a network
• File Access vs. File Transfer
• File Transfer Characteristics
• File transfer makes independent copies of files
  on different hosts
• Copies the entire file and access is generally
  not available until the copy is completely
  executed
• File access methods privileges are not the same
  for local and remote resources (so its not
  transparent)
• Allows for storage of intermediate results,
  distributed management control, and data
  protection

3
File Access Services

• File Access Characteristics
• User access is only to a selected portion of a
  file the complete master copy of the file is
  stored on a single system
• Images of file are not independent there is
  only one real copy stored on a single system
• File access methods and permissions are typically
  the same for local and remote resources
• Typically provides quick up-to-date access to
  files and prevents different versions of a file
  from getting out of synchronization
• Multi-user access to files must be carefully
  managed!!

4
Examples of Major File Access Services

• The Networked File System (NFS)
• The goal of NFS is to provide transparent file
  access for clients to files and filesystems on
  NFS servers
• From both a user a technical perspective uses a
  client-server architecture
• Developed by Sun Microsystems to allow
  filesystems to be distributed across multiple
  UNIX/Linux systems
• Current specification (version 4) is in RFC 3530.
  still a number of version 3 implementations (RFC
  1813)
• NFS clients available for Windows and Mac
  Operating Systems
• Theoretically any application on the client that
  works with local file access should also work
  with NFS file access
• Technical details of NFS will be examined later

5
Examples of Major File Access Services

• Microsoft File Services
• File access and sharing has been around as a
  separate product for a decade (LANManager)
  support for networked file access now an integral
  part of Windows 95/98/NT and later versions
• Originally proprietary but Microsoft has
  transitioned to TCP/IP (actually UDP) and
  published the interface specifications in RFC
  1001 and 1002
• Major components of Microsoft file access
  services
• The Server Message Block API
• The NetBIOS API
• Browser service
• The Universal Naming Convention (UNC)

6
Examples of Major File Access Services

• Microsoft File Services (Continued)
• The client typically maps a networked file or
  filesystem by binding the resource (via the UNC)
  to a local drive letter at that point access is
  the same as a local resource
• Also Resources can be mapped directly with the
  UNC
• Clients available for Windows, OS\2, Mac, and
  many desktop UNIX/Linux systems (via SAMBA, etc.)
• Microsoft has introduced a new component called
  DFS (Distributed File Services)
• Provides a unified virtual namespace that makes
  distributed resources look like local resources
• Provides unified access across multiple real
  file services like SMB, Netware, and NFS
• Still proprietary to Microsoft

7
Examples of Major File Access Services

• Apple (Appleshare and the Appletalk Filing
  Protocol)
• Created as part of a proprietary network
  architecture for Macs
• Created for Apple File sharing client-server or
  peer-to-peer
• Current versions run on TCP/IP
• Novell Netware
• Created to support Novell File Servers Clients
• Proprietary client-server architecture though
  Novell does do 3rd party licensing
• Access to remote resources via drive mapping
  where a remote resource is bound to a local drive
  letter
• Current versions run on TCP/IP
• Clients for DOS, Win3.1, Win95/98, NT, OS/2, and
  Mac

8
The Remote Procedure Call (RPC)

• The foundation of NFS The Remote Procedure Call
  (RPC)
• NFS is based on the Remote Procedure Call, a
  generic API that allows procedures called on one
  system to execute on another
• If a client is RPC aware nothing needs to be
  done by a user to access resources via NFS
• There are two different RPC varieties Sun RPC
  and the OSF Distributed Computing Environment
  (DCE)
• Sun RPC specifications published in RFC 1057
  version 2 open specifications published in RFC
  1831 (version 2)
• Sun RPC, the foundation of most NFS
  implementations, uses the Sockets API to access
  TCP and UDP transport services
• The RPC specifications consist of two main
  pieces
• The transfer protocol (message format exchange)
• Data representation encoding

9
The Remote Procedure Call (RPC)

• General Diagram of RPC Operation

10
The Remote Procedure Call (RPC)

• RPC Message format
• Two messages used a Request and a Reply
• Fields in an RPC Request
• Transaction ID (XID) field 4 bytes a field
  initialized by the client w/ a unique number used
  to match Replies w/Requests
• Call field 4 bytes set to zero for a Request
  one for a Reply
• RPC version field 4 bytes specifies what
  version of RPC is in use typically either two
  (version 2) or three (for version 3)
• Program numbers, version, and procedure fields 4
  bytes each used by the server to determine
  what specific procedure the client wants to
  access remotely

11
The Remote Procedure Call (RPC)

• RPC Message Format (Continued)
• Credentials field up to 408 bytes
• Used to identify the client to the server
• Use is optional typically the User ID and Group
  ID of the client process is included
• Can be used for the server to enforce access
  restrictions on the RPC
• Actual length of field encoded at the beginning
  of the field value
• Verifier field up to 408 bytes used with
  secure RPC to pass encryption related information
• Parameters field variable length holds all
  parameter information necessary for the procedure
  call to be executed by the server

