Today I’d like to share two tools that I have found indespensible in the course of my studies. Ironically, these have nothing to do with geophysics or seismic at all. But they’re out to save you.
Git is a very low-key version control tool. At certain points in the process of creating you can tell git to make a snapshot and store it. Git can then share these offsite, if you want to. It takes a week at most to get used to git and a quick google away, you’ll find all the help you need, to work with it.
It happened to me many times.
I have been writing a seminar paper, or writing some code, changed this, changed that, hit save, closed and the history was gone. Those changes, however, broke the entire code and there was no way back. It took me hours to remember, what I have done and sometimes it would not work at all.
With git I could just revert to a snapshot I took before the changes and I’m good to go.
Github is a service that can store your git snapshots, called repositories or short repos, on their servers. With easy encrypted access and they even support creating a DOI for citation of your code.
One caveat is that standard repos are open to anyone. However, as a student or someone in academia you can apply for an education account. They have given me 5 private repos, which is more than enough. After completion of a topic, I usually publish them, or take them down. Easy as that.
And github has saved my butt in a different way too.
At one point in my studies I was called to a professors office with my study group. The mail was dead on point “the is a plagiarism issue with your project”. I was incredibly nervous. I gathered any proof I could get that this code was in fact the work of our group and made sure with everyone in the group that their contribution was well cited or original content (it was). Then it dawned on me, we were collaborating through a github-hosted repo. A server, I could under no circumstances manipulate with the entire version history of our project. I printed it and took it all to the professor. Turns out, someone copied the project from an unprotected account in the university network, changed the variable names of the code and handed it in. They weren’t even able to explain the code, plus our watertight proof of authroship. That was a moment, where I was very happy about github.
Yesterday I came back to my computer being absolutely unresponsive. After restarting, a complete check of the harddrives was forced. Usually this would have wrecked my nerves. All my work gone? Au revoir thesis!
However, a couple months ago I went for a automatic backup software that is well recommended: Crashplan. I never thought I’d be able to get this to work on Linux, but they have an incredibly easy install script, which does in fact work without root. Now, I’m working with sensitive data, so off-site backups are a topic of security. Nevertheless, when the entire backup is encrypted with a 448 bit key, I’m not really worried, that there is any trouble coming up on the horizon.
When I started to study a picture of a poster on a bulletin board circulated.
?you can keep and sell the laptop I don’t care, but please return the CD in the tray marked ‘Diploma thesis’. No questions asked.
I didn’t know much by then, but I knew it did never want to be in that position. This gives me security I will never get there.
Backups and Version control
These two tools will improve your work structure and even if your entire system burns to the ground, you can recover the data, from the continuous backup archives of Crashplan.
What kind of tools save your nerves on a daily basis?
Latest posts by Jesper Dramsch (see all)
- Geophysical Assumptions - 2018-06-06
- Uncertainty in Geoscience – Bayes and Expert opinions - 2018-02-15
- Making Full-Waveform Inversion Uncool again - 2018-01-29