brokenco.de/_posts/2017-02-28-take-notes.md

62 lines
2.5 KiB
Markdown

---
layout: post
title: "Keep a lab notebook"
tags:
- opinion
---
I have been "[farming](https://twitter.com/croyfamilyfarms)" for a few years
now and as the beginning of the 2017 season in northern California approaches,
I wanted to share some advice to consider, regardless of whether you're a
gardener or not.
**Always keep a lab notebook**
For just about any hobby project I have found a lab notebook can be invaluable.
What I consider a "lab notebook" is just a simple spiral notebook, with dated
log entries, on what I have done, observed, etc.
Currently I have two different hobby lab notebooks floating around, one is my
"Farm Book" wherein I will log:
* What has been planted in which sector
* When I planted seeds, starters, etc
* When seeds germinated.
* Notable conditions. A particularly hot week, wet days, etc.
* When plants begin to fruit, and the quantities.
Over the seasons, I can refer back to my Farm Book and improve how, where, and
when I plant. Perhaps more importantly however, I can refer back to the Farm
Book after I have been "out" for a while, whether traveling or just not paying
attention, to figure out what projects/tasks I left off with.
For my [other hobby
project](https://www.flickr.com/photos/agentdero/29706573822/), my lab notebook
has been even more valuable. Not only acting as a truck-repair-log but also
storing data on diagnostics performed at varying intervals. In the case of my
old truck, without a running log of what experiments and diagnostics I have
performed, repairing it would be a futile effort.
I was first introduced to this practice by [Bluthe
Rocher](https://twitter.com/blithe) in her talk [The Scientific Method of
Troubleshooting](https://www.youtube.com/watch?v=h9YZXuUjyOs) in the context of
software development and troubleshooting. Of course you can, and should, use
something approximating a lab notebook during the course of building software.
> **Sidenote**: Using Git commits and a `NOTES.adoc` in a repository is a great
> way to accomplish this. Many times, on my own projects, I will commit an
> experiment that failed with a commit message explaining what didn't work, and
> then revert that commit.
While it's easy to understand the value of additional rigor in software
development, I urge you to consider keeping a lab notebook for any other hobby
or side-project of any importance to you.
Inevitably, you'll find yourself staring at your project on a Saturday morning
wondering "what the hell was I doing here?"
Keep a lab notebook, and give yourself the answer.