At 11:18 AM -0500 3/9/06, Mike Woodworth wrote:
all makes sense so far, right? but it's this regression testing that has me confused. My understanding is, I want to do regression testing at a very high level - testing for the symptoms of the bug... so that if a new bug crops up with similar results it will be caught, even if it stems from a different underlying problem.
I'd call that more "integration testing" than "unit testing," but yes, it's a good idea when you can do it.
well, my app is a client server app - completely async communication, etc. and most of the bugs i'm really worried about writing regression tests for are the ones that show up as a problem between multiple clients keeping data in sync, or objects unsyncing later... all async stuff that occurs across multiple applications - they tend to be messaging problems.
That sort of stuff is going to be very hard to test for automatically. If I were you, I'd probably settle for this:
use the process of fixing the bug to insert unit tests in all the code i'm changing, and hope later problems get caught at a lower level?
HTH, - Joe -- Joseph J. Strout [EMAIL PROTECTED] _______________________________________________ Unsubscribe or switch delivery mode: <http://www.realsoftware.com/support/listmanager/> Search the archives of this list here: <http://support.realsoftware.com/listarchives/lists.html>
