Why N3 is better than XML

One of the more interesting pieces of functionality I added to DELI was a scraper that queries Google in order to locate new UAProf profiles. The scraper generated an XML file that could be transformed via XSLT to human readable XHTML or RDF. Having some RDF that links together all known profiles is potentially very useful, although you there are a lot of devices out there, so it is necessary to automate creating and maintaining this data wherever possible.

After a year's break (according to CVS) I came back to look at the XSLT code today, found it did not work, and I looked at it and I could not understand it. XSLT for me is a WRITE only language.

So rather than spend a lot of time trying to work out what I wrote, I decided to get rid of the XML and use N3 instead. N3 is another way of writing RDF and its a lot neater than RDF/XML. I've been rewriting the scraper to generate this format.

I also registered a PURL for DELI: http://purl.oclc/org/NET/butlermh/deli which currently maps onto http://www.cems.uwe.ac.uk/~mhbutler/deli. DELI was using URNs in order to generate unique identifiers for manufacturers and devices, but this is not ideal.

There is more work to be done - hopefully I will get time soon.