James Espie, Test Specialist

Grylls and Thrills

Using systematic and speculative tracking to solve software problems

Note: this blog post contains minor spoilers for the Netflix series You vs wild (s1e3). Maybe don’t read on if you care about that sort of thing.

You vs wild

Bear Grylls’ has a new interactive TV series on Netflix called You vs Wild.

It’s interactive! At points during the show, the viewer is called upon to make decisions about Bear’s next action.

For example, he’s only got room in his pack for either a slingshot or a grappling hook. You decide which one to take, and that shapes the rest of the episode. This can determine whether Bear succeeds in his mission or not.

It’s great fun! A bit like a Choose Your Own Adventure book you can do with the whole family.

(You can’t kill Bear Grylls at any point, as far as I can tell. We’ve tried.)

A lost dog

In one episode, Bear’s mission is to find a missing rescue dog in the Swiss Alps.

At a certain point in the story, Bear needs to decide how he’s going to try and track the dog.

The options are:

These are techniques we apply in software all the time!

They come in useful when trying to diagnose issues from a production environment. For example, when a customer reports a problem that you can’t reproduce. Or, if an unexpected alert fires from your monitoring tools.

Systematic Tracking

Systematic tracking means looking at the clues to find out what happened. For example, you might look at:

Using these sorts of clues, it’s often possible to piece together a picture of what the user did to cause an error. It also demonstrates the importance of having good monitoring and logging in place!

Speculative Tracking

Speculative tracking means trying to think like a user. Asking yourself, if I was using this product, what might I be doing to get myself in this situation?

This can mean getting creative. Some things you could consider:

Why not both?

The great thing is, unlike in You vs Wild, we don’t have to pick one method or the other. Often when my own team is confronted by a problem, we tag-team it. I start tracking down the problem using speculative techniques. Meanwhile, one of my teammates tries to tackle it using a systematic approach. (or, vice versa). This way, we can collaborate to figure out the root cause of our problem.

In my experience, it works really well!