12
The Remote Procedure Call (RPC)

• RPC Data Representation
• RPC uses a special encoding system call External
  Data Representation (XDR) for encoding the data
  values in RPC fields
• Allows heterogeneous systems to communicate using
  a common data representation
• Field structure including bit and byte order
  specified for each field in the RPC messages
• As an example, in XDR all integers are encoded as
  four byte values
• XDR specifications published in RFC 4506

13
The Remote Procedure Call (RPC)

• The RPC Port Mapper (411 for RPC)
• RPC server programs use ephemeral ports instead
  of well-known ports, so some kind of registrar
  is necessary for client stub processes to
  determine how to communicate with server
  processes
• The port mapper is an RPC server process all
  other RPC server processes on that particular
  server registers with when they initialize
• The port mapper listens for client calls on
  well-known port 111 (both UDP and TCP)

14
The Technical Details of NFS

• General NFS Technical information
• NFS uses a subset of defined RPCs to provide
  distributed file access (RPC can do much more
  than that)
• Files and filesystems are referenced through the
  use of file handles objects or structures
  used to represent the files or filesystems
• NFS servers are stateless (version 3 and
  earlier) servers do not keep track of which
  clients are accessing its resources or which
  resources are being accessed
• Version 4 introduces file locking, which
  introduces stateful operation and requires more
  complex operational procedures

15
The Technical Details of NFS

• An Example of NFS Operation

16
The Technical Details of NFS

• NFS Procedures
• NFS Version 4 has 39 procedures for file
  directory manipulation and access as well as an
  extensible security framework
• Built on RPCSEC_GSS (RPC 2203)
• Provides integrity, privacy, and authentication
• Allows for negotiation of security options
• NFS Version 3 has 20 procedures for
  file/directory operations
• Important NFS procedures (common to v3 and v4)
• OPEN CLOSE compound commands for file
  manipulation
• ACCESS check file access permissions
• READ WRITE reads and writes to/from files
  options for either synchronous or asynchronous
  writes
• GETATTR SETATTR get and set attributes on a
  file
• MKDIR RMDIR make directory delete directory

17
The Technical Details of NFS

• Mounting Drives
• In order for a client to access remote files via
  NFS it must use the NFS mount procedure
• The mount procedure returns a handle which the
  client uses to reference the filesystem
• The client then uses the handle to integrate the
  remote filesystem into its local filesystem
  structure (analogous to mapping a networked
  drive)
• As with other NFS procedures the port mapper must
  be called first to find the UDP or TCP port being
  used on the NFS server
• Note this procedure works differently under
  the hood in v4, as no portmapper is used and the
  mount protocol has been incorporated into the
  core NFS specification

18
The Technical Details of NFS

• NFS over TCP
• Originally NFS was strictly a LAN technology
  rarely implemented over wide area connections, so
  UDP was used to provide efficient throughput
• With the globalization of companies NFS is being
  used more often over wide geographic areas
• Using TCP provides NFS with more robust
  implementations over WAN connections
• The default transport for NFS version 4 is TCP

19
The Technical Details of NFS

• An NFS Example Reading the file testplan.txt
  from mountpoint/smiley/testplan.exe

20
Directory Services
21
Directory Services

• Definition
• A means of searching a database of information
  easily rapidly using keywords to match
  attributes stored in the database
• With the explosion of networking and the Internet
  very important for finding people, information,
  services
• Directory services must be
• Easy to use
• Contain correct data
• Provide value
• Directory services are categorized as White or
  Yellow Pages
• White Pages Directory Services oriented toward
  information associated with individuals
• Yellow Pages Directory Services oriented toward
  finding resources (networked printers, servers,
  etc) or services

22
Directory Service Protocols

• There are 4 Major Directory Service Protocols
  used in the Internet
• DNS,
• NIS,
• X.500, and
• LDAP
• And yes, NDS, WINS, and AD could be included
  too
• The Domain Name Service
• Yes, this can be considered a very limited
  Directory Service!
• Can be adapted for other uses such as spam
  blacklists
• Provides only name resolution services but some
  information can be gleaned from the name
  structure
• Has almost global accessibility
• Already covered earlier in exhaustive detail

23
Directory Service Protocols

• The Network Information Service (NIS)
• NIS was developed by Sun Microsystems as a
  companion to NFS to ease the burden on
  administrators of distributed computing systems
• Uses a client-server architecture for
  distribution of Yellow Pages information
• NIS allows the use of a single administrative
  repository for important network-wide
  information such as password files.
• Enables the use of a single username password
  across multiple systems
• Also includes information search and retrieval
  commands that can be used in the development of
  non-administrative distributed directories and
  databases
• Sun is trying to migrate from NIS to the
  LDAP-enabled Sun Directory Services (SDS)
  theres a large installed base of users so this
  will likely take a while

