This blog of course contains too much noise to really act as a CV, but lets pretend. Most posts on this blog are written fairly quickly and for fun, from idea to publishing, and would benefit from a fair bit of editing and occasional purging. I'm basically practicing writing in English - which is obviously not my native language. I aim to strike that careful balance between Strunk and White and John Fowles. Yoda: that is why you fail.
I will now look back on two posts from 2003 and 2004. I started to program 1984, professionally 1997 and did my first proper XP project around 2003 - we started trying a few years earlier. I joined ThoughtWorks in 2005.
Giving the build environment some love (2003)
As you can see on all the bold sentences, I was on some serious Jacob Nielsen trip at the time. This post talks about using Ruby to build Java and how to do Acceptance Test Driven Development. It was around the same time Rake came out, but we had our own thing called Raven (not to be confused with this tool), which supported Maven repos - the XML-based tools referred to are CruiseControl and Maven. During this period there was somewhat of a backlash against DHTML, so testing end-to-end without Javascript was possible.
The browser driver talked about was somewhat similar to the HtmlUnit version of WebDriver, and had a client API that mirrored our HTML components. On top of this there was what these days would be called an internal DSL for testing. Like many other similar setups, we used Jetty to start the actual application during testing. This way you could often ATDD a story in a few hours without ever looking at it in a browser but still be quite certain that it worked as intended. I worked quite a bit while living in San Francisco trying to solve this problem again for AJAX applications, but the result is not public. (Update: ThoughtWorks open sourced it in March 2011.)
We really liked Ruby and were past the "internal-tools first" adoption strategy, so mails like this seemed promising. We had written several Java web frameworks, but something was definitely lacking, and I naively predicted a quick death of Java web development. Sadly, I was wrong.
The source code is like a house, where objects live (2004)
A somewhat pretentious title and post, but it's dealing with software's wider effects on society, which is a very important subject. It discusses Christopher Alexander's views on both this and the fact that his thinking has been greatly influential in the software world. I'm currently quite uninterested in patterns and object orientation in general, as I'm caught up in the current functional programming hype, but like all things, this goes in waves and I learned Scheme and Java at the same time. I predict that I will re-read Smalltalk - Best Practice Patterns in 2013 and think, "this is awesome stuff!" as memory fades.
The overall point of this post still resonates very strongly with me seven years later. Have we been moving in the right direction during this time, to be better equipped to face what I called Alexander’s challenge - and he wrote this 1996 - or have we lost track while delivering business value?
I can of course also provide a normal CV at the last responsible moment. The title is from The Office, where Karen is one of two secretaries David interviews (but this clip misses the actual line). Enough about myself for a while. Time to move on!