ghopp
/
server_playground
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
Another abandoned server code base... this is kind of an ancestor of taskrambler.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
38 Commits
1 Branch
0 Tags
51 MiB
 
 
 
 
 
 
Tree: de7e390c78
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from 'de7e390c78'
${ noResults }
server_playground/doc/www.w3.org/2005/Talks/1026park/index.html

528 lines
17 KiB
Raw Blame History

<?xml version="1.0" encoding="UTF-8"?><!--*- nxml -*-->
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd">
<html xmlns="http://www.w3.org/1999/xhtml">
<head profile="http://www.w3.org/2003/g/data-view">
<title>Web Architecture</title>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8" />
<link rel="transformation" href="http://www.w3.org/2002/12/cal/glean-hcal.xsl" />
<link rel="stylesheet" href="http://www.w3.org/2005/03/slideshow.css"
type="text/css" media="screen, projection" />
<script src="http://www.w3.org/2005/03/slideshow.js" type="text/javascript">
</script>
</head>
<body>
<div class="slide"><h1>Web Architecture</h1>
<address class="vcard">
<a class="url fn n" href="http://www.w3.org/People/Connolly/">
<span class="given-name">Dan</span>
<span class="family-name">Connolly</span></a><br />
</address>
<div class="vevent">
<a class="url summary" href="http://captain.park.edu/ics/index.shtml">Park University Department of Information and Computer Science</a><br />
<abbr class="dtstart" title="2005-10-26">26 Oct 2005</abbr>
</div>
<p class="acks">with thanks to Ian Jacobs for <a href="http://www.w3.org/2004/Talks/0121-ij-xmlgov/all.htm">his xml.gov presentation</a></p>
<p><em>postscript: The Park ACM club is hosting a <a
href="mms://flag.park.edu/acm/10_27.wmv">archived video stream</a> of
the presentation.</em></p>
<div><small>$Revision: 1.4 $
of $Date: 2005/10/28 17:30:21 $</small></div>
</div>
<div class="slide"><h1>Overview</h1>
<ol>
<li>W3C's Technical Architecture Group (TAG)</li>
<li>Overview of Architecture of the World Wide Web</li>
<li>Architecture Review of US Patent Office Web Site</li>
</ol>
</div>
<div class="slide"><h1>W3C's Technical Architecture Group</h1>
<p>W3C's <a href="http://www.w3.org/2001/tag/">Technical Architecture
Group</a> was <a
href="http://www.w3.org/2001/07/19-tag">chartered</a> in 2001:</p>
<blockquote>
<p><em>"to document and build consensus around principles of Web
architecture and to interpret and clarify these principles when
necessary."</em></p>
</blockquote>
<p>Roles: write, coordinate, mediate</p>
</div>
<div class="slide"><h1>TAG Participants</h1>
<p>TAG participants elected and appointed at our September 2002
meeting in Vancouver:</p>
<div class="photo"><img align="right"
alt="TAG, posing in front of same motorbike as in Sep 2002, in Vancouver"
src="http://www.w3.org/2003/07/tag-20030723.png" />
</div>
<ol>
<li><a href="http://norman.walsh.name/">Norman Walsh</a>, Sun Microsystems. Docbook guy</li>
<li>Paul Cotton, Microsoft. fulltext industry</li>
<li>Chris Lilley, W3C. SVG lead</li>
<li><a href="http://www.pacificspirit.com/">David Orchard</a>, BEA. Web Services. </li>
<li>Roy Fielding. HTTP spec editor. REST thesis</li>
<li>Tim Berners-Lee, W3C Director</li>
<li>Stuart Williams, HP</li>
<li>Dan Connolly, W3C</li>
<li>Tim Bray, Sun. XML 1.0 co-editor</li>
<li>Ian Jacobs, W3C. Tech writer</li>
