September 21st, 2004

The Wheels Go Round and Round

I spent most of the day trying to figure out how to modify our error handler class so that multiple threads could write to the same log file. (This is not improved by the fact that our "log file" is actually a bunch of string arrays in memory for most of its life.) After playing with various designs for the stacked error handlers (which install on top of one another on a per-thread basis), I finally realized that the easiest thing to do was to dump the contents of each thread's log back to the parent thread's log when the thread terminates.

Of course, this meant all of the code I'd been playing with was completely useless. But Doug and I did get to explain the concept to the programmer who's going to implement it.

Now, I'm back to the Debt Scheduler, where I've fixed another parser problem in the new functions that I added. I'd ask "Why doesn't anyone test these things?", except that no one has started testing this bit of functionality yet and I'm still unit testing as I clean up around the edges of the feature.

The good news is that it mostly works. :)

One Project Done

Several days ago, I started trying to replace the battery in the mixer's CPU. Unfortunately, the CPU is rackmounted in a position which is out of the way -- so much so that wrestling it out, removing the 16 screws, opening the case, replacing the battery, closing the case, replacing the screws, and putting it back in the rack is not just a job, it's an adventure.

When I was done, the CPU didn't recognize the keyboard, so I couldn't reset the BIOS. I concluded that the keyboard had come partially undone and needed to be plugged back in. The only problem was that it was underneath a trans-Atlantic cable worth of patch cords plugged into the back of the patchbay above the CPU. And those cords were taut enough that they tended to pop out. *grumble*

Tonight, I unplugged each cable from the back of the patchbay, labeled it, and set it aside. Then I plugged the keyboard back in, verified that it was recognized, reset the mixer BIOS, and booted successfully. Then I plugged all of the patch cords back in. In theory, everything should be working now.

I hope.

In other news, Gretchen and I started working on the last sketch that we're responsible for writing for the ConClave show, but Gretchen was feeling a bit under the weather, so she crawled off to bed after we watched "Scrubs", which left me free to go work on Cablefest in the basement.

With luck, she'll feel better in the morning after a good night's sleep.