Sixty Hours later...

I have to tell you a funny story. OK, two funny stories.

***

Everything was working fine with the software by late Wednesday night. Or early Thursday morning. I started an overnight test, went home, grabbed a few hours sleep, got back to work to check the results. Everything looked good. Seems like we were set for the Release.

Got a call from Configuration Management. Did I realize that the revision number was incorrect? Seems that the previous person in charge of this project never actually did a release, so it should've been Version 1.

Hmmm. That's interesting. The previous person in charge of this project had told me we were on release 8, which made this upcoming release version 9.

Checking into the release database, it turns out they had never actually released a PRE-PRODUCTION version, only EXPERIMENTAL versions. (There are four types of releases: development, experimental, pre-production, and production.) We are required to release a pre-production version for Qualification tests.

But ... they ran Qualification tests on the first version last year, right? Well, I guess not. At least, not with a pre-production version of the software.

Oops.

Since the version number is embedded in the software, I had to rebuild it and re-run all my tests. So the release was delayed til the afternoon. But it got done. Now we could all relax and let the Qual testing begin.


***

The build was ready, checked in, submitted to Configuration Management, and available for the Hardware folks to use in their pre-Qual tests.

It was going to be so nice; I was going to be going home on time tonight!

Then I got a call around 5 p.m. The Hardware pre-Qual tests failed. My presence was requested in the other building (where they were running the tests). So, instead of going home as I had planned, I detoured over to the other building.

The whole lot of them were gathered around the test station, wringing their hands and casting aspersions at the software. I reviewed the test failure log and shook my head; couldn't figure out how that error had occurred. Unless ...

Then one of the guys suddenly got this weird look on his face and said, "I think I know what's wrong." We all looked at him as he walked around behind the test station and starting fumbling around with the cables, muttering. He was looking for something and not finding it.

Turns out, the test station wasn't set up right. Some signals weren't getting through because a signal converter box was missing.

The software had actually detected a hardware failure!

And that took us til 7:30.

Can I go home now???