Personal tools
You are here: Home Tech Tidbits Site Maps

Site Maps

Site maps can be a simple listing of pages with or without links to the pages. There is a standard xml format that Google/Yahoo/MSN and others use to aid with discovering web pages on your site.

Site maps are helpful for people trying to find a particular section, page or even just try to grasp the logic of how the site is set up. Site maps can also assist search engines find all the pages on the site in a systematic way. There is even a standard XML format that several search engines have adopted which aid their discovery system (see http://www.sitemaps.org/). RSS is related, but is primarily used as a discovery method for recently added pages.

But for the One Big Database, a site map should be in XML and allow the navigation of the site to be exposed in addition to linking to all the pages and documents. For example, if a section of your site contains press releases the structure of the site should indicate this (e.g. www.organization.org/pressreleases/date-and-or-name.html ). In this way the structure of the site becomes a mini object database that can be used for robust, but simple, XQuery queries. See the example of using the navigation box on this site as a site index/map for the purpose of bringing back links from those resultant pages.

So I propose (or would appreciate someone alerting me to another) creating a new standard for site maps and indexes. See an early example of what this might look like below. Note that the web addresses/URL's are the only text that is outside of the tags and that the nesting of the site folders/directories indicate where the slashes should be.

<site-map>
<base-uri encode = "UTF-8" >http://www.site-mapped.org</base-uri>
<section type = "folder" title = "" method = "uri-list" >
    <uri type = "http://www.w3.org/1999/xhtml" metadata = "http://purl.org/dublincore" xml:base = "http://www.site-mapped.org/" title = "Home Page">index.html</uri>
    <section type = "folder" title = "1st Section" method = "uri-list" ><folder>first-section</folder>
        <uri type = "http://www.w3.org/1999/xhtml" metadata = "http://purl.org/dublincore" xml:base = "http://www.site-mapped.org/first-section" title = "Section">index.html</uri>
        <uri type = "http://www.w3.org/1999/xhtml" metadata = "http://purl.org/dublincore" xml:base = "http://www.site-mapped.org/first-section" title = "Help">page1.html</uri>
        <uri type = "http://www.w3.org/1999/xhtml" metadata = "http://purl.org/dublincore" xml:base = "http://www.site-mapped.org/first-section" title = "More Help">page2.html</uri>
    </section>
    <uri type = "http://purl.org/rss/1.0" metadata = "http://purl.org/dublincore" xml:base = "http://www.site-mapped.org/" title = "RSS Newsfeed">rss.rdf</uri>
</section>
<site-map>

Note that sites that are built without the One Big Database or ReSTful architecture often do not use the folder/directory system (i.ed. slashes dividing section contents) to divide sections. Instead they use SQL column and row call outs (e.g. www.nonrestfulsite.com/page.php?section=21&page=43&randomsqlinject=droptable ). In understanding the One Big Database, it informs the structure of sites. The proposed standard would essentially allow a site to be a highly structured database, from the URL exposing the sections and the actual web pages/documents having easily discoverable content.

Document Actions
What's News