Ethereal-dev: [Ethereal-dev] Re: MX/IP dissector

Note: This archive is from the project's previous web site, ethereal.com. This list is no longer active.

From: ronnie sahlberg <ronniesahlberg@xxxxxxxxx>
Date: Fri, 8 Apr 2005 06:47:09 -0400
Can you do the following changes to the dissector and resubmit for review:

1,
get_mxip_pdu_type()
seems really to do just what a normal value_string val_to_str()
combination does in normal dissector.

Can you refactor and remove all these functions and replace them with
normal value_strings and use real hf_fields using the
VALS(value_string) construct so that the dissector becomes better and
so that the filtering gui works better?

et_mxip_action is another one of those.
Please as far as possible get rid of all these
translate_value_to_string() functions and replace them with a
value_string.   use this value_string in the VALS() section of the
hf_field.

See lots of other dissectors on how thios works.


2,
change dissect_mxip() into a new style dissector that returns int and not void
and which returns 0 if this does not look like mxip protocol.
This will make it work properly when you have conversations between
two well known ports.
I.e. when both source and destination port are ones with registered
protocol dissectors.   MXIP would then look at the packet   see it
doesnt look like mxip and return 0,   ethereal would then see  oh crap
 mxip said its not mxip   ill better try the other poissible dissector
then.







On Thu, 7 Apr 2005 09:07:31 +0200, "Wemelsfelder, Armand"
<Armand.Wemelsfelder@xxxxxx> wrote:
> Hi all,
>  
> I recently wrote a new dissector for Ethereal (version 0.10.9) that 
> analyses our (proprietary but open) MX/IP protocol which lives on top 
> of UDP. It uses port number 8655 which is in the process of getting 
> registered. MX/IP is a management protocol that is developed by NKF 
> Electronics ( <http://www.nkfelectronics.com> www.nkfelectronics.com) to
> control their network video equipment. The 
> dissector has been tested and verified and I sincerely believe it is 
> correct. I only tested compilation on the MSVC/Windows platform, but I 
> also believe the code is portable to the other platforms as I tried to 
> adhere to the Ethereal coding rules with respect to portability.
> 
> There is one new file residing in the epan/dissectors directory:
> packet-mxip.c
>  
> The corresponding makefile.common in the same directory needs to be patched
> with makefile.common.patch to include this new file.
>  
> Best regards,
>  
>     Armand Wemelsfelder
>  
> 
>