I'll include the fix soon as an optional feature. PS to Craig: this issue is partially addressed by the OpenH323 stack. In addition, it should not send empty Facility with H.245 tunneled message when communicating between >=V4 endpoints. Instead it should use Facility_UUIE with reason set to transportedInformation. But the gatekeeper in routed mode should have an option to translate Facility messages carrying H.245 tunneled elements (body: empty->facility(transportedInformation) and facility(transportedInformation)->empty) to ensure that older (probably badly designed and not checking Facility reason) H.323 implementations understand the purpose of Facility sent. ----- Original Message ----- From: "iTS Consult@ncy" <itsc99@xxxxxxxxx> Sent: Thursday, February 26, 2004 9:24 PM > My question is if this can or should be applied to gnugk like some > autosensing feature without having to guess why your calls get disconnected > when dealing with non v4 complient endpoints under certain circumstances. > ( tunneled) > > I have made my fix but outsite gnugk and simply would like to know if it can > be more transparent and for everyone who is using gnugk. After all is looks > like a backward compatability problem to me. > > Tjapko. > > > -----Original Message----- > From: openh323gk-users-admin@xxxxxxxxxxxxxxxxxxxxx > [mailto:openh323gk-users-admin@xxxxxxxxxxxxxxxxxxxxx]On Behalf Of > Zygmuntowicz Michal > Sent: Jueves, 26 de Febrero de 2004 10:22 p.m. > To: openh323gk-users@xxxxxxxxxxxxxxxxxxxxx > Subject: Re: Interoperability Issues > > > That's not related to the terminalCapability set. You need > to set Facility body to empty for V1,2,3 endpoints (some work > without this, some does not) when sending H.245 tunneled > messages. The V4 recommendation requires the body > to carry Facility_UUIE with reason code set to transportedInformation. > The problem is that some older endpoints receiving Facility > with non-empty body think it is a call redirection. > > ----- Original Message ----- > From: "iTS Consult@ncy" <itsc99@xxxxxxxxx> > Sent: Thursday, February 26, 2004 3:45 PM > > > > I experienced similar problems and posted this a while ago and I think it > is > > coming from a H323 protocol mismatch between version 2 and 4. I am not > > really sure about this. > > > > Bottom line: If you can hack into the code than be sure that you replace > the > > empty with a facility or the facility with an empty TerminalCapabilitySet > > message to make it all happen. > > > > I have found this in the ITU h323 manual as a change in the latest v4 > h323 > > protocol and they mention that there are problems with backward > > compatibility. > > > > Unfortunately I am not a coder myself just a dumpanalyser so if you can > > solve this issue than build it into the gnugk. > > > > Tjapko. ------------------------------------------------------- SF.Net is sponsored by: Speed Start Your Linux Apps Now. Build and deploy apps & Web services for Linux with a free DVD software kit from IBM. Click Now! http://ads.osdn.com/?ad_id=1356&alloc_id=3438&op=click _______________________________________________ List: Openh323gk-users@xxxxxxxxxxxxxxxxxxxxx Archive: http://sourceforge.net/mailarