The site should be built using travis-ci every time a commit is pushed to the source branch. In this way I can just commit to origin/source and not have to execute any code locally.

Perhaps the nested structure for short filenames should more closely follow the structure for long filenames.

YYYYMMDD/experimenter/sample/experiment/HHMM.extension

In this way, its easier to simply copy this nested structure into data.

The master branch needs a README.md file as follows:

Repository for Joshua Smith's public blog.

This repo contains at least two branches: master and source. The master branch is a container for the HTML created by pelican and used to serve the site via github sites. The source branch contains the source files for the blog and can be thought of as the main branch of development.

If you are first cloning this repository, you will want to set up the source branch like so:

git checkout -b source origin/source

This issue pertains to my Python testing style guidelines post.

Arbuckle's "Python Testing: Beginner's Guide" does an excellent job of explaining how python's unittest framework works as well as how to think about testing a python module. My post is mostly about how to organize tests.

Overview

Rename master branch to main for obvious reasons.

Update to commit f7e4fefa

Consider a sample die that's been diced from a wafer. At some point, devices are going to be fabbed on that die, and it will be useful to give each a UID. Having a UID per device is especially useful when an experiment is done on a single device that generates a single data file.

Fancybox is an example.

My CV needs to be included.