So damned close

I worked a lot this weekend on this database project for United Health.  I was there until about 8:00 Saturday night and about 6:30 last night.  I lost a good 6 hours to what appeared to be a design flaw.  Turned out it was crud in the sample data that should never happen in real life.  The bright side is that there's now all sorts of error trapping for unlikely errors.  *sigh.  The down side is that I wanted that project done this weekend and now I still have a few hours to go.