Hi
On Wed, Apr 13, 2011 at 5:40 PM, Chris Maynard <chris.maynard@xxxxxxxxx> wrote:
> Roland Knall <rknall@...> writes:
>
>> I have provided some time ago a patch for submission into wireshark
>> (Bug #5753). Over the course of the next two weeks a new version of
>> this patch would be completed, which would enable the dissector to
>> talk Modbus/TCP as well. But this version would be very preliminary,
>> and not stable for a longer time.
>
> "Not stable" or "not complete"? I think incomplete dissectors are OK as they
> can always be improved upon later as time permits, but unstable is not
> acceptable.
It is stable. It is in use around the office with various
configurations, and it has been fuzz tested. For the fuzz test I used
various network configurations to create capture files, and than
tested with that. Therefore, I can say with certainty, that it is
stable.
Also, there is nothing missing protocol implementation wise. Although
some though already went into implementing conversations and taps, but
besides that, it is feature complete.
> The latest patch you provided seems to have been reviewed by Guy, Jeff and
> Jakub. I *think* it's probably close to the point of being accepted, so I'd say
> rather than introducing another patch that "might not be stable for a longer
> time", that you should hold off until after this patch has been accepted.
> Another patch at this point might prolong the time it takes to be accepted, but
> I guess it depends somewhat on the extent of your new changes and if you really
> meant "not complete" rather than "not stable".
Ok, than I'll wait.
kind regards,
Roland