Sunday, April 29, 2012

The Beginner's Mind

"In the beginner's mind there are many possibilities, in the expert's mind there are few." - Shunryu Suzuki

For a long time, I refused to track test requirements at any level of detail. The rationale was that it took too long and had limited value. I was able to convince many people that I was right and that this wasn't needed. I 'knew' that this wasn't useful, so I didn't do it.

What really happened was that I had tried a couple times to track requirements at a very (very, very) granular level and gave up when the maintenance of these requirements was too much. I made the mistake of believing that my was was the only way and that since my was was not workable, no was was workable.

I've spent too much of my career 'knowing' what to do. This is not to say that I've come across as a know-it-all or not being cooperative when other (and better) ideas come along. I mean to say that I've not taken the opportunity to seek other opinions, especially from those with a 'beginner's mind'.

The 'beginner's mind' is not only found in beginners, but those that allow the many possibilities to be given full consideration, even when our experience tells us otherwise. We use our experience to help guide us in making better decisions with less work. However, these shortcuts can get in our way, especially when the basis for some of these shortcuts no longer exists.

Since these early days, I've been able to open my eyes to the possibilities and come up with a method of tracking test requirements that is actually maintainable. (go figure)

Are there things you 'know' to be true about testing software?
Have you examined those 'truths' to ensure their basis is still valid?

Examine your beliefs, see what you find.

Monday, April 09, 2012

If you know of any company that is looking to hire summer QA interns, this program looks like a good opportunity to get support for that and get connected with people interested in getting into the field.

If you are or know of someone interested in QA, check out the link!

Tuesday, April 03, 2012

Auntie Cueway

From time to time, people send me things to further my education, to bring a laugh to my day or to generally improve my moral standing. I've started getting these letters anonymously that seem to be from a 'Auntie Cueway' to what appears to be someone she is mentoring. She speaks as if she is mentoring a nephew, but I'm not entirely sure if this is accurate or not.

Rather than give you my impressions of these letters, I will reproduce them here and let you decide.

P.S. Apologies to C.S. Lewis. Any resemblance to 'The Screwtape Letters' may be intentional...

SUBJECT: Congratulations!

My Dearest Gunther,

Congratulations on your new role as team lead! It took quite a bit of patience and effort on my part to get the director to believe it was his idea to put you in charge of the team after your previous manager's 'accident'.  It's now your turn to put in the work we've been planning. Even you should be able to understand how important the kickbacks I get for our current test tool licenses are, so be on your toes.

Your loving Auntie Cueway

SUBJECT: Metrics


I was looking over the proposed Metrics Standards document you drafted and am very disappointed. It's all very well to require five pages of metrics each week for each team, but you must be more careful in picking your metrics.

Since I know you will be unable to figure this out on your own, let me be blunt. Take out the Defect Detection Percentage and the whole section on test requirement coverage. These are far too helpful for the director and the other metrics are seemingly helpful enough that he won't complain about the content. If he does, you need to be sure to speak to him as if he were a fool and suggest that if he doesn't understand these metrics, maybe he's not suitable for his position. That may be a bit risky, but if you use the right nuance, then he will internalize the comment as if he had come to that conclusion himself.

Your deovted Auntie Cueway


SUBJECT: Automation

My Darling Gunther,

From what I hear from the developers, your test automation strategy is coming along nicely. The tests are just unstable enough to often work, but not so unstable to warrant a re-write. They still require lots of manual intervention to start and then to review results. My favorite part is that you were able to convince the director that the work needed to automatically put the results into the test management system are so onerous, that it's not really worth the effort. If my calculations are correct, the automation is twice as time consuming as if you were running the tests manually. This is quite an achievement and will ensure that our test tool licenses will continue to be funded (along with my kickbacks).

But I do have a word of warning, the director is not so simple-minded as to be fooled for long. You must be able to quickly show improvements in the automation so that he doesn't bring in a consultant or hire someone that can see through your ruse.

Always your servant,
Auntie Cueway

SUBJECT: Not your best work

My little Gunther,

I'm not sure why you have disregarded my previous warnings about the director. I'm hearing rumors that he's beginning to reconsider your appointment. You realize that if you fail me, these lucrative kickbacks will stop and you will be out of a job!

I have always provided you with a good position and have led the way for you to rise within Corp's ranks while being able to keep my own involvement a secret. This is how you repay me? Your father tried to warn me when I wanted to hire you. He said you were not up to the task, but I stood by you. My only hope now is that you can show enough improvements in your team to head off a disaster.

It's clear that you are unable to think of these things yourself, so here is what you need to do. You need to push your team to write and run as many tests as they can, regardless of their value. Then, you need to write as many bug reports as possible. Take that notebook you've been keeping on system quirks and write up a half dozen bug reports for each of these. Now pay attention, this next part is critical. For each bug, you must follow your bug reporting guidelines perfectly. There should be no question that each bug report is well-written and represents a true quirk. You should give each version of these bugs a wide range of priorities so not to draw attention to what you are actually doing. Hopefully, with this explosion of new tests and defect reports, the Director will ease his suspicions.

Your Loving Auntie Cueway


NOTE: These are only the first few letters and I've not been able to read further. As the remaining letters are organized, I will post them for you to read.