When the Fire Comes

⇐ previous
Posted
next ⇒

I grew up in a logging town in British Columbia. There weren’t as many forest fires as there are now, and they weren’t as severe, but as July turned to August people would make sure they always had a full tank of gas and a change of clothes packed for the kids. They’d stop on the street to talk to their neighbors about who was going to look in on Mrs. Kerr to make sure she was OK, or ask the pharmacy for an extra order of tyheir prescription because, well, you never know, do you?

Forty-five years later it’s time to start making plans like these for our research projects. The Trump administration has launched an unprecedented assault on science; thousands of people have been fired, thousands of others have had their funding cut, and things are going to get a lot worse before they (hopefully) get better. What should you do if you’re a PI and you get the email saying that some coked-up tech bro in Washington has cut your funding because you used the word “diverse” in a grant application eight years ago?

Richard Littauer and others are putting together a paper tentatively titled “10 quick tips for making your code last beyond your current job”, and I’ll share a link to it when it’s ready to read, but its recommendations are for the medium and long term. What do you do now if you have until the end of the week to clean up and clear out?

Let’s assume your team includes one post-doc, three grad students, and half an RSE, and that you are collaborating closely with people at two other institutions. Your data, code, papers, and collective memory are scattered across Google Drive, Slack, GMail, GitHub, and the departmental server. Here’s what I think you should do:

  1. Remember that the only plan that works is one you’ve practiced (or as my dad once said, a full tank of gas is no use if you can’t find the car keys). Once you and your team have made a plan, try it out: you will immediately discover two things that don’t actually work and three others you have overlooked.

  2. Make sure you can communicate with each other. My family didn’t own a walkie-talkie, but a lot of others did. Your team will need a way to coordinate without Slack or departmental email; a Signal group is one option, but a mailing list that includes everyone’s work and personal addresses is probably a better option. (Think attachments, search, etc.)

  3. Check your licenses. I hope you put open licenses on everything so that you’re legally allowed to do this; if not, bookmark this post and come back once you’ve done that.

  4. Copy everything. Sure, your code is in GitHub1, but what about your papers? Your grant applications? Your Slack conversations? It will take five minutes to make a (partial) list of everything you’ve looked at in the past six months; do that, then figure out how to dump it somewhere that you control and that isn’t (as) likely to be impacted by random acts of government2. (For example, I now rely on a Canadian company called Sync to store my data.) In an ideal world you would use this as an opportunity to tidy everything up, but in an ideal world we wouldn’t be having this conversation.

  5. Write things down as they happen. You’re going to be making important decisions under pressure, with partial information, and with other people; point-form notes of everything your institution told you to do, everything you actually did, and everything you’ve learned along the way are going to be as important to getting through this as the logs that paramedics, nurses, and doctors keep in medical emergencies.

  6. Tell people what’s happening. Some of your colleagues will already have gone through this and will be able to offer advice; others who haven’t might still be able to help in unexpected ways. (“Why don’t you and your students come and spend a few days with my team?” wouldn’t normally bring tears to my eyes, but neither would someone offering to take the kids for ice cream while you’re setting up a tent.)

  7. Help others. Grassroots efforts like the Data Rescue Project are trying to coordinate work like this; if you aren’t immediately impacted, please help them out. And please ask your professional or scientific association to grow a spine and do something more than issue a press release.

Later: see also A Dollar a Minute and this partial list of places to run to:

  1. There is no legal or technical reason why the US government couldn’t turn off access to GitHub (or Zoom, or Google Cloud) for people outside the US, or (more plausibly) put tariffs on their use. I’ll be doing another post about that scenario soon. ↩

  2. I really wish BioTorrents had caught on. I bet a lot of other people do now too. ↩