Making Plants Thrive

on October 23 2019, 7:38PM

It's often lamented that software projects become dead plants in an unloved garden: we excitedly keep buying new plants but we don't put in the time to see them thrive.

The appeal of building something new, playing with some fancy dependency or tool, trying out some new process; if only we could resist the temptation. But we shouldn't resist the temptation because this is the sign of healthy experimentation! It's far better to experiment in your spare time than to use your career as an excuse to try out the next shiny thing.

I'm a huge fan of "laboratories" where questions you have regarding code are answered by creating code and committing them to a central repository. Making them multi-language helps by reducing friction for testing things out. A graveyard of good intentions becomes a collection of prior discoveries.

This doesn't change the fact that we feel guilty that we can't keep the plant alive. It takes a little discipline, and maybe for some, a bit of prior knowledge, but it's not too hard to get things into place. In the same way we reduced friction by making a project multi-language, introducing automation to reduce toil is the best way for us to combat bitrot; if we can come back to projects knowing full-well they build, we are much more willing to continue to "water the plants". Making a project thrive comes in a few major parts:

  1. Testing and building the code before it reaches trunk/master
  2. Artifacts (library, binary, etc.) are created and published
  3. Said artifacts may be deployed to a server to run

Many other automations can be done too: linting, dependency updates, scheduled builds, et. al. Scheduled builds are cool (and underrated) because they continuously show projects are building and tests are passing. You now have extra capacity from all the built artifacts to handle services going down or security updates having been released. If you automate away the toil, you can treat a project less as a chore (by focusing less on the accidental complexity) and more as a labor of love (by focusing on the inherent complexity of the problem you are trying to solve).