Ethereal-dev: RE: [Ethereal-dev] Patch to packet SIP to make content type lower case, Patch to
>-----Original Message-----
>From: Olivier Biot [mailto:ethereal@xxxxxxxxxx]
>Sent: den 20 januari 2004 00:11
>To: Lars Roland; Anders Broman (TN/EAB)
>Cc: ethereal-dev@xxxxxxxxxxxx
>Subject: Re: [Ethereal-dev] Patch to packet SIP to make content type
>lower case, Patch to packet-isup to avoid owerwriting 'SIP' in Proto
>column
>From: "Lars Roland"
>|
>| Hello all,
>|
>| Anders Broman (TN/EAB) wrote:
>| > Hi,
>| > Just to clearify the intent:
>| >
>| > Currently the SDP dissector adds /SDP in the protocoll column e.g
>| > a MEGACO message with SDP descriptions would show up as
>| > MEGACO/SDP in the protocol column, this sort of make sense as SDP
>just decodes
>| > a part of the MEGACO message "written in SDP" the same is true for
>SIP and BICC.
>| > I was thinking that if there was a SIP message with ISUP ( and
>SDP ) parts included in it
>| > it would make sense to show it as SIP/ISUP/SDP as the "main"
>message still is SIP or perhaps only "SIP" ?
>| >
>| > Currently the message will show "application/ISUP/SDP" ( as SIP is
>owerwritten )at least it should be changed
>| > to "ISUP/SDP" or "ISUP(ITU)/SDP" in my opinion.
>| >
>| > Comments wellcome.
>| >
>|
>| I agree. Subdissectors should not overwrite "SIP" in the protocol
>| column. However I would set the protocol column to "SIP-T" and not
>to
>| "SIP/ISUP", but both solutions are OK.
>
>So tell me what I have to do here :)
Check in the ISUP patch :)),
Seriously: Puting SIP-T in the protocol column isn't so straight forward I think,
( I don't have a SIP-T trace so I don't know how a complete call would look)
as that would have to be dependant on Content type beeing equal to multipart/mixed
and probably would mean that a call sequence would have a mix of "SIP" and "SIP-T/SDP"
( asuming ISUP puts nothing there )entries in the protocol column ?
The alternative would be:
"SIP" and
"SIP/ISUP(ITU)/SDP"
I don't realy have any strong opinion either way.
Best regards
Anders
| BTW, last patch to packet-sip.c introduced a bug. Patch is attached.
The SIP patch has been checked in.
Regards,
Olivier