Switching between teams is brain wrenching. The regression test last weekend turned up a some anomalies. The first few were easily explained by the differences in implementation of C libraries on Windows and UNIX. Been there, seen it, conquered it, have a drawer full of T-shirts. The other couple of issues took most of yesterday and today. An hour was spent looking at the 20 lines of code that are responsible, trying to find obvious flaws such as variables not initialized, incorrect values, and a few other things that might account for it. Nothing, the code was solid. Turning to the test itself, it finally hit me. The tester set up the scenario with a bunch of inter-related data. If the data does not start out in the correct state, or the order in which it run is not correct, the test would act the way it did. I rechecked that everything ran in the correct order. Yup, nothing wrong there. Checking the initial state of the data, reconstructing what went on when, and sure enough, the problem was uncovered. Next, the hour and a half explanation of what went on to the interested parties. The day is mostly shot and my brain is still churning. The bright side is that the release is now shipping a day ahead of schedule. I pick up the kids and I need to go over the spelling words with Rachel. I find two lists in her folder. I know one is from last week, and one is the current one. My brain is balking at another "what happened when" exercise. It should be something as simple as finding a date on the paper, but I can't do it. After Rachel is in bed, April calls and tells me which one I should be looking at.
End of Entry
No comments:
Post a Comment