Re: Advice on software testing?

["Peter Neilson" <neilson -at- windstream -dot- net>]

Here's a scenario that occasionally happens, and that should underlie any  
test plan using unusual hardware:

- Tuesday: The sole beta-test version of the custom hardware device  
promised for Monday is ready for you to begin testing. You've read the  
specs and the code, and you have a plan. You have two weeks before it gets  
taken away from you.

- Wednesday: You're revising the test plan to cover the new items you  
discovered in the hardware and the latest changes to the software.

- Thursday: You begin actual testing, entering bugs into the reporting  
system.

- Friday: Testing is going well. You report that you might be done well  
within time frame.

- Monday: You discover that over the weekend the system was shipped out.  
"We needed to make it billable."

- Tuesday: You try to get your boss to push through to a level of  
management that can make the decision on what to do.

- Wednesday: The consequences of the decision now belong to you. Whether  
you have to spend the next two months in Irkutsk, or test remotely over a  
dial-up connection, or publish a lie that reports no bugs, it's not  
something you want to own.

How do I know this stuff can happen? Kansas City is not quite Irkutsk, but  
I'd not planned to be debugging software on a machine sitting on a factory  
floor. On the other hand, we did discover that the SCR-controlled forklift  
trucks created electrical noise that crashed the custom device. Nothing in  
the original specs had anticipated that problem.

On Tue, 09 Aug 2011 15:11:28 -0400, Cardimon, Craig <ccardimon -at- m-s-g -dot- com>  
wrote: