>> With all due respect ... I've tried it both ways, and it _does_ make
>> a difference for me.  But as always, YMMV.
>
>Ken I'd be interested in knowing if you've monitored for overruns and
>other errors during your tests.

Ummm ... I think that perhaps I made the impression that I did a thorough
scientific analysis.  That's not the case :-)

Here's exactly what happened:

- I remember reading (I _thought_ it was in Steven's TCP/IP) that a lower
  MTU would improve interactive response time.  So I set it down, and
  things _seemed_ better, but I didn't have any hard data.

- At a later date, I thought about it, decided that it really didn't make
  sense, and I set the MTU back up in my PPP startup script, but I didn't
  restart PPP.

- The next time I rebooted, I said, "why are my sessions so terrible in
  response time?".  I poked around, realized that I set the MTU higher
  and forgot about it.  I set it back down, and things started behaving
  the way I remembered that they did.

That's it.  Note that a couple of months passed between #1 and #2.  However,
the fact that I noticed a difference when I had forgotten that I changed
it convinced me.

--Ken