</ol>
</div>
<div class="slide"><h1>TAG Participants, all dressed up</h1>
<div class="photo">
<a href="http://www.flickr.com/photos/ndw/sets/998264/">
<img align="right" alt="TAG members in ties!" src="http://static.flickr.com/33/45604663_72c19787da.jpg" /></a>
</div>
<ol>
<li>Henry Thompson, W3C and U. Edinburgh. XML Schema editor</li>
<li>Norman Walsh</li>
<li>David Orchard</li>
<li>Vincent Quint, INRIA</li>
<li>Tim Berners-Lee</li>
<li>Dan Connolly</li>
<li>Roy Fielding</li>
<li>Noah Mendelsohn, IBM</li>
</ol>
</div>
<div class="slide"><h1>Why an Architecture Document?</h1>
<ul>
<li>To distill ten years of experience with the hypertext Web</li>
<li>To help developers of Web technologies avoid pitfalls</li>
<li>To provide guidance to users, site managers, software designers
on promoting a robust Web</li>
<li>To build consensus around concepts and terms</li>
<li>To learn humility...</li>
</ul>
</div>
<div class="slide"><h1>Community Brings Issues to TAG</h1>
<ul>
<li>23 Jan 2002: <a
href="http://lists.w3.org/Archives/Public/www-tag/2002Jan/0182">Potential architecture</a> issue brought
to TAG's attention on public list
www-tag@w3.org
(<a href="http://lists.w3.org/Archives/Public/www-tag/">archive</a>):
<ul>
<li>
HTTP GET deprecated in <a
href="http://www.w3.org/TR/2002/WD-xforms-20020118/">XForms Last Call
Working Draft</a> (<a
href="http://lists.w3.org/Archives/Public/www-tag/2002Jan/0199">background</a>)</li>
</ul>
</li>
<li>29 Jan 2002: TAG <a
href="http://lists.w3.org/Archives/Public/www-tag/2002Jan/0223">accepts</a>
issue <a
href="http://www.w3.org/2001/tag/issues.html?type=1#whenToUseGet-7">whenToUseGet-7</a>.</li>
</ul>
<p>Teleconference and mailing list discussions ensue.</p>
</div>
<div class="slide"><h1>TAG Explores Problem Space</h1>
<p>What makes HTTP GET important?</p>
<ul>
<li>HTTP GET designed so that URI alone encodes interaction; allows linking</li>
<li>Safe/unsafe distinction in protocol enables user agent
support</li>
<li>Requests with no side-effects enable caching of results:
<ul>
<li>At global networking scale, require fast data exchange</li>
<li>Use caching to improve performance (see <a
href="http://www.w3.org/Protocols/HTTP/Performance/Pipeline.html">HTTP/1.1
study</a>)</li>
<li>Design protocols that support caching</li>
</ul>
</li>
</ul>
<p>8 Apr 2002: Dan Connolly <a
href="http://www.w3.org/2002/04/08-tag-summary#When">receives
assignment</a> to write strawman proposal. This evolves into
a draft <a
href="http://www.w3.org/2001/tag/findings"><strong>finding</strong></a>.</p>
</div>
<div class="slide"><h1>TAG Coordinates to Build Consensus</h1>
<ul>
<li>4 May 2002: Making connection to Web Services, David Orchard proposes
<a href="http://www.w3.org/2001/tag/doc/ws-uri-05042002.html">SOAP
HTTP GET Binding Version 0.1</a></li>
<li>3 Jun 2002: David Orchard <a href="http://lists.w3.org/Archives/Public/www-tag/2002Jun/0019">receives
assignment</a> to request of XMLP Working Group that
SOAP 1.2 (then a Working Draft) HTTP binding include GET method.</li>
<li>10 Jun 2002: TAG approves finding
<cite><a href="http://www.w3.org/2001/tag/doc/get7.html">URIs,
Addressability, and the use of HTTP GET</a></cite>.</li>
<li>10 Nov 2002: TAG <a
href="http://www.w3.org/2002/11/25-tag-summary">announces agreement</a>
regarding use of GET.</li>
</ul>
</div>
<div class="slide"><h1>Groups Document Consensus</h1>
