If you haven’t noticed by now I’ve been writing quite a few questions of the day for SQLServerCentral. It started, as things often do, as just something to do that was different (before this year, my last one was in 2006!) but it’s grown beyond that. It’s a really interesting short form of writing and it’s writing that gets intensely reviewed by readers for content (because every word matters when you’re doing a puzzle). I’m consciously experimenting with style and substance. I’m trying to write easy questions and hard ones. Trying to write hard questions about easy topics and vice versa. Writing questions that require experience and inferences to solve. Writing longer scenario questions. Experimenting with tricky and not-so-tricky. I can’t say I’ve boiled it down to a formula, but I have a better feel for what will work and what doesn’t. Still, the goal is to push the limits of the box, or redefine it – what can I learn while teaching something useful?

Last week I wrote a question that on it’s own that is mildly interesting, but I wrapped it in the trappings of a Sherlock Holmes mystery. Partly for fun, partly to remind people (especially question authors) that writing questions is often about providing subtle clues. For this question the story doesn’t add much to the question, it’s really window dressing, but there is a second question that I ask them to figure out that may at turns be far too hard or far too easy, depending on the question taker. Window dressing is fun. Fun to write, fun to read, but almost always stuff you can cross off when it comes to figuring out the problem, yet many people in “real” troubleshooting sessions often get distracted by that same kind of stuff. It will probably show up in 6 weeks or so and we’ll see how it does.

The big goal I’m thinking on is how to write a question that spans a week. Something that starts Monday and continues every day and ideally gets harder because it builds on the answers and understanding of questions earlier in the week.