24
Directory Service Protocols

• Technical Details of NIS
• Like NFS, NIS is built upon the Remote Procedure
  Call (RPC) and NIS procedures operate very much
  like NFS procedures
• Built around map files, which are the data or
  directory information repositories and maps,
  which are unique views or indexes for the data
  files
• Map files are stored in individual directories on
  the server (example the password file is
  typically stored on a server at /var/yp/passwd
• Servers come in two varieties master and slave
• The master server is the single administration
  point for the map files analogous to a primary
  DNS server
• There is only one master NIS server for a set of
  map files
• Slaves read all their map file information from
  the Master
• Clients can be serviced equally well by either a
  Master or a Slave

25
Directory Service Protocols

• Technical Details of NIS (continued)
• Organizations that want multiple access policies
  for directory or distributed database information
  can segregate NIS servers into Domains
• A domain defined as a set of servers that access
  and distribute information from a unique set of
  maps
• Each domain has a single Master server and as
  many slaves as necessary to adequately service
  client requests
• Domains can overlap (an NIS server could be a
  Master for multiple domains) and clients can
  easily switch between domains
• NIS Database Operations come in two flavors
  distributed filesystems and explicit commands
• With the distributed filesystem mode of operation
  NIS clients read data for administrative
  functions from the NIS server instead of a local
  file
• This mode is used for such functions as password
  or license files it allows a single file
  administered from the Master to be used across
  multiple NIS client systems

26
Directory Service Protocols

• Technical Details of NIS (continued)
• NIS clients can also be set up so they will check
  their local file first and if a match is not
  found do an NIS lookup
• NIS services (maps) may also be accessed using
  explicit commands
• This is very useful for building distributed
  database or directory applications that have
  nothing to do with system administration (i.e.
  systemwide personnel directory, etc.)
• The most important NIS user command is ypmatch
  this allows the search of a map using a keyword
• Three groups of internal NIS procedure calls
  perform the work for either mode of operation
• Client-Lookups key-driven procedures (match,
  get-first, get-next, get-all)
• Maintenance-calls checking server names status
  (get-master, get-order)
• Internal NIS calls commands between servers
  such as a map transfer from Master to Slave

27
Directory Service Protocols

• The X.500 Directory Service (aka DAP)
• ITU standard for storing, accessing
  distributing directory information
• What X.500 provides
• A standards-based directory
• A structured information framework
• A single global namespace
• Powerful search capabilities
• Decentralized maintenance
• Encompasses seven recommendations (X.501, X.509,
  X.511, X.518, X.519, X.520, and X.521) besides
  X.500 the standards cover
• Directory Service Quality access rules,
  authentication, filters, etc.
• Directory Queries read, compare, list, search
  operations, etc.
• Directory Modification add, rename, modify
  operations, etc.
• Error Reporting definition of of error
  conditions and responses
• Referrals relationships between Directory
  servers other external objects
• Originally designed for OSI protocols but later
  modified for TCP/IP

28
Directory Service Protocols

• X.500 Technology
• X.500 Directory Structure and Terminology
• Information is held in a Directory Information
  Base (DIB)
• A DIB consists of individual entries organized in
  a tree structure called the Directory Information
  Tree (DIT)
• Can also have relational characteristics
• Each entry is composed of attributes and has a
  Distinguished Name (DN) that uniquely identifies
  it (User friendly Distinguished Names are very
  important to X.500)
• Each attribute composed of a type and one or more
  associated values
• The type specifies a particular syntax and data
  type for the value (boolean, integer, etc.)
• DIT entries have some object-oriented
  characteristics like inheritance
• The DIT also has a schema in X.500 the schema is
  a rule-set that ensures the DIT maintains its
  logical structure during modifications

29
Directory Service Protocols

• An X.500 DIT example

30
Directory Service Protocols

• X.500 Technology
• The X.500 communication protocols
• X.500 client (Directory User Agent or DUA)
  communicates with an X.500 server (Directory
  Server Agent or DSA) using the Directory Access
  Protocol (DAP)
• The X.500 DAP uses the standard OSI presentation
  layer Remote Operations Service Element (ROSE)
  and Association Control Service Element (ASCE)
  for communications
• RFC 1249 defines a standard for interface the
  X.500 service to the UDP and TCP transport layers
  for use over the Internet
• OSI networking is ridiculously complex and X.500
  over TCP/IP isnt much better so thats all were
  going to cover on this subject

