PURLs of Wisdom
URL's are the glue that binds the web together. Being unique, they can also serve as definitive "name" and be used like "tags."
Everybody on the Internet uses URLs (Uniform Resource Locators). In web browsers, they are the glue that binds documents together and allows those documents to live in a rich environment connected to every other document. URLs in common usage refers to the subset known as web addresses or links. Web browsers allow people to locate and read documents based on the URL of the document. However, this seemingly obvious observation clouds some of the real understanding and power of the web. URLs in the RESTful way of thinking about things, are representations of things. Many web developers leave the creation of URLs to the software employed to run the website which can produce mangled or meaningless URLs (usually the web developer uses the <title> tag to give an understandable name to the document or even to just put a big headline at the top of the web page). The best developers and software helps create understandable, archival and well architected URLs. Beyond better bookmarks is the possibility of the URL to be human readable and even memorable. And the URL can even help explain the architecture of the web site which can help human visitors as well as software agents/bots/web scrappers/RESTful APIs.
(note: the creator of the web cares very much how URLs are constructed: see http://www.w3.org/Provider/Style/URI)
Quite often in web documents there are objects mentioned that might be worth further exploration. For example, a book might be mentioned. Normally the title of the book will appear. Very rarely will the authors, publishers, ISBN or other important identifying information be included. However, the title of the book might be linked by means of a URL to a web page on Amazon, the publishers page for the book, a Wikipedia page describing the book and very occasionally to an authoritative online electronic copy of the book. But these are not universal, unique and proper representations of the book. For example, the title alone gives barely a real help as there may be other things with the same name or title, like books, movies, peoples names, etc. And as most books are not yet truly online (and some have characteristics that make it impossible anyway), any link mentioned is limited by nature. Note that the <cite> tag itself is very lacking as well as not implemented in any real way.