>Jim,
>I am beginning to think that this may have been a problem with a specific compile of my app. I am curious about the connection between printer activity and code execution. What is the nature of the relationship?

I don't know all the details - this was pursued here about 3 years ago.
But the facts of the situation had the same exact 'feature' as yours as regarded a known to be present/existing/clean/working elsewhere simultaneously index would suddenly come up as not found. The case also had other oddities that were similar in nature (i.e. not an index not found, but other things illogically causing errors).

As I said, their 'fix' was found somewhat accidentally, as I recall and they elected to simply move the print server function elsewhere. While I know that this was the 'definitive' solution in that case I don't think the WHY was ever discovered.

good luck