Taking Stock: October 2002

Wednesday, 09 October 2002

Let's be honest, my technological experiments have stopped dead in their tracks. Current job uncertainty, the fact I'm not sure what country I'll be living in next year, plus a Japanese exam approaching... it's quite obvious that something had to give. My research into .NET and Linux is suspended, for the time being, until life gets back to normal. Until then, I don't expect to post new articles to Wander regularly.

However, I can give an update on what has happened.

Linux Sure Looks Nice

My problem with Linux is that I haven't got a project. I installed Mandrake Linux 8.2 onto my hard drive and then realised I wasn't sure what to do next. I have to come up with a pet project and then I'll start getting somewhere. Maybe I'll look at existing open source code out there, or start following instructions on how to compile your own kernel. I came across Linux From Scratch recently, which tempts me.

Stuck on Planet .NET

I had begun to work on the language-learning tool I wanted to build in Visual Basic .NET. However, I got stuck very early in the process. I wanted to use a database that would hold all of the words for the application. Using the handy JET (Microsoft Access) database seemed like a good idea, and I thought using my old copy of Access 97 would be enough. However, I want to store Japanese characters and that has caused big problems.

I don't know whether it is the version of SQL Query I have got, the Access version I'm using (Access 2000 uses Unicode), or perhaps just the general non-Japaneseness of my Windows setup. I'm sure I could use something heavy-duty like MSDE 2000, but expecting users to install SQL Server on their systems just for a language test seems like overkill. I could fall back on using a flat file, but remember that the whole point was to learn ADO.NET.

So I paused, because I wanted to do this right. However, time has run out, and now the project is on indefinite pause.


While I was working at Wilco Tokyo, I had been carrying around a floppy disk. On this disk, amongst other things, was a small nondescript file called project notes.txt, in which I recorded all of my observations of project management. The contents ranged from the mundane, like getting workstations and logins ready for new team members, to the difficulties experienced when e-mail is the primary method of communication. I kept promising myself to tidy this random sprawl of comments into something more readable, but recently...

I finally got around to reading Peopleware, by Tom DeMarco and Timothy Lister, widely acknowledged as the book to read on software project management. After reading it, I wonder whether I should just discard most of my project notes. Peopleware is such an excellent book. It covers many important elements of project life but its central thesis is that projects are all about people - the development team - and not the technology.

Reading the book was an uplifting experience. It validated a number of my suspicions regarding projects but, even better, forced me to re-evaluate some of my assumptions. It's full of easily-digestible anecdotes that clearly illustrate the ideas being conveyed. Further, many of the conclusions arise from actual research rather than simply relying on the authors' personal experiences; this adds much credibility to the propositions put forward in the book.

For example, there's a whole chapter that gives a far better exposition of the Zone than I did. De Marco and Lister's term for it is taken from psychology - a state of flow. In fact, a significant portion of the book is dedicated to the problem of getting teams into flow. For those who have been made redundant, the chapter on "Human Capital" makes very interesting reading, indicating that there is much to be changed in the way that companies are perceived by the stock market.

I can see a lot of similarities with W. Edwards Deming's work (see The New Economics, which condemns the practice of performance measurement) and also that by Gallup (see Buckingham & Coffman's First, Break All The Rules, use employees' existing talents because talent cannot be taught). It is no surprise to discover that DeMarco and Lister read Deming before writing Peopleware.

Today's recommendation: if you're interested in project management, then read Peopleware. It may be slim and a little expensive, but it's worth it.