Wrong, In Context

On the front page of the September 9, 2014, Wall Street Journal, we find that GE has exited the kitchen with the sale of its appliance business:

GE Exits

However, on page B3 the same day, the drop quote in an article would seem to indicate otherwise:

GE buys a food company

Of course, General Electric is not buying the food company Annie’s; the headline makes clear that the General in this case is General Mills (stock symbol: GIS).

But there’s nothing in the drop-quote to indicate something is wrong within its own context. Maybe General Electric often pays premiums like that during an acquisition. Maybe the copy editor or whomever did this dropquote finished the GE story from page 1 just minutes before working on the General Mills story.

However, we’ve got to retain context when testing and proofreading.

Where does this come into play in testing?

The foremost example in my mind is when we’re doing things to trigger error conditions to make sure that an error message displays. It’s possible that the system will throw up the wrong error message and we’ll miss it. I once wrote automated tests that triggered error conditions and parsed the error message (mostly to make sure an error message applicable to the screen and operation displayed). However, I did not write it smart enough to compare the error message that displayed to the error message expected. So when the application started failing by displaying the wrong message for the occasion, the tests didn’t catch it.

So you’ve got to remember to see the forest and the trees–along with the underbrush, the soil, the other flora, and the carnivorous fauna–when you’re testing.

Comments are closed.


wordpress visitors