31
Directory Service Protocols

• The Lightweight Directory Access Protocol (LDAP)
• Even with the adaptation of X.500 to TCP/IP very
  few organizations actually adopted it
• Required too much computing and OSI engineering
  expertise
• Most implementations were kludgey and complex
• Engineers researchers at the University of
  Michigan decided a completely new protocol for
  accessing X.500 services from Internet clients
  would hasten deployment of global directory
  services
• The Lightweight Directory Access Protocol (LDAP)
  was originally designed to be a gateway
  service an X.500 back-end was still necessary
• Eventually the designers saw the value of LDAP as
  a completely independent directory services
  access protocol
• LDAP version 3 specifications published as RFC
  2252
• Recent update in RFC 4510 provide new features
  protocol extensibility

32
Directory Service Protocols

• The Lightweight Directory Access Protocol (LDAP)
• What the current version of LDAP provides
• Centralized administrative tasks within an
  organization
• Storage of sensitive information in a centrally
  secure repository
• Authentication and security services
• Centralized procedure for determining the status
  and role of an individual in an organization
• Standardizes the search and retrieval of white
  yellow page data
• Multi-platform, multi-vendor open standard with
  published APIs for software development
• URL format specifications for accessing LDAP data
  via browsers
• Gateway services to other directory services like
  NIS and NDS
• Vendors with current LDAP enabled products
  Redhat, Innosoft, OpenLDAP, Sun, Microsoft, and
  Novell (list not exhaustive)

33
Directory Service Protocols

• The Lightweight Directory Access Protocol (LDAP)
• LDAP Technology
• LDAP standard addresses 3 areas API, data
  format, and access protocol
• There are several APIs published and in use
• The base API developed by the U. of Michigan
  published as RFC 1823
• Other C and Perl based APIs readily available
• LDAP data format
• LDAP has the same object format and definitions
  as X.500
• Directory Information Base (DIB) Directory
  Information Tree (DIT)
• Entries and Attributes
• LDAP DIB structure includes both hierarchal
  relational structure
• Four LDAP object types are currently defined
  Person, Organizational Unit, Group, and Domain

34
Directory Service Protocols

• Example LDAP
• directory structure

35
Directory Service Protocols

• The Lightweight Directory Access Protocol (LDAP)
• LDAP access protocol
• Conforms to a client-server architecture (clients
  performing operations against servers)
• Does not define how data is stored in server
  just how to access it in a standard manner
• Uses TCP client can issue multiple queries over
  a TCP connection
• Basic LDAP Operations
• Binding an LDAP client to a server
• Required operation before a client can search
  information on a LDAP server
• Authentication and access controls are parts of
  this operation
• Unbinding and rebinding can occur over a single
  TCP connection allowing a client to change access
  privileges

36
Directory Service Protocols

• The Lightweight Directory Access Protocol (LDAP)
• Basic LDAP Operations
• Server search
• Single or multiple keyword
• Wildcards are allowed
• Compare entries
• Add entry or entries
• Modify existing entry or entries
• Delete existing entry or entries

37
Directory Service Protocols

• The Lightweight Directory Access Protocol (LDAP)
• Other LDAP Operations
• Referral allows one server to pass searches to
  another server
• Replication allows multiple servers to handle
  requests for a DIT enhances availability via
  load balancing and redundancy
• Encryption Security supports Kerberos, SASL,
  and Secure Sockets Layer (SSL)
• DIT discovery allows a client to query a server
  for its structure and entity relationships

38
Introduction to the World Wide Web
39
Introduction to the World Wide Web

• Goals
• Developed to provide an easy way to distribute
  info to members of a geographically dispersed
  group
• Especially well-suited for multimedia
  information
• Meant to allow software, hardware, system
  independent information sharing!
• Built on the concept of hypertext and hypermedia
  cross-linked groups of documents or objects
• Plenty of good Web related information available
  at http//www.w3c.org

40
Introduction to the World Wide Web

• History
• The WWW concept was developed in 1989 by Tim
  Berners-Lee at CERN
• He wanted a way to easy distribute high energy
  physics data to researchers located around the
  world
• Concept was developed and culminated in the
  release of the first WWW browser in 1993 (Mosaic)
• With the incorporation of Netscape commercial
  development of the WWW gained momentum
• WWW is actually based on an earlier text-based
  system called gopher that was developed in the
  mid-80s

41
Introduction to the World Wide Web

• General Operation Concepts
• WWW has client-server architecture, with browsers
  (clients) pulling data from Web servers
• Browsers request information (web pages) using a
  unique 'address' for each page called a Uniform
  Resource Locator (URL)
• Servers deliver either an error indication or the
  requested web page information to browser
• The WWW uses TCP for transport (i.e. one-to-one
  communications)
• WWW is mostly a 'pull' technology only
  client-requested info is sent
• Web pages are linked together in a hierarchical
  'mesh' by embedding URLs in web pages
  hyperlinks
• Caching is an integral part of the WWW can take
  place in many locations
• Server-side in web farms (load balancers
  caching appliances)
• In the network at firewalls and proxy servers
• Client-side in web browsers

