Horizontal Vs. Vertical Knowledge

24 January 2004

This will bore some (many?) of you. Some of my recent experiences with technology have dealt with very arcane configuration parameters in a piece of software used for supply-chain work. If a company sells a widget that requires two sub-assemblies to be built prior to final assembly and there are pay-points or labor costs for each assembly and six or seven raw materials for each assembly, how do you make (this particular) software track that.

I haven’t done much of this kind of work since the mid-1980’s. You’ll recall how frustrated I became recently. There were many moments when I felt as if I was learning things I would never use again. They were one-off solutions to problems I didn’t even want to be solving. But, there’s always a silver lining…

This morning I’ve been tinkering with Movable Type and MT-Medic. MT-Medic is a nice tool for someone (like me) who is ultra-cautious about trampling around on the insides of a Movable Type installation. I don’t know why the configuration report from MT-Medic shows some different settings from the ones I know I’ve changed in mt.cfg, but it does. That’s a different subject.

Here’s what hit me: why am I enjoying mucking around in Movable Type, but not in a supply-chain software package? Several reasons come to mind, but the most important is, ”Movable Type is horizontal knowledge. Understanding how it installs on a web server and how permissions are set on files and how to make it behave as a content management system for virtually any business is radically different work from configuring a supply-chain package to handle a unique set of widgets.”

One of these is clearly horizontal knowledge that will be in demand for many years to come. The other is specific, vertical knowledge that I’m having to develop for a single client. It just feels better to be working on something that you might be able to apply over and over. Ultimately, the answer lies in doing work that fits this kind of requirement.

Filed under: