[VOIPSEC] 4G Issue Map: signaling complexity

[Craig Southeren]

On Fri, 25 Aug 2006 01:42:18 -0400
"Hadriel Kaplan" <HKaplan at acmepacket.com> wrote:

..deleted

> > Only 2 or 3 times in my life have I ever seen the binary decode operation
> > fail.  The reason is that people generally use well-tested commercial
> > compilers and libraries from companies like OSS.
> 
> Wow.  We must live in different worlds.  I get "H.323" traces all the time
> Ethereal can't decode and I end up asking the vendor for their secret
> decoder ring.  This literally happened last week.  When Ethereal can't parse
> someone's SIP, I still can. (well, unless sigcomp is used ;)  

These are two different problems.

Failing to decode PER is a rare problem, but does happen. The SIP
equivalent would be failing accept both CRLF or just CR as end of line
markers (yes, I know the standard specifies what to use, but that does
not mean that all terminals do it right). Regardless, it's a bug.

Being unable to decode a vendor-specific extension is a completely
different issue, and can occur regardless of the protocol being used.

It's true that not having the "magic decoder ring" (i.e. the ASN.1
description) for a H.323 extension makes these difficult to
reverse-engineer, but that is no worse than putting binary into the
Content-Body of a SIP PDU, or having opaque hex fields in MIME headers.

I've this done in systems that use SIP for signalling but then add their
own "special sauce" in order to deny access to any clients other than
their own.

   Craig


-----------------------------------------------------------------------
 Craig Southeren          Post Increment – VoIP Consulting and Software
 craigs at postincrement.com.au                   www.postincrement.com.au

 Phone:  +61 243654666      ICQ: #86852844
 Fax:    +61 243656905      MSN: craig_southeren at hotmail.com
 Mobile: +61 417231046      Jabber: craigs at jabber.org

 "It takes a man to suffer ignorance and smile.
  Be yourself, no matter what they say."   Sting