42
World Wide Web Addressing

• There are three basic parts to the WWW
  Addressing, Transport, Presentation
• Addressing (URLs)
• One of most important aspects of the WWW is its
  global addressing scheme called the Uniform
  Resource Locator (URL)
• URLs provide a unique address allowing direct
  access to web pages
• URLs consist of three parts protocol, host
  name, and document name
• The URL Protocol part
• Web browsers can handle multiple protocols to
  provide backwards compatibility flexibility
• Depending on the browser these additional
  protocols can be handled internally in the
  browser or through the use of 'helper'
  applications
• Additional protocols handled telnet, ftp,
  gopher, news (NNTP), mail (SMTP), and local file
  access

43
World Wide Web Addressing

• The URL Host Name part
• Second part is the DNS name or IP address of the
  server where the web page resides
• Use of IP addresses is valid and works but is
  discouraged
• If no TCP port number is specified with the name
  or address, port 80 is implied
• If you wish to run web services on another port
  the port is specified after the DNS name
  (example http//www.noname.org7777/admin/system
  .html)
• The URL document name
• Final part of URL specifies the location of the
  web page on the server
• Usually consists of directory path followed by
  web page filename
• Path is typically relative to a default directory
  on the web server (operating system dependent)
• Path structure and filenames are operating system
  specific

44
Transport The Hypertext Transfer Protocol (HTTP)

• Introduction
• The Hypertext Transfer Protocol is the
  application layer protocol responsible for
  delivery of WWW data between browsers servers
• Current HTTP release is version 1.1 (RFC 2616)
  though a significant number of systems run the
  first standard 'production' release (v1.0)
• HTTP is a very simple text based request-response
  protocol like SMTP MIME like header fields and
  encoding conventions are used
• RFC-822/2822 format headers and MIME extensions
  are used to define the web page content,
  negotiate page parameters, support conditional
  requests
• HTTP uses TCP for reliable transport layer
  services
• HTTP v1.0 works in a stop and wait fashion - only
  one HTTP command can be outstanding requires a
  separate TCP connection for each object
• For each object (file) on a page a separate TCP
  connection is established
• HTTP v1.1 allows persistent connections and
  pipelining

45
Transport The Hypertext Transfer Protocol (HTTP)

• Important HTTP Characteristics
• Application Level
• Request/Response
• Stateless
• Bi-directional Transfer
• Capability Negotiation
• Support for Caching
• Support for Intermediaries/Proxies

46
Transport The Hypertext Transfer Protocol (HTTP)

• Important HTTP commands
• GET transfers a file from server to browser
• HEAD transfers the HTML header information for
  a web page from server to browser
• PUT uploads a file from browser to server (used
  a lot with web-based mailers)
• POST append information to a URL (used with
  forms)
• LINK create a link between files on a web
  server
• UNLINK delete a link between files on a web
  server
• DELETE delete a file on the server
• ALL of these commands are subject to access
  restrictions enforced by both the web server
  software and the server operating system!

47
Transport The Hypertext Transfer Protocol (HTTP)

• HTTP in action

48
Transport The Hypertext Transfer Protocol (HTTP)

• HTTP v1.1 Technical Details
• Persistent TCP Connections
• Allow multiple HTTP commands (and responses) to
  be sent over a single TCP connection can greatly
  increase response times and interactivity
• Reduces load opening closing TCP connections
• Fall back to v1.0 if necessary
• Application-layer compression
• Does not resend HTTP headers on sequential
  requests to the same server
• HTTP Command Pipelining
• Multiple HTTP commands can be issued without a
  response
• The server must respond to each request in the
  order it was received
• Expanded and improved directives for more
  intelligent Web caching
• Enhanced support for Web content lifetimes and
  expiration dates
• Validation algorithms commands to allow a cache
  to check on web page expiration
• Better support for transfer of dynamically
  generated web pages
• Chunked transfers files sent as a sequential
  stream of chunks with a defined size
• HTTP 1.1 has explicit header fields for defining
  chunk mode transfers and expressing the size of
  the chunks

49
Presentation The Hypertext Markup Language
(HTML)

• Introduction
• HTML is concerned with the structure and
  formatting of web pages
• HTML was developed from an ISO standard called
  SGML a metalanguage developed in 1974 (you
  may ask why SGML isnt used for the WWW instead
  of writing a new language its because SGML is
  very general and terribly complex)
• Current version is 4.01 (December 1999)
• HTML is designed to be backwards compatible to
  previous versions
• It was the original objective of HTML to allow
  the browser to control formatting presentation
  of page
• This was instituted to allow better data sharing
  in a heterogeneous environment
• Content providers want more control over the
  presentation of their information so the original
  object of browser control over presentation is
  being gradually eroded in each new release of HTML

50
Presentation The Hypertext Markup Language
(HTML)

• HTML page structure
• An HTML page consists of ASCII text and is
  divided into two main parts a header and a body
• The header contains information about the
  document
• The body contains the actual information to
  display
• HTML depends on embedded commands in the web page
  called TAGS
• Tags define both a command action and a scope
• Tags are enclosed in start and end symbols (lt
  and gt) to distinguish them from normal text
  that is for client display
• Tags, depending on their specific actions, can
  have optional attributes (size, align, action,
  etc.)

