Patch testing

In Darril's book, he states that the goal of basic testing "is to ensure that a patch does not introduce new problems." Describing regression testing, he states that it is "used to detect any new errors."

To me, that sounds like they are one and the same. Isn't an error also a problem? Just looking for a little clarification.

Find more posts tagged with

Comments

paulgswanson

I just read that part about an hour ago, I ws wondering the same thing. They sound synonomous.

kriscamaro68

The errors usually are related to what is being patched/specific to what the patch is patching or if the patch fails during deployment. As far as introducing new problems that can be a different story. Your patch may work great but could break something completely different. Say you patch something related to I.E. everything patches fine and there are no error's but with that patch something changed that now breaks Microsoft Office. Now you have a new problem.

RoyalTech

I kind of understood that. What I didn't understand is how the testing process would be any different. What you are explaining is simply the result of the patch. At least that's how I understand your reply.

ChooseLife

Regression testing verifies that everything that used to work still works after the patch.
Basic testing is a vague term and can mean different things, although in the industry it is often used interchangably with regression testing.

RoyalTech

That's pretty much how I interpreted it. Thanks.

paulgswanson

Regression testing verifies that everything that used to work still works after the patch.

Thats a great explanation! Thats going right on its own flashcard.