and I'm keeping score on paper how many fail and how many succeed.
Last Friday evening, before I went home I had 123 testcases that went well, and 15 that failed.
This morning, without grabbing the latest copy of the application, and without changing anything to my test set (it's all run locally so I'm complete lord and master of my test-environment) I suddenly had just 118 testcases that went well, and 20 that failed.
But after I checked a few things in the database (still not altering anything) I ran the complete set again, and I again had 123 good testcases and 15 that failed.
It bugs me because:
A) I can't reproduce the problem.
B) It doesn't give me much confidence in the quality of the system.
C) I can't explain that problem to my boss. |