Saturday, May 14, 2016

Arguing Micheal Bolton's 20 statements about testing

Michael Bolton, one of the most recognized testers worldwide (assuming we take into account the number of followers), has recently published 20 very interesting tweets on his account. This leads to a very interesting discussion, which I recommend reading too. Unfortunately, Twitter wasn't designed for such lengthy conversations and it's hard to grasp what's going on there. Don't get me wrong though, I like Twitter and you should too. It seems like most of the IT professionals use it every day and share quality stuff there.

The thing is, as much as I consider those statements valuable and interesting (hence they're all cited here), I can't fully agree with all of them. Twitter char limitation (140) isn't enough to express my doubts, so let's move the discussion to the blog world.

Edit: It seems like I may have taken Michael's words out of context, which I hadn't been aware of. Turns out, his views on testing are pretty much like mine. See the comments for clarification.
Such a negative and disempowering statement. I had already talked about the ever-changing tester's role these days. Yesterday I had a pleasure to listen to Micheal Feathers' talk on Geecon titled 'Testing patience'. I managed to capture the key slide from the talk (see below). Why would we strip ourselves from the power to redefine testers' role in IT? It's pretty much in our hands.


Fully agree. Testers aren't magicians and saviors. They won't change the flawed design by night. They can however point out the problems early.

This is again a surprisingly disempowering statement. Testers should be part of the team with all the privileges and responsibilities. If production is down I don't see why testers should not actively seek root cause and solution. Especially when they have vast experience using the system.

Agree. Some time ago it fashionable to treat testers as quality gatekeepers. It simply didn't work.

Classic fingerprinting problem. That's why testers were integrated into development teams and gatekeeper concept buried. Everyone is responsible for quality.


Nice. I like that.

Yeah, sometimes it's not easy to show problems. Especially when the teams are remote and we have additional cultural differences. That's why books like the psychology of software testing are needed.


Again, psychological stuff. Software development is a social activity with all its nuances...

Not sure about that. It seems like an error during exerting control. In mature companies, this shouldn't be the case.

Obvious. There is no such thing as unwelcome help.

We fix the problems. We make the decisions. Why would we distance ourselves from the problems?

We change things, we prevent the problems. Testers help throughout every project phase.

Don't like that. Strict role borders lead to handoff waste. We should aim at as much blurring of roles as possible in software development. That's much more efficient.


We can prevent problems. Just like everyone in the project. Not sure why we would disrespect someone by doing that.

Those statements seem like a contradiction to #13 and all those disempowering before.  Finally, an acknowledgment that the testing role can be very broad. Fully agree.

We should take credit and responsibility just like everyone else in the company (including HR girls). Testers are no exception here.