Re: sipgate problem workaround

On Mon, 17 Jan 2005, Thomas Ries wrote:

> I can just try to guess what code lines you are talking about (a diff
> might have been better). IF (as I assume) the code you commented out
> lies in the function msg_make_template_reply() this would make me
> wonder, BECAUSE this function is solely used for locally processed
> requests (phone -> siproxd) and not requests directed to the provider
> (phone -> siproxd -> sipgate). A request beeing processed there will
> never be sent to the outside world.

strange... i wonder if they actually changed their authentication scheme
while i tried out changes on the code...

argghhh... now it works with the original source, so they might have 
actually changed something!!!!!

> So I would be very interested in details what you found out about
> sipgate and using VIA headers for authentication. Once the problem is
> fully understood, it should be possible to provide a
> "clean" fix/workaround that does not break other things.

well, basically the sipgate support told me that they used the VIA header
for authentication purposes and that this would not work because
siproxd added another VIA header. so I wrote them back that they should
simply check all via headers in the queue instead of the first one, but 
did not get an answer. So I guess they changed something today, and now 
it works with the regular siproxd, and it was pure coincidence that I was
experimenting with changing the source code today...

-- 
peter koellner <[email protected]>


-------------------------------------------------------
The SF.Net email is sponsored by: Beat the post-holiday blues
Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt