I’ve never used Twitter or Facebook. Am I missing something important?
I still use all of the above sometimes, but mostly I use these free, open-source alternatives:
A few years ago, I got tired of Windows upgrades breaking stuff that had been working, so I switched to this user-friendly version of Linux. It’s probably less susceptible to viruses than Windows, if only because it’s less popular, and I find it less annoying. Mint comes bundled with lots of useful programs, including ones that can work with .docx and .xlsx files, though not all Office features are included and there can be some changes to formatting, placement of images, etc. At first, I installed it along side Windows, but stopped doing that after I found I rarely choose to boot into Windows rather than Mint.
I do still have a few programs that only run in Windows, such as the desktop software that lets me search and view hand-written lab notes from my Echo Livescribe smartpen. So I installed VirtualBox. That let’s me run XP (easy) and Windows 7 (trickier) in a protected “sandbox” within Linux Mint, where I control which directories the guest operating system shares with Mint and whether it can access the internet.
This is the search engine I use most often. Google still has lots of my personal information to sell, because my university makes us use Gmail, GoogleDrive, etc. “You’re not the customer; you’re the product.” But Google’s motto is (was?) “Don’t be Evil” so I guess I shouldn’t worry.
Encrypted synchronization of files among my computers, without storing it on anyone else’s server. It can only synchronize computers that are on at the same time, so I set up an always-on Raspberry Pi as a private server. This still doesn’t give me some of Dropbox’s backup and sharing options.
Python makes it easy to do simple, but time-consuming tasks, to Automate the Boring Stuff.
Recently, I’ve used ChatGPT to write simple Python programs, but it hasn’t done so well (for me) on more-complex programs. Here are some examples of major projects I’ve done with Python:
- Refree.py, a text-based Python program to manage references. This was inspired by, once again, having the commercial reference-manager software I was using becoming unavailable.
The graph above, of root-nodule fresh weight per soybean plant over time, with various mixtures of rhizobia bacteria, represents output from a pipeline of several Python programs:
- PhotosNweights.py runs on a Raspberry Pi in my lab. It guides students through weighing plants to measure their water use (a proxy for leaf area) and photographs their root systems.
- RootPainter (written and supported by Abraham Smith) identifies nodules in the root photos and outputs a file with size and position of each nodule. This is client-server software, so it’s a little tricky to set up, but then easy to use.
- NoduleTracker.py uses output from RootPainter to estimate nodule mass per plant.
- PickNplot.py uses the data file from NoduleTracker and a metadata file to pull out data for specific treatments and plot them, as shown above. I’m currently using the same program to manage data from Minnesota’s three-site Long-Term Agricultural Research Network, following these data-management guidelines from my colleagues, Eric Seabloom and Elizabeth Borer.
darwinianagriculture.com 