<ul>
<li>24 Jun 2003: <a
href="http://www.w3.org/TR/2003/REC-soap12-part2-20030624/">SOAP
Version 1.2 Part 2: Adjuncts</a> becomes a W3C Recommendation,
with GET as part of HTTP binding (see <a href="http://www.w3.org/TR/2003/REC-soap12-part2-20030624/#RPCResourceRetrieval">section 4.1.2</a>)</li>
<li>22 Sep 2003: TAG accepts revised finding
<cite><a href="http://www.w3.org/2001/tag/doc/whenToUseGet-20030922">URIs, Addressability, and the use of HTTP GET and POST</a></cite></li>
<li>14 Oct 2003: <a
href="http://www.w3.org/TR/2003/REC-xforms-20031014/">XForms 1.0</a>
becomes a W3C Recommendation, with support for HTTP GET.</li>
<li>9 Dec 2003: <a
href="http://www.w3.org/TR/2003/WD-webarch-20031209/">Architecture Document to
Last Call</a> with discussion of safe interactions
in <a
href="http://www.w3.org/TR/2003/WD-webarch-20031209/#safe-interaction">section
3.5</a></li>
</ul>
<p>Ongoing: Marking safe operations in WSDL</p>
</div>
<div class="slide"><h1>Negotiation tactics</h1>
<ul>
<li>March 2004: TAG struggling to get webarch document done</li>
<li>Dan was more optimistic than Paul Cotton:
<blockquote>
Let the record show that PC and DC have bet a dinner that at this rate we won't (PC) or we will (DC) get to Recommendation by 2005.
<a href="http://www.w3.org/2004/03/02-tagmem-irc.html#T10-27-36">W3C Tag meeting notes March 2004</a>
</blockquote>
</li>
</ul>
</div>
<div class="slide"><h1>It's a REC! Party!</h1>
<p><a href="http://dm93.org/2005/0501brio-plaza/0501brio-plaza.html"><img align="right" alt="Connollys on the Plaza" src="http://dm93.org/2005/0501brio-plaza/0501brio-plaza-Thumbnails/6.jpg" /></a></p>
<ul>
<li>December 15: with a week or two to spare, W3C issued webarch as a Recommendation. (with <a href="http://www.w3.org/2004/12/webarch-pressrelease.html.en">press release</a> and all)</li>
<li>Paul graciously conceded the bet</li>
<li>Connolly family dines on Paul's nickel</li>
</ul>
</div>
<div class="slide"><h1>What type of information is in the Architecture Document,
Findings?</h1>
<ul>
<li>Properties we desire of the Web, and </li>
<li>Design choices to achieve them.</li>
</ul>
<p>Example related to previous issue:</p>
<dl>
<dt>Principles, Constraints, Good Practice Notes</dt>
<dd>
<ul>
<li>Agents do not incur obligations by retrieving a representation
("GET is safe").</li>
</ul>
</dd>
<dt>Rationale</dt>
<dd>
<ul>
<li>Benefits of URI addressability: linking, bookmarking,
caching</li>
<li>Benefits of distinction in protocols of safe/unsafe: user
agent alerts, caching</li>
</ul>
</dd>
<dt>Stories and examples</dt>
<dd>
<ul>
<li>Examples of safe (lookup) and unsafe (credit card purchase) interactions</li>
<li>Considerations for sensitive data</li>
<li>Practical considerations, ephemeral limitations</li>
</ul>
</dd>
</dl>
</div>
<div class="slide"><h1>Architecture Tripod</h1>
<ol>
<li>Identification</li>
<li>Interaction</li>
<li>Representation</li>
</ol>
</div>
<div class="slide"><h1>Identification I: Why URIs?</h1>
<p>Value of common syntax for global identifiers:</p>
<blockquote>
"Great multiplicative power of reuse derives from the fact that all
languages use URIs as identifiers: This allows things written in one
language to refer to things defined in another language. The use of
URIs allows a language to leverage the many forms of persistence,
identity, and various forms of equivalence."
