[jitsi-dev] SIP Via/Contact headers have wrong IP address on multhomed Mac


#1

Emil,

:That's exactly what we do. Apparently it is failing in the new Jitsi
:stable. This could be due to Java 7 or something else. We should have
:a look in the following days since many people are seeing the issue.

FWIW, Jitsi 2.4-latest works. I'd be glad to try out the theory that
it's an embedded JRE issue, but it's unclear to me how to point Jitsi
at a different JRE (I have Apple Java6 and Oracle Java7 installed). I
would be glad to do some bisectional testing if there were much-older
2014 builds of Jitsi available.

I started looking at the code a bit, but I'm no Java or OO programmer.
About the only thing I concluded is that getListeningPoint() (which is
invoked quite early on in the ContactHeader/ViaHeader routines) would
appear to be obsolete with current JAIN SIP, that getListeningPoints()
should be used. I have no idea of how (ir)relevant that is, here.

-Mike

···

--
Michael J. O'Connor mjo@dojo.mi.org
=--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--==--=
"True observation begins when one is devoid of set patterns." -Bruce Lee


#2

If building the sources is an option you can switch between JRE's by changing the JAVA_HOME variable, e.g.:
$ export JAVA_HOME=/System/Library/Frameworks/JavaVM.framework/Versions/1.6/Home/
$ ant run

Regards,
Boris

···

On 08/02/15 21:50, Mike O'Connor wrote:

Emil,

:That's exactly what we do. Apparently it is failing in the new Jitsi
:stable. This could be due to Java 7 or something else. We should have
:a look in the following days since many people are seeing the issue.

FWIW, Jitsi 2.4-latest works. I'd be glad to try out the theory that
it's an embedded JRE issue, but it's unclear to me how to point Jitsi
at a different JRE (I have Apple Java6 and Oracle Java7 installed).