Wood, HTML & Action Research

The process of constructing this bliki, or knowledge-base, or whatever I am calling it this week, is both frustrating and illuminating. Last Friday I set up the new version of La Jurnalo the weekly newspaper for La Mentala Rosario, the project we are running in Second Life. I used the blogging software WordPress and the process was very smooth and very simple. The five minute installation really did take five minutes and the structure of the templates made customisation extraordinarily simple.

I found a template that looked more or less like a newspaper and then customised it so that it looked more or less like I wanted it. It took a couple of hours to explore the php and css and change what I wanted to change. At the end of a morning I had a site that was up and running and (I suppose) a part of the blogosphere. Certainly it is capable of managing trackbacks, sending RSS feeds, linking to Technorati, and so on.

This is the opposite of my experiences here, where everything is taking a long time, and nothing works automatically. As I have explained elsewhere this is deliberate, and I was reminded of why I am doing it this way when I reread an old article about Jeff Hawkins that talks about his famous wooden blocks.

When he was developing the Palm Pilot Hawkins spent six months wandering around with a wooden block that he “used” every time he wanted to use his pda. According to Wired people thought Jeff Hawkins was crazy when they saw him taking notes, checking appointments, and synchronizing a small block of wood with his PC, pretending all the while that the block was a handheld computer.

“If I wanted to check the calendar I’d take it out and press the wooden button,” the father of handheld computing told an audience of Palm developers at PalmSource 99 on Wednesday.

Hawkins did the same thing when he designed his new Handspring handheld, which has an expansion slot for adding hardware extras like a mobile phone.

“I walked around answering phone calls with this block of wood, and of course it didn’t do anything,” he said. “I did it to see if it worked. I decided it worked pretty well.”

This is something similar to the strategy that I have adopted, although my wooden block does actually do something. I started with a (very good) bare-bones wiki which had no blogging features. I then simulated functions by adding them manually. You can see some of this in the page of ready made code chunks where there are some fragments of html that I use to put the keywords at the bottom of the pages. This is ridiculous, of course. A process like this is precisely what php is supposed to automate.

I also add the information to the recent entries page by hand too: copying and pasting it from the top of each page, and then formatting it by hand.

I have learned a lot by doing this. I have learned which functions I need, and why I need them. Adding everything by hand is an ideal way to test if it really is worth the effort or not. There is a lot of incentive to not bother, which leads directly to questioning whether the effort is worth the time.

I realised last week that some of the conclusions I posted seem obvious, if not completely self-evident when presented at the end of an essay. Rereading the essay I realised that they can even sound like straw men: fictitious problems that I have invented in order to demonstrate how clever I am solving them. This is not the case, and the current organisation of this site bears this out.

The menu structure divides the site into sections. The first section is called blogging, because when I set up the layout this seemed an obvious section to have. Now it seems unnecessary and confusing.

The reality is that there is no “blog”. There are simply pages of writing with different subject matters, and differing levels of topicality. Experimentation, using the wooden block principle, has proved to my satisfaction that nothing would be lost, and many things gained, by scrapping the blogging category completely.

I will do this in the next few days, when I have a couple of hours to spare. It will mean rewriting the menu, which will take five minutes. However it will also mean going through all the so-called blog posts and reassigning them to different sections and categories, which will take a lot longer.

Some presumably count as research, while most will be writing of one sort or another. This raises another question I need to look at: evolving a consistent taxonomy to allow me to find material later. The whole keyword (tagging) system needs some wooden blocks of its own…