Change your system only when real work breaks it

How do we make updates to our knowledge systems?

One really good way is to stop expecting your system to work perfectly, and to stop fixing what isn't broken. Take a functional approach instead. Change the system only in response to some sort of problem or bottleneck that appears while you're using it to do real work.

When you approach a change, clearly articulate what isn't working and what it should look like when it does work. This simple rule of thumb keeps procrastination at bay. You stop changing things just to make them nice, and focus on the goals the system helps you reach.

Sometimes I like to leave a dysfunction alone for weeks, observing how it misbehaves and what it prevents from happening. Gradually, a solution crystallises in my mind. Given that time, I may come up with an interesting solution, one that ultimately improves the system much more than if I had addressed the dysfunction straight away.

Yours in Markdown,
Zhenia