Wolf Pack – a collaborative test compilation Martin Jansson
You are part of a pack of wolves.
You are hungry and have not found food for several weeks.
When you move, you run covering lots of ground quickly.
You are out hunting, cooperating and collaborating with the rest of your pack.
You are seeking the big game, not a flea, nor a rabbit or rat.
An elk is ok, a mammoth is great, but you look for a stranded whale or a leviathan.
You might take note of the smaller game, but as a pack it is not your focus.
When you find tracks or clues of the bigger game, you howl and notify the rest of the pack.
As a pack, you circle and take down the prey.
The lone wolf is no hero in this context.
Can this be an effective compilation when you get a new build, as a form of smoke test, acceptance or regression test. It could train you in collaboration, not crying wolf for small prey, but instead wanting the bigger juice.
It’s an interesting idea but we must look for the problems with any allegorical model to avoid potential problems in the language in which we think so here’s my take…
The places the allegory weaken, for me, are as follows.
1. Small game is small game. You can measure it with a ruler, but (speaking as a heuristic) there’s no way to objectively measure the size of a bug’s impact or importance. What is small game to you may be large to me. What is larger to you may be smaller to me.
2. The same problem applies to finding evidence of large game. What is this evidence? Do other wolves agree? What if the pack splits into 5 groups who think we should head in different directions? A direction-less, immobile pack of wolves is in a worse position than one with an alpha directing the hunt.
3. Lone wolf behaviour should be limited, but not demonised. Some of the best testing/learning work a tester can do is opportunistic exploratory work, going off at a tangent and perhaps finding big game (or evidence that may lead towards it).
4. What is the game in this allegory? Bugs? That assumes the testers are looking for bugs which isn’t what I believe a tester should do. My job is to explore, shed light on the state of the system, learn more about it and identify and mitigate risk. It’s in the identification of this that we provide value… so lone wolves all looking in various undocumented and unplanned directions could be the most valuable way for your pack to work. We don’t just check to see if a product under test can work, but how it might not work.
Some ideas for you, hopefully it’ll help.
Thanks for you comment Chris. I was vague on purpose to let the reader apply her own experience and ideas to this.
I think a tester usually have an idea what is a small game or a big game. As a tester you often indicate the importance of bugs or other test information you gather. If you are unsure, then collaboration and discussion with your peers and stakeholders is essential.
If you think you have found big game and want the rest of the pack to gather, then you are bound to have a discussion and sometimes disagreement. I think it is effective to look at different things, but when someone think they have found a big game you pause what you did before and make an all-out attack together to exhaust as many variations as possible. I see testers as intelligent who does not need to be pointed in new directions at all times, instead I expect them to follow hunches and embrace serendipity.
This is a compilation of collaborative testing. It is not something you do at all times, just like you do not use one tool for everything. If the key to this is collaboration, communication and cooperation then a lone wolf who does not follow the base principals is not providing value, in this context as I see it.
For me, Game in this context is what is valuable to find. It could be bugs, but a hunt for information is the core of what we do as testers. A lone wolf who does not share his findings is not helping the bigger goal. A group of testers bring their own backgrounds and perspectives to the area in focus. This enables the group to learn more and to circle problems in different ways than a single member could do.