Earlier last year, in November, I was assigned my first project… the day it was due on. No worries, I was told, it was a low priority one and we would have no problem with it being late. It was a simple project, all I had to do was to create a new output file that would contain raw transaction data for a client. It was simple conditional statement and a write statement to add to the program. How hard could it be?
I wrote all the code changes in one afternoon, fixed the bugs the next day, and took an entire week figuring out where I could find test data that actually had that specific client in it (apparently, in a whole month worth of transaction, they only showed three times). Once that was done, we put it in the test cycles, thinking that it was a job well done. It wasn’t. We forgot to put the date in the outgoing file, but then we were sure it would be good to go. And it was. Except for the fact that the tester are blithering idiots…
It has been almost two and a half month since the thing was in the test cycle, it was supposed to be approved for install on Tuesday at noon, but the test team did not start looking into it until a week before it was due. At first it was: “Hey, what are we supposed to test? What cycle is this running on?” and we sent them the design specs, which they should already have, and told them to verify the content of the outgoing file with the client. The next day, they replied :” The file is empty. What are suppose to verify?”. So we had to explain to them that the client did not show up frequently but that they could easily look up report to find the data they needed, like we did a couple month before. The week end has passed and it is now Monday before the deadline. “Could you post some transaction for us to test with?”, “No, that is not our responsibility, go find your own data.” Tuesday : “Guys we have tested anything yet, and the deadline is at noon!” And finally: “We did not get the approvals required for pre-prod sign-off, we’ll have to revert the changes”.
Great job guys! Next time, maybe you should read the design docs more than a week before your deadline. The application analyst I worked with on this project was not impressed either. We ended giving them the data I pulled so that they could test it, which is kinda dumb. It is a bit weird having all these rigid procedures for quality insurance and testing when, as far as I can tell, no one really bothers to follow them. I can already tell that this is only the beginning…
The Adventure of Alexandre in Nebraska! 