Tuesday, August 30, 2011

Hybrid Manual/Automation Tests

I've had this idea before. Suppose you have a set of tests that are difficult to automate because some middle step needs either human interaction or would benefit from a human performing the validation.

Well, Amazon's Mechanical Turk is one option. Another option was just announced as well. With these tools, you can add in a human task to your automation tests to make your Manual / Automation hybrid tests.

So... Any tool authors out there that want to make this available?

Inspected by A-12345

(from http://www.qclabels.com/)

Every so often, I find a slip of paper in a new pair of jeans that reads "Inspected by 12" (or something similar). Then I realize that someone, or multiple someones, inspected this product for some level of quality before it was shipped.

What if you looked at the help menu entry for a particular feature and it would tell you who tested it? What do you think you'd notice? Would you attribute the good or bad quality of that feature with that tester's ability? If you were that tester, would you feel pride in a job well done?

I feel somewhat like I'm living 3Ms motto, "We don't make the products you use, we make the products you use better".

Tuesday, August 23, 2011

Structured Creativity Lessons?

I saw this book, published in 2007 called "Wreck This Journal". It is a series of journal entries where you are asked to basically demolish the book, one page at a time, in a very specific way.

The instructions include things like "Rub dirt on this page", "Paint with coffee", "Paste a picture here and change it"

It may seem totally random, but I expect that it will encourage creativity in a fun way.

Learn to Code With Friends

If you've wanted to learn how to write code, but didn't have a place to start. Check out http://www.codecademy.com/ . This is a way to learn to code where you are given feedback as you learn in a very intuitive way. You can also do this with friends to see how they are doing. Check it out!

Thursday, August 18, 2011

Wednesday, August 17, 2011

Robert's Rule of Risks

"When a risk manifests itself, it becomes an issue. For big issues, the real question is 'Was this from a known risk?' "

Friday, August 05, 2011

Is SmoothTeddy Really 'work-related'?

In the end, I came up with this banner as a sample for a 'logo contest' for our internal QA community. Let me give some background in how this came about.

I ran across 'Smooth Teddy' a few years ago as a drop-dead simple way to create interesting 3D figures that don't look like they came from a CAD system. Here is a bit more of the 3D 'bug' I was able to make. (Sure, it's not particularly 'buggy' but hopefully, you can see the scope of the things that can be done.)

If you end up using Smooth Teddy, be sure to read the documentation because some of the coolest features are not obvious. (The program itself is a bit unstable, so save often).

The remaining question is 'Is this 3D tool work-related'. Well, only if you want to allow people to find their way to learn. I'm certain it has nothing to do with the 'fun' factor of using the tool :)

Wednesday, August 03, 2011

Framework to Allow User to Recover Gracefully from Hung Applications

So some researchers were looking into how to get MS Word to recover when it gets hung and came up with a general framework to allow users to attempt to get hung applications unstuck.


This is a great case of a truly useful tool that can be 'an enabler' for bad programming practices. Surely, this is a good tool for end-users. But developers providing this as a workaround for problematic code can be a bad strategy.

Tuesday, August 02, 2011

Removing Bias From Your Strategy

While this article is geared towards corporate strategy, many of the concepts can be applied to product strategy, test strategy, etc.

http://www.mckinseyquarterly.com/The_case_for_behavioral_strategy_2551 (requires free registration)

Here are some of the tips provided.
- The process of developing the strategy is one of the largest contributors to success
- Change your 'angle of vision' to see the issue from different perspectives. For example, Apple is famous for fully developing multiple ideas for the same feature to come up with a single implementation that fully implements the new feature.
- In most organizations, simply expressing high confidence in a plan is enough to get it approved. However, that does not result in success. Approval should be based on a clear recognition of the uncertainty involved.
- Shake things up, such as setting 'stretch goals' that are not possible given the current ways of doing things. This can spark innovation and efficiencies.
- Call out 'Silo thinking'. If departments hold onto their own departmental interests at the expense of company interests, call it out.
- Allow for 'Safe Debate'. Group-think can elimnate dissent, but sometimes dissent is needed. The key is to be able to do it without personalizing it.

There are more interesting topics on this site. Look around and enjoy.