Wireshark-dev: Re: [Wireshark-dev] GVCP Dissector

From: Joerg Mayer <jmayer@xxxxxxxxx>
Date: Wed, 17 Aug 2011 16:58:12 +0200
On Tue, Aug 16, 2011 at 11:04:30AM -0700, Mat Gara wrote:
> We are working with the GigE Vision standards committee to develop a wireshark plugin that will dissect the GVCP protocol.  At this point we have a plugin that works and we would like to check it into official wireshark repositories.  Currently there is an incomplete GVCP dissector that is part of wireshark and we would like to replace it.  How would we go about doing this?

- Make sure the formal stuff is ok (files have correct Wireshark copyright headers).
- Turn the plugin into a regular dissector (there are few cases except "rapid
  development" where a plugin really makes sense).
- Run the checks (at least tools/checkAPIs.pl, tools/checkhf.pl and the fuzz
  tester tools/fuzz-test.sh).
- Open a bug at http://bugs.wireshark.org/, upload the source code and (if not
  classified) a sample capture that can be used to run automated fuzztests in
  the future and set the button "review_for_checkin".

> Additionally in the future we want to make this dissector easily configurable so that hardware providers could add support for dissecting their custom register reads.  To do this we want to make use of XML configuration files and hence we would need to make use of an XML parser in this dissector.  One xml parser we have been experimenting with is Expat.  What would it take to integrate Expat support into wireshark?

Does it really make sense to do something like this? Why not add the code directly
into Wireshark. For one, interpreting stuff at runtime is bad performance wise,
also it will be nice to have support in Wireshark for as many product specific
extensions as possible.

Ciao
       Jörg
-- 
Joerg Mayer                                           <jmayer@xxxxxxxxx>
We are stuck with technology when what we really want is just stuff that
works. Some say that should read Microsoft instead of technology.