[Asterisk-Dev] libsrtp

[Derek Smithies]

Hi,
  I read in 

http://srtp.sourceforge.net/README

* The sequence number in the rtp packet is used as the low 16 bits
    of the sender's local packet index. Note that RTP will start its
    sequence number in a random place, and the SRTP layer just jumps
    forward to that number at its first invocation. 

There is no RTP sequence number in the iax packets. Yes, there is a 
sequence number, but that is iax specific.
 I think srtp might require some mangling to over come this.



=================

If srtp is revised-bsd-like, how does that go with the GPL license in * ?

My current understaing is the GPL requires that:
   all code linking to GPL code must also be GPL ?

Apologies if I have misquoted, but here are two quotes....
=========

See, http://news.com.com/2100-7344-5198117.html
One tricky part of the GPL is the "linking" question: How tightly coupled 
may a proprietary software package be with GPL-covered software? If a 
programmer wants to include GPL-covered software as part of a larger 
program, the GPL requires that the larger program also be released under 
the GPL--an effect some have termed the "viral" nature of the GPL.

But things aren't as clear when there's a looser link between the two 
bodies of software. 

===========

Linus wrote in:: 
http://www.uwsg.iu.edu/hypermail/linux/kernel/0312.0/1282.html
But if somebody else writes a piece of code for Linux, we now have a
situation where Linux is comprised of two pieces (a) the original work
(for simplicity, let's say it's copyright by person A) and (b) the new
work (copyright person B).

Now, (a+b) is the new work (joint copyrighted by A+B), and if the original
code (a) required the GPL, then (a+b) requires the GPL too due to the
"viral" nature of the GPL. You are right so far. 
 

====
My view is that it is a grey issue.. I would rather not have the lawyers 
using  * as a "test case".



Derek.


 ===================================================================
On Fri, 14 May 2004, James H. Cloos Jr. wrote:

> Is there any contra-indication to including libsrtp in the * dist and
> using it for encrypting rtp and rtsp streams?
> 
> The license is revised-bsd-like so it should be OK under both of *'s
> licenses.
> 
> The current version (1.3.20) is rfc 3711 compliant.
> 
> The api is simple enough; once the sessions are started you only need
> to call srtp_protect() on each outgoing packet and srtp_unprotect()
> on each incoming packet.
> 
> I can post a patch in mantis if there is interest; initially just one
> to incorporate the lib, later to actually use it.
> 
> -JimC
> 
> References in order of appearance:
> 
> http://srtp.sf.net/
> http://srtp.sf.net/license.html
> http://srtp.sf.net/srtp-1.3.20.tgz
> http://www.ietf.org/rfc/3711.txt
> http://srtp.sf.net/libsrtp.pdf
> 
>