>>I agree.....hard-coding names is a bad idea. Guess what, I've seen source code from commercial applications that do it.

I've seen several examples of code in commercial application source where the code could not possibly have been tested because it would have crashed had it been executed. I only found them because I had the misfortune to create the weird cases that executed the code and caused it to crash.
That's not crappy coding. That's just human limitation.

I tell my clients that I work by trial and error, but it's mostly error.
If I type 10 lines of code, at least 6 of them will have a mistake in them. One of the reasons I love the VS IDE is that the intellisense finds most of them for me.
But even if they're corrected, they're still errors and I know that there are plenty of logic errors that intellisense misses.
As a result I probably spend more time designing test cases than the average programmer.
Just a couple of weeks ago someone exposed a bug in some of my code that's been executed at least a thousand times daily for the past ten years

Error-free code is an oxymoron and a dangerous myth.