Guy Harris <guy@...> writes:
> The capture_xxx() functions are for doing a very quick and dirty dissection of
packets while we're
> capturing traffic, so that instead of just showing a count of packets as we
capture, we can show a count of
> particular types of packets. Therefore, the right thing to do with invalid
packets would be to count them
> as "other" packets and not bother doing any further dissection - just as we do
when doing a real dissection.
Right, thanks. I normally hide the capture info dialog, so I forgot about those
statistics.
I checked in a change to explicitly increment the "other" count for these
invalid Ethernet frames, even though they would have been counted the same way
anyway in capture_ethertype(), but I think it's a little clearer this way.