For obscure reasons, the data for these races is currently stored in an LDAP database and various WWW forms fronting PHP scripts are used to maintain this database.

Then, an ldapsearch program is used to extract all the data in the LDAP database to build an XML document. In this way, the XML document describes all the races.

Every night and also on demand, six XSL Transformations are used to produce the five HTML pages and an RSS page (next.htm, newnext.htm, prev.htm, pb.htm, rest.htm, next.rdf) from the XML document.

Here is a link to the XML document containing details of all the races. Exactly what happens if you visit this link depends on the browser you are using. Modern browsers will attempt to display the XML. With older browsers, you may have to View the Page Source. Depending on your connection to the internet, it may take some time to download and for it to be displayed as the document is currently (July 2006) about 900KB.

The "future races" page displays an iCalendar button for each race. When this button is used, some iCalendar information will be produced. This is done by executing some XSLT code. The XSLT code was provided by Brian Suda under the W3C Open Source License. The XSLT code is described at http://suda.co.uk/projects/X2V/ and it is available from http://suda.co.uk/projects/X2V/xhtml2vcard.xsl.

The iCalendar information is produced using the mime type text/calendar. Exactly what happens at this point depends on how the visitor's browser has been configured. Either an application will be launched, or the visitor will be asked whether they want to start an application, or they will be asked whether they want to save the iCalendar information to a file.

   post  Barry Cornelius, Computing Services, University of Oxford, 13 Banbury Road, Oxford OX2 6NN, UK
phone +44 1865 273267      fax +44 1865 273275      email barry.cornelius@oucs.ox.ac.uk      links facebook flickr