On Fri, Sep 06, 2002 at 12:51:35PM +0200, Hannes Gredler wrote:
> not the one that got committed - [have a look to the CVS version and
> you'll see that all the interesting fields get registered [not just text]
i suppose its a deficiency in ethereal that pdu-like fields can't be
added into a registered field (each pdu would overwrite the previous,
right?). is there a proper way to handle this that i'm missing?
> maybe can you add your v{1,7,8} support,which is great, on top of the CVS
> version pls ?
why would i waste hours doing that when mine already: supports everything
the existing one does, registers all the fields found within the packet,
supports all versions seen in the wild, has better bounds checking (you
can walk past the packet in the checked in version with an invalid pdu
count in the header), is already factored to support multiple versions
easily without code replication, uses the netmask info to make a prefix,
properly calculates the date/time fields.... the list goes on and on.
even with the new version that was checked in that supports v1 and v7,
mine still supports v8 and all the subformats within v8.
and as i've said before[1]:
# alternativly[sic], if you give me cvs access i'll just maintain mine
# in the ethereal tree instead of my own perforce repository.
i won't, however, waste my time just to fit it on top of someone else's
incomplete work, simply because they beat me to the punch of saying
"please check this in". even the version available when i first notified[2]
this list (back in june) of my dissector was leaps and bounds more
complete then the currently checked in dissector.
i have yet to hear a single reason why the packet-netflow.c in cvs
shouldn't simply be replaced wholesale (w/o the module glue) with my
dissector[3].
sorry if the above seems a bit hostile, but i'm completely oblivious as
to why this isn't a no-brainer.
--
- bill fumerola / fumerola@xxxxxxxxxxxxx / billf@xxxxxxxxxxx / billf@xxxxxx
1. http://www.ethereal.com/lists/ethereal-dev/200209/msg00068.html
2. http://www.ethereal.com/lists/ethereal-dev/200206/msg00260.html
3. http://www.mu.org/~billf/packet-netflow/packet-netflow.c
