Leading with Simplicity
TL;DR: pass all process operations through a DAG (Directed Acyclic Graph).
Assuming a weighty program, something with many 1,000s of requirements, we must start somewhere.
In development finding a starting point is fairly straightforward, building must begin with the foundation, then layers of logic are added on.
Get Todd Thomas’s stories in your inbox
Join Medium for free to get updates from this writer.
But, when testing a catastrophic failure that cascades into many subsequent failures we also must begin somewhere. The development case is easy to navigate; the testing path is less so but the same principle applies to both.
Given n number of failures, trace up the stack, of course. But, when multiple failures from separate functional components all contribute to the greater failure then the starting point is simpler than it seems:
Begin by testing affected components from the least number of dependencies
to the greatest.
I’m sure this is a principle in some comp-sci book somewhere but it seems to have drowned in a sea of noise. I’d like for the concept to find air again and regain its rightful place in the tool belt; it’s been saving me more and more lately.
I hope for this to become a common comp-sci axiom one day.
