(I'm replying to this by cutting and pasting from the original message
on the eGroups page, because the mail didn't show up in my mailbox.
Mail at NetApp, right now, doesn't seem to be as reliable as one might
desire....
I'd ask whether anybody got any bounce messages for mail sent to me, but
if mail to NetApp isn't getting delivered, I guess they'd have to answer
the question by phone....)
> The same code works for dissect_pop, and does not run into problems with
> the three way handshake (it does not get called).
In the captures in which it worked, what was the link layer of the
capture, and were there any packets that were, say, padded to the
minimum packet size on that medium? If not, that might explain it - the
problem I saw was with an Ethernet packet padded to 60 bytes; the IP
code currently doesn't use its notion of the packet length and use that
to limit what dissectors below it see (I think I mentioned this in the
subsequent mail you mentioned, which I think I sent just to you, not to
the list).