51
Presentation The Hypertext Markup Language
(HTML)

• Basic tags
• Page tags ltHTMLgt and lt/HTMLgt tags delimit the
  complete page
• Header and Body tags ltHEADERgt, lt/HEADERgt,
  ltBODYgt, and lt/BODYgt
• Hyperlinks ltA HREFURLgt and lt/Agt
• Images ltIMG srcURLgt and lt/IMGgt
• JPEG format (.jpg)
• GIF format (.gif)
• Formatting
• Physical formatting
• Specifies an exact display parameter for
  rendering web text
• Examples are Bold (ltBgt and lt/Bgt) and Font Size
• Logical formatting
• Specifies special treatment of the enclosed text
  but the exact physical rendering is left up to
  the browser or defined in a style sheet

52
Presentation The Hypertext Markup Language
(HTML)

• Style Sheets
• Provide a way to uniformly define how logical
  styles are assigned physical display attributes
  (without changing the configuration of your
  browser)
• Allows the web page author to apply typographic
  styles and spacing instructions for elements on a
  web page
• The advantages of style sheets are that they
  remove presentation information from the HTML
  document (which should be concerned only with
  structure) and one style sheet could be reused
  by linking it to multiple HTML documents
• The disadvantage of style sheets is that they
  take presentation control away from the browser
  -- if the content creator is not careful their
  web pages may not display properly
• Style sheets can be embedded into a web page or
  linked externally through a link or import
  statement in the header
• Generic Style Sheet entry syntax (example later)
  selector propertyvalue

53
Presentation The Hypertext Markup Language
(HTML)

• More Tags
• Menus, lists, and tables
• Menus
• Creates a compact list of choices with no
  numbering or bullets
• Relevant tags ltMENUgt, lt/MENUgt, and ltLIgt
• Lists
• Ordered Lists
• Ordered lists sequentially list items
• Relevant tags ltOLgt, lt/OLgt, and ltLIgt
• Unordered Lists
• Items are listed with bullets
• Relevant tags ltULgt, lt/ULgt, and ltLIgt

54
Presentation The Hypertext Markup Language
(HTML)

• More Tags
• Tables
• Creates a 2-D collection of data organized in
  rows and columns
• The ltTABLEgt and lt/TABLEgt tags bound the table
  data
• The rows are specified by ltTRgt tags
• Captions can be added using the ltCAPTIONgt tag
• Cell data delimited by the ltTDgt or ltTHgt tags
• Headings
• HTML Headings provide a set of logical styles
  for use in documents
• These are logical styles actual display
  properties set by the browser or a style sheet!
• Relevant tags ltH1gt through ltH6gt with end tags
  lt/H1gt through lt/H6gt
• Paragraphs, Line Breaks, Horizontal Rules
• These are the most commonly used tags that do not
  come in pairs
• Relevant Tags Paragraph ltPgt, Line Break ltBRgt,
  Horizontal Rule ltHRgt

55
Presentation The Hypertext Markup Language
(HTML)

• Example HTML Page
• lt!doctype html public "-//w3c//dtd html 4.0
  transitional//en"gt
• lthtmlgtltheadgt
• ltmeta http-equiv"Content-Type"
  content"text/html charsetiso-8859-1"gt
• ltmeta name"Author" content"John Romano"gt
• ltmeta name"GENERATOR" content"Mozilla/4.5
  en (Win95 U) Netscape"gt
• lttitlegtSample Web Pagelt/titlegt
• lt/headgtltbodygt
• lth4gtThis is a sample Web Page for my Internet
  Classlt/h4gt
• lth5gtI'll start off with a list of what toys I
  want for Christmaslt/h5gt
• ltolgt
• ltligtA trip to Tahitilt/ligt
• ltligtA fast expensive convertiblelt/ligt
• lt/olgt
• lthr WIDTH"100"gtltimg SRC"bix2b.JPG" height86
  width144gt
• lth5gtHere's a table for comparisonlt/h5gtnbsp
• ltcentergtlttable BORDER COLS3 WIDTH"100" gtlttrgt