-- <cite><a
href="http://www.w3.org/2001/tag/doc/whenToUseGet-20030922">URIs, Addressability, and the use of HTTP GET and POST</a></cite>
</blockquote>
<ul>
<li>Power in the network effect.</li>
<li>Extension through URI schemes</li>
</ul>
</div>
<div class="slide"><h1>Identification II: URI Usage</h1>
<ul>
<li>Comparison. Key to Semantic Web, caches</li>
<li>Dereference. Discussed below in Interactions</li>
</ul>
<p>Due to global scope, URIs also used outside of Web protocols (e.g.,
as database keys).</p>
</div>
<div class="slide"><h1>Interaction I: Dereferencing a URI</h1>
<ul>
<li>Communication between agents involves URIs, messages, data</li>
<li>Dereference a URI, get back a representation of resource state</li>
<li>Representation consists of representation data
and metadata (e.g., Internet Media Type).</li>
</ul>
</div>
<div class="slide"><h1>Interaction II: Dereferencing a URI (illustration)</h1>
<p class="center"><img src="http://www.w3.org/TR/2003/WD-webarch-20031209/uri-res-rep.png" alt="A resource (Oaxaca Weather Info) is
identified by a particular URI and is represented by pseudo-HTML
content"/></p>
</div>
<div class="slide"><h1>Interaction III: Managing Representations</h1>
<ul>
<li>Internet Media Type</li>
<li>Representations evolve over time as resource, technology
evolves</li>
<li>Consistency in representation increases trust in URI</li>
<li>Content negotiation facilitates evolution</li>
<li>Fragment identifier semantics and content negotiation</li>
</ul>
</div>
<div class="slide"><h1>Interaction IV: Issues Raised by Interaction</h1>
<ul>
<li>Safe, unsafe interactions</li>
<li>Sensitive data</li>
<li>Access control independent of identification (cf. <a
href="http://www.w3.org/2001/tag/doc/deeplinking-20030911">Deep
Linking Finding</a>)</li>
<li>Metadata from representation provider authoritative.
Other behavior ok, but requires transparency for user.</li>
</ul>
</div>
<div class="slide"><h1>Representation: Data formats</h1>
<ul>
<li>Data formats used to organize representation data</li>
<li>Data format considerations: binary v. text, extensibility,
versioning, composition, modularization</li>
<li>Hypertext</li>
<li>XML-based data formats: links, namespaces, qnames, media types</li>
</ul>
</div>
<div class="slide"><h1>Architecture Review of US Patent Office Web Site</h1>
<p>Review of the <a href="http://www.uspto.gov/">United States
Patent and Trademark Office</a> revealed:</p>
<ol>
<li>HTTP GET used for database lookup (good)</li>
<li>HTTP GET used for unsafe interactions (not good)</li>
<li>URI for patent is actually URI for search (not optimal)</li>
<li>POST used to protect sensitive login data (design choice)</li>
</ol>
</div>
<div class="slide"><h1>HTTP GET used for database lookup</h1>
<p>HTML "GET" form used for <a
href="http://patft.uspto.gov/netahtml/search-bool.html">database
lookup</a>:</p>
<pre>
&lt;form action="/netacgi/nph-Parser" method="GET"&gt;
</pre>
<p>Use GET for queries, searches, database lookups.</p>
</div>
<div class="slide"><h1>HTTP GET used for unsafe interactions (not good)</h1>
<p>Modifying state of shopping cart is unsafe since produces
side-effect:</p>
<ul>
<li><a
href="http://patft.uspto.gov/netahtml/search-bool.html">Search</a>
with keyword "hypertext"</li>
<li>Select <a href="http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&amp;Sect2=HITOFF&amp;p=1&amp;u=/netahtml/search-bool.html&amp;r=1&amp;f=G&amp;l=50&amp;co1=AND&amp;d=ptxt&amp;s1=hypertext&amp;OS=hypertext&amp;RS=hypertext">Patent 6,678,889</a>: "Systems, methods and
computer program ...."</li>
<li>Add to cart, view "Quantity" (1)</li>
<li>Hit back button, add to cart, view "Quantity" (2)</li>
</ul>
<p>"Add to Cart" an HTML link:</p>
<pre>
&lt;a href=".../AddToShoppingCart?docNumber=6,678,889..."&gt;...
