John,

I'm running into a problem with the KMF system. I know the doc
variable is populated in slurp_doc(), but it seems to be getting
corrupted sometime before the final analysis step (in
actually_perform_final_analysis()). I'm not sure what's going on.

I've verified that slurp_doc() is doing what it should -- the dump of
doc after slurp_doc() is the first attachment. The second attachment
is doc right before actually_perform_final_analysis().

I tried following the execution path through the cleanup step, and
everything seems kosher there. It seems to be happening in the thread
dispath, which is really odd. I talked to Ken, and he said talk to
you.

Any ideas?

Nick