56
Presentation The Hypertext Markup Language
(HTML)

• Advanced HTML
• Forms
• Basic HTML is one way information transfer
  between server and browser interactivity is very
  important and was added in version 2.0
• Forms allow data to be pushed back to the server
  for processing (via some other specification such
  as CGI)
• Forms delimited using the ltFORMgt and lt/FORMgt tags
• The Method and Action attributes of the FORM tag
  used to specify how and where to upload the form
  data
• Example ltFORM ACTION http//www.xyz.com/cgi-bin
  /order METHODPOSTgt
• Individual fields in the form specified using the
  ltINPUTgt tag
• The ltINPUTgt tag has many attributes to control
  labeling and data input
• The Name attribute
• The Size attribute
• The Type attribute

57
Presentation The Hypertext Markup Language
(HTML)

• Cascading Style Sheets
• A new feature that imbues style sheets in object
  oriented functionality
• Style sheets can be linked together in a tree
  structure
• Style sheets can inherent properties from parent
  to child
• Cascading style sheets sometimes have different
  behaviors between browsers!

58
A Look Ahead Addressing

• Addressing (URIs)
• A Uniform Resource Identifier is a formatted
  string which identifies, via name, location, or
  any other unique property, a resource
• URLs are a subset of URIs that specify a location
  this restriction can be a problem
• URIs will in some sense allow anycasting of
  information across the WWW
• Will allow requests for information without
  worrying about where it came from
• Will help eliminate problems with network and
  server outages
• URI composition
• A URI is composed of four major parts scheme,
  authority, path, and query
• The URI syntax is ltschemegt//ltauthoritygtltpathgt?lt
  querygt

59
A Look Ahead Addressing

• URI composition (Continued)
• Scheme
• Only the scheme part of the URI is mandatory
  other parts are optional
• Could identify a protocol or some other defined
  or standardized namespace
• The remainder of the URI syntax depends on the
  scheme!
• Authority
• May be defined as a Internet-based service or by
  a Scheme-specific naming authority
• Internet based-servers usually have the form
  ltuserinfogt_at_lthostgtltportgt
• Userinfo is optional but must be composed of
  legal URI characters
• Host would typically be a valid DNS name or IPv4
  address
• Port would be a valid TCP port number
• A scheme specific naming authority could be
  anything valid in the syntax of the scheme as
  long as it is composed of legal URI characters

60
A Look Ahead Addressing

• URI composition (Continued)
• Path
• The path component contains data, specific to the
  scheme and authority, that identifies the
  resource within the scope of the scheme and
  authority
• Typical paths in an Internet server-based URI
  (like a URL) navigate the directory structure of
  the server
• Query
• The query component is a string of information to
  be interpreted by the resource
• Typically used to provide keywords or parameters
  for formatting output

61
A Look Ahead - Presentation

• The Extensible Markup Language (XML)
• A new way to structure and format data on the
  Internet
• Goal make information transmitted across the
  Internet self-describing
• Consists of a ruleset to describe how to define
  your own markup language
• Like HTML and other markup languages tags are
  used -- however they are there more to describe
  what the information is instead of how it should
  look
• XML History
• Development of XML began in 1996 and culminated
  in early 1998 with the adoption of version 1.0
  standard by the World Wide Web Consortium (W3C)
• Like HTML, XML is based on the Standardized
  General Markup Language (SGML) developed by the
  ISO in the mid-1970s
• SGML is far too general and complex (the standard
  runs over 500 pages) for easy and widespread
  deployment XML is a stripped down version
  suitable for use on a wide range of networked
  systems
• HTML has been re-written into XML (aka XHTML
  1.1)

62
A Look Ahead - XML

• HTML shortcomings
• Even with the vast bandwidth in the Internet the
  WWW is often plagued by lack of interactivity and
  poor response times
• Explosive growth of the WWW has made it difficult
  to find what youre looking for (metatags web
  crawlers dont cut it anymore)
• HTML is inflexible it takes months to years for
  new standards with additional tags and
  functionality to be published
• XML solutions to HTML problems
• Speed with tags specifying what the delivered
  content means and what should be done to it a
  significant amount of processing can be done on
  the client
• Saves network bandwidth
• Offloads overloaded servers
• Searching XML makes searches more intelligent
  by allowing the structure and meaning of data to
  be included as keywords or search parameters
• Flexibility XML allows a content creator to
  define tags in a standard way so new tags can be
  created on the fly without rewriting standards

63
A Look Ahead - XML

• XML Implementation
• The XML standard is, by comparison to SGML or
  HTML, a very concise document
• XML uses Unicode a standard set of characters
  that encompasses all of the worlds major
  languages
• Like HTML, XML is built on tags
• Tags must be in start/end pairs that enclose the
  text to which they apply
• Tags cannot overlap, but they can be nested
• Nested tags create a tree structure for the XML
  document
• Information typically required in a new XML
  document
• What tags are allowed
• How they can and cannot be nested
• How tags should be processed
• The first two of the preceding three items are
  usually covered in the Document Type Definition
  (DTD)

64
A Look Ahead - XML

• A Real World XML Document (A RealServer license
  key)
• ltLicenseKeyDefinitiongt
• lt!-- Warning Do not Edit this file! Editing
  will invalidate your Real Server License Key. --gt
• ltList Name"License"gt
• ltList Name"Definition"gt
• ltVar Name"Evaluation" Value"True"/gt
• ltVar Name"Manufacturer"
  Value"RealNetworks"/gt
• ltVar Name"LicenseID"
  Value"01-0102-0031-57074"/gt
• ltVar Name"ProductID" Value"0"/gt
• ltVar Name"MajorVersion" Value"6"/gt
• ltVar Name"MinorVersion" Value"0"/gt
• ltVar Name"StartDate"
  Value"01/01/1997"/gt
• ltVar Name"EndDate"
  Value"01/01/2030"/gt
• lt/Listgt
• lt/Listgt
• ltList Name"General"gt
• ltVar Name"ClientConnections"
  Value"10"/gt
• ltVar Name"Live" Value"True"/gt
• lt/Listgt

65
A Look Ahead - XSL

• Other important parts of XML
• The Extensible Stylesheet Language (XSL)
• Defines a set of rules for presentation of XML
  data allowing the development of a write once,
  publish everywhere information infrastructure
• XSL stylesheets can be defined for various
  display media when an XML document is downloaded
  the appropriate XSL stylesheet is used to
  reformat the data to display data on the local
  system
• XSL stylesheets can be defined for presentation
  of data in non-visual formats using XSL an XML
  web page could be listened to or converted to
  braille
• XSL/XML document use

XML Document (Structure)
XSL StyleSheet (Presention)
Users display at Browser
XML Document 2 (Structure)
XSL StyleSheet 2 (Presention)
66
A Look Ahead XML XSL

• XML XSL Document Example
• lt?xml version 1.0 ?gt
• lt?xml-stylesheet typetext/xsl
  hrefeditors.xsl ?gt
• lt!-- This is a partial DTD - I havent defined
  everything necessary for this example! --gt
• lt!ELEMENT street (PCDATA)gt
• lt!ELEMENT city (PCDATA)gt
• lt!ELEMENT state (PCDATA)gt
• lt!ELEMENT zip (PCDATA)gt
• lt!ELEMENT address (street, city, state, zip)gt
• lteditor_contactsgt
• lteditorgt
• ltfirst_namegtJonathanlt/first_namegt
• ltlast_namegtSmithlt/last_namegt
• lttitlegtSenior Engineerlt/titlegt
• ltorganizationgtJohns Hopkinslt/organizationgt
• ltaddressgt
• ltstreetgt139 N Charles Streetlt/streetgt

67
A Look Ahead XML XSL

• XML XSL Document Example
• lt?xml version 1.0 ?gt
• lt!-- This stylesheet takes a simple XML doc and
  displays it in HTML --gt
• ltxslstylesheet xmlnsxslhttp//www.w3.org/TR/WD
  -xslgt
• lt!- this defines the xsl namespace this
  stylesheet belongs to -!gt
• ltxsltemplate match/gt
• lt!- Apply template to everything in the XML
  document -!gt
• ltHTMLgt
• ltBODYgt
• ltH1gtEditor Contactslt/H1gt
• ltxslfor-each selecteditor_contacts/editorgt
• ltH2gtName ltxslvalue-of selectfirst_name/gt
  ltxslvalue-of selectlast_name/gt lt/H2gt\
• ltPgt Title ltxslvalue-of selecttitle/gtlt/Pgt
• ltPgt Title ltxslvalue-of selectorganization/gtlt
  /Pgt
• ltPgt Title ltxslvalue-of selectaddress/street/
  gtlt/Pgt
• ltPgt Title ltxslvalue-of selectaddress/city/gtlt
  /Pgt
• ltPgt Title ltxslvalue-of selectaddress/state/gt
  lt/Pgt
• ltPgt Title ltxslvalue-of selectaddress/zip/gtlt/
  Pgt
• ltPgt Title ltxslvalue-of selecte_mail/gtlt/Pgt

68
A Look Ahead - XLink

• Improved Hyperlinks (XLink)
• Another W3C standard to provide more intelligent
  hyperlinks
• Version 1.0 approved June 2001
• Unlike current hyperlinks that can only link to
  another single physical location XLinks will be
  able to do much more
• XLinks could allow a choice of multiple actions
  or destinations
• Other XLinks will allow the information to be
  embedded directly in the page (maybe a floating
  dialog box) instead of forcing the viewer to
  leave that page
• XLinks could be indirect links, allowing changes
  to links to be made at a single database record
  instead of where ever the link is referenced

69
Reading Homework

• Reading
• Chapter 25 25.3, 25.13, and 25.14
• Chapter 27 HTTP
• Theres a lot more to the WWW XML than
  presented here see http//www.w3c.org for more!
• Next week advanced web applications