On 7/12/07 9:45 AM, "Gerald Combs" <gerald@xxxxxxxxxxxxx> wrote:
> If you release a
> binary dissection module, what's to stop someone from
feeding it fuzzed and
> randomized data, mapping the input and output, and
reverse-engineering the
> protocol itself? Dissectors aren't device
drivers, and their very nature
> (i.e. "tell me everything you know about
the data at offset X") means that
> once you release a dissector, you've
also released the details of your
> protocol.
A very good point, but I think the issue many face is less about actually
keeping things secret as it is about getting the lawyers and other
pointy-hair types to approve of releasing the code.
--
Phil