</pre>
<p>I cannot link to shopping cart from this slide; a search engine or
pre-fetching agent might increment counter (cf. <a href="http://www.w3.org/TR/SVG12/#prefetch">SVG 1.2, section 11.8</a>.</p>
<p>In HTML, use "POST" form for unsafe operations.</p>
</div>
<div class="slide"><h1>URI for patent is actually URI for search (not optimal)</h1>
<p>What might a URI for a patent look like?</p>
<pre>
http://www.uspto.gov/patents/p6678889
</pre>
<p>Note that this is globally unambiguous; better than "6678889"</p>
<p>Search produced this URI for search on "hypertext":</p>
<pre>
http://patft.uspto.gov/...s1=hypertext&amp;OS=hypertext...
</pre>
<p>Search produced this URI for search by patent number 6,678,899:</p>
<pre>
http://patft.uspto.gov/...s1=6,678,889.WKU....
</pre>
<p>Why are these URIs different if this is the same patent?</p>
</div>
<div class="slide"><h1>Cost of Arbitrarily Different URIs</h1>
<p>At first, I thought these URIs were arbitrarily different
URIs for the same resource. If so, machines cannot
compare reliably, so:</p>
<ul>
<li>Interferes with caching</li>
<li>Semantic Web does not work</li>
<li>Site management more complex</li>
</ul>
</div>
<div class="slide"><h1>Identify Results of Search, not Search</h1>
<p>Resource only indirectly identified as query result.</p>
<ul>
<li>My expectation is not to bookmark search, but result of search</li>
<li>Search might return different results another day; I want to
refer to the patent.</li>
</ul>
<p>Related in Architecture Document:</p>
<ul>
<li>Avoid URI ambiguity.</li>
<li>Publishers of a URI SHOULD provide representations of the
identified resource consistently and predictably.</li>
<li>URI persistence</li>
</ul>
</div>
<div class="slide"><h1>POST used to protect sensitive login data (design choice)</h1>
<ul>
<li>GET allows URIs (bookmarking, back button), but
we don't want sensitive data in URI.</li>
<li>Choices include:
<ul>
<li>GET with HTTP Basic Authentication over SSL: Sensitive
data in HTTP headers, so allows bookmarking. User agent
manages passwords.</li>
<li>POST over SSL</li>
</ul>
</li>
<li>However, cost to SSL as well</li>
</ul>
<p>Think about these architecture issues, tradeoffs during design!
See <cite><a
href="http://www.w3.org/2001/tag/doc/whenToUseGet-20030922">URIs,
Addressability, and the use of HTTP GET and POST</a></cite></p>
</div>
<div class="slide"><h1>Future work</h1>
<ul>
<li><a href="http://www.w3.org/2001/tag/issues.html">Issues</a> the TAG has not resolved for First Edition</li>
<li>Other systems that make use of URI space: Web
Services, Semantic Web</li>
<li>Internationalized URIs (<a href="http://www.w3.org/2001/tag/issues.html#IRIEverywhere-27">IRIEverywhere-27</a>)</li>
<li>XML canonicalization</li>
<li>Binary XML (<a href="http://lists.w3.org/Archives/Public/www-tag/2002Jan/0182">binaryXML-30</a>)</li>
<li>Mixing XML Namespaces (<a href="http://www.w3.org/2001/tag/issues.html#mixedUIXMLNamespace-33">mixedUIXMLNamespace</a>)</li>
</ul>
</div>
<div class="slide"><h1>Questions?</h1>
<p>For more info:</p>
<ul>
<li><a href="http://www.w3.org/">W3C</a></li>
<li><cite><a href="http://www.w3.org/TR/webarch/">Architecture of the World Wide Web, Volume One</a></cite><br />
W3C Recommendation 15 December 2004</li>
<li><a href="http://www.w3.org/2001/tag/">W3C Technical Architecture Group</a></li>
</ul>
</div>
</body>
</html>