Web Architecture

Web Architecture in 1 line

When TimBl invented the web he wrote three things: HTTP, HTML and the URI. The URI was the most important thing because it allowed nearly everything to relate to nearly everything else, where:
scheme://host/path/resource.extension#fragment

"Just as there are many different methods of access to resources, there are a variety of schemes for identifying such resources." [RFC2396]

ftp://ftp.is.co.za/rfc/rfc1808.txt -- ftp scheme for File Transfer Protocol services
gopher://spinaltap.micro.umn.edu/00/Weather/California/Los%20Angeles -- gopher scheme for Gopher and Gopher+ Protocol services
http://www.math.uio.no/faq/compression-faq/part1.html -- http scheme for Hypertext Transfer Protocol services mailto:mduerst@ifi.unizh.ch -- mailto scheme for electronic mail addresses
news:comp.infosystems.www.servers.unix -- news scheme for USENET news groups and articles
telnet://melvyl.ucop.edu/ -- telnet scheme for interactive services via the TELNET Protocol

Few schemes which are centrally allocated by IETF/IANA.

www.w3.org is a Domain Name and corresponds to an Internet address
this name is allocated via hierarchical delegation
the .org registrar gave us w3, and we created www. (www is not required, merely a common convention. info.cern.ch was the first web host)
many, many, debates about who controls this system.

this frequently corresponds to a file system path, but doesn't have to as you can configure the server to redirect or serve anything for a given URI.
the host determines this an usually isn't that contentious though many organizations have strict publishing requirements internally.

Typically there is some resource that corresponds to a URI, that is dereferenced and used.
This example happens to be an HTML file; it could be a GIF, JPEG, flash, or mpeg file (these are registered as MIME types).
HTML is an application of SGML: SGML is a language for allowing others to create markup applications. SGML tells you how to create an element or attribute: HTML is a set of specific elements and attributes with a meaning like <H1>foo<H1> is a heading.
XML is a simplification and extension of SGML that has an initial low entry cost, and more complex, but cool, functionality.

A fragment doesn't mean anything to the scheme, it's a function of the resource's MIME type.
HTTP sends the whole resource even if the user only wants to look at the fragment.
The browser, knowing it is an HTML file, does the appropriate thing, in this case pages to the correct portion of the resource.

Does the design lead to centralized assignment, or distributed flexibility? If someone doesn't like your site, could your registrar take it away from you? (What are the policy/scalability implications?)
What happens if someone doesn't do the design correctly? Have you ever tried to bookmark part of a Flash site? You can't.
Interests try to bias the architecture to their own interest. New York city planner Robert Moses designed his roads and over-passes so as to exclude the 12-foot high public transit buses that carried people -- often poor or of color -- to the parks and beaches he also designed. AvantoGo uses a conduit to make PDA accessible content unavailable on the Web. (Give me the URL to the palm accessible New York Times? You probably can't find it. Balkanization of the Web serves their interest.)

Simplicity -- "Keep it simple, stupid!"
Module Design
Tolerance -- "Be liberal in what you require but conservative in what you do"
Decentralization
Test of Independent Invention -- If someone else had already invented your system, would theirs work with yours?
Principle of Least Power

Universality -- Any resource anywhere can be given a URI
Global uniqueness -- It doesn't matter to whom or where you specify that URI, it will have the same
Sameness -- a URI will repeatably refer to "the same" thing (does this mean the NYTimes front page, or a specific article in their archive?)
Identity -- the significance of identity for a given URI is determined by the person who owns the URI, who first determined what it points to.
Not a unique space, just universal -- URI space does not have to be the only universal space

Simple axioms, but great philosophical wars over the details:

Universal Resource Identifier (URI) -- includes all of the following. Your identity comes from the fact that you a unique identifier: 02139-0405
Uniform Resource Locators (URL) -- a location that can be dereferenced. Your identity come from your characteristics of your "location": Main St., Technology Square.
Uniform Resource Names (URN) -- " are intended to serve as persistent, location-independent, resource identifiers." Your identity comes from an allocated name by a recognized authority: NE43-350.
Uniform Resource Characteristics (URC) -- Your probabilistic identify comes from your characteristics: tall ugly building in the middle of all the construction.