2 min read

Lessons from Jamesway

Table of Contents

[template]

Looking back, a few things stuck with me from Jamesway, technical and otherwise.

What worked

  • Keeping the collector and the view separate paid off, because I could change one without touching the other.
  • Talking to the operators early meant I built the thing they actually wanted.

What I would do differently

  • I would write the mock hardware sooner. I spent too long testing against real machines before I had a fake one.

The takeaway

The biggest thing was that software near hardware has to expect the hardware to fail. I still design for that now, well before I write the happy path.