Just some random thoughts.

Pages:  1 2 3
New Subdomain
  Created - 2012/12/15 18:06:36+0000

I have launched a new subdomain, http://ebchecker.mattunderscore.com/ . This is devoted to the EbChecker model checker I created for my Master's dissertation. It is currently a little thin but I will add content to it as I go. I am announcing something I did yesterday (see date).

Disk Failure
  Created - 2012/11/25 20:48:42+0000

Within weeks of my website going live my host suffered disk failure. The raid was unable to recover the data. Fortunately after so short a time I have barely noticed. Particularly as my website is still subject to change. I'm back up now (obviously) and I have instituted an automated backup process.

Old Ray Tracer
  Created - 2012/11/25 20:57:36+0000

In the final year of my bachelor's degree I implemented a simple ray tracer using Java. In my final year I was taught about best practice and correct software engineering methods. I recently went to a job interview (I have been going to a lot of those recently) that made me think back to it. So I found the code and was horrified at what I had written. It was inefficient, there was dead code, badly named methods and variables, poorly chosen access modifier and questionable design. Watch out for new graduates.

Linked Data
  Created - 2012/11/20 19:42:11+0000

I have successfully published a little bit of linked data , using the 303 redirect pattern . Take a look at http://id.mattunderscore.com/person/mattchampion and http://id.mattunderscore.com/project/ebchecker ; don't forget to use a RDF browser if you want to see the content negotiation work. You can replace the "id" with "rdf" and put ".rdf" as the extension if you want to see the raw RDF. I will add more as soon as I get the chance but I am very pleased with how this test has gone. I have implemented it using good practice in the style of the University of Southampton (which you might expect as I have just graduated from there). I have implemented my own Filter for the content negotiation. Paul Tuckey's UrlRewriteFilter is very nice but the problem I had with it was not that it does 302 redirects, as some have noted, but that it uses regular expressions to match the accept header, not true content negotiation. This filter is now available as a maven artifact .

  Created - 2012/11/08 20:27:15+0000

I went to a job interview a little while ago for a company that uses Node.js. I'd always quite liked JavaScript (apart from the DOM methods) and it stood out in my mind as it is the most common prototype based object oriented language generally in use. This made me think about using it as a general purpose scripting language. I downloaded and built Google's V8 JavaScript engine and quickly had a standalone shell for JavaScript. I plan to play with this.