Translate

Sunday, December 7, 2014

365QOD- Day1400



Fuzzy Random Inputs

"Fuzzing applying random inputs and looking for some unexpected outcomes"- Unknown

We tend to follow known rules often in our lives.   Once we figure out that given input A well result in output B, we can predict the output
  It becomes predictable and we move on learning another rule.

But there is a difference between predictable and optimal. In order to get to optimal level we must introduce a random fuzzy input A*.  Then we look for some unexpected result.   It could be B* or even Z*.  This randomness will nudge you towards that optimal that you will never get to by using predictable inputs.

A great example of this idea is when software is tested by people who would not normally use it.  They could introduce inputs that the normal user might not because they know better.   This is important to find bugs that traditional testers would not catch.

Today's question is:
"Do you test your ideas with fuzzy random?"

No comments:

Post a Comment