Monday, July 11, 2005

Return from EuroPLoP 2005

I’ve just returned from the EuroPLoP conference in Germany. For those who’ve never been to a PLoP conference you don’t know what your missing - get along now!

Anyone will do - except for the fact that I hear the original PLoP (outside of Chicago) has changed its atmosphere in the last couple of years so maybe that is not a first choice. VikingPLoP is a great little conference but it is small. Still EuroPLoP isn’t that big, about 45 people this year, much less than the 65 when I first went in 2003.

I return from EuroPLoP as usual, exhausted physically (too much beer, too many late nights) and mentally (so much to think about, so much to read, so many good conversations) but also inspired and ready to move on to the next year.

(Unfortunately while I was at EuroPLoP some bombs exploded in London - you’ve probably heard about this already. London is my home town, I was born close to Liverpool but I now consider London home. It feels like someone has attacked a member of my family. Still, I didn’t let it ruin my conference too.)

A few things deserve mention here.

First, much to my surprise, I was award the 2005 Neil Harrison Shepherding Award. This is a great honour for me, although I feel “I’m not worthy”. I look at the names on the Staff and they include great people: Norm Kerth, Joe Bergin, James Noble and Frank Buschmann. Thank you to the programme committee, I am honoured.

Second: I ran a long focus group (5 hours over 2 days) with Lise Hvatum on the subject of Conway’s Law. This was a fascinating exercise that opened up all kinds of ideas and insights. We learned too much to write it all here. We will be writing a report for the conference proceedings in the near future. For now lets just say Conway’s Law isn’t necessarily a Law.

Next: there where several analysis patterns at the conference, including a couple in my workshop. Didi Schutz asked me: “Are Analysis Patterns really Patterns?” I think we all tend to accept Analysis Patterns as Patterns but I’m no longer sure they are Patterns.

A pattern should tell me what to build, it should tell me how to go about building it, the pattern is named after what you build. The activity of Analysis is not about building, it is about understanding, decomposing and deconstructing. Therefore, how can you have a pattern about this?

Because patterns are about building they are about synthesis. Obviously, analysis is not synthesis. In fact, I’m in very close to revisiting Henry Mintzberg’s argument about strategic planning. (I recommend The Rise and Fall of Strategy Planning for more detail.)

Finally: I’m conscious that there is much pattern lore that is handed down by word-of-mouth, and in a somewhat random fashion. If you have the right conversation you will learn about “the but form of forces” but if your unlucky nobody will tell you about the “noun phrase” rule for pattern naming. At a time when individual patterns are becoming less important and pattern sequences more so I feel there is a need to get this information to new pattern writers.

Writing a pattern should not be hard. You should not require years of learning. Patterns are usually written by domain experts, if you need to be an expert pattern writer too then we are going to loose a lot of opportunities for patterns.

I could write more but I’m going to draw the line there for today.