Sometime people forget testing is a means to an end.  Testing does not serve any purpose in and of itself.  It is a means to create code that you high confidence in and that is easy to maintain, and extend.  Just because code is untested doesn't mean that you shouldn't use it.  Only that you be aware of the trade-offs.  Here I make that point in a highly sarcastic manner.

Good old memory leaks.  It's not even the finding them that's the problem.  In most cases it's figuring out why their occurring.  I'm finding memory leaks in a perl codebase currently.  I've even fixed a few already.

I'm somewhat ambivalent about being able to fix some old code at work.  On the one hand I get to rip out a bunch code that was essentially masking some naive sql.  On the other hand it should have never happened.  We should have dealt with the underlying problem directly when it first came up. 

I had some thoughts about software development best practices along with the suprising lack of certain key tools and infrastructure at many companies.

I finished setting up a simple DDNS configuration with on my OpenBSD box. I wrote a quick summary.

I'm trying out Google's site builder to see if I can make a decent personal set of webpages for myself.