Hi, PAul.

PE> FM> Did you do anything about the buffering? That's what *seemed* to be
PE> FM> taking the time, and may not show up in your profiling.

PE> Say that again slowly.  BFN.  Paul.

IMHO (only) the slowness of PKTJOIN and PKT2QWK seemed to come from the
very considerable amount of physical disk access - it seemed like there
must have been minimal/no buffering in those programs.

And since that's not your program's code, I thought it *may* not have
been evident in your profile results - if it's only timing the stuff
which is your actual code.

However, since Brenton made my disk accesses hum with smarter use of
smartdrive, I don't have the problem.

Regards, FIM.

 * * Any fool can tell the truth, it takes sense to lie well.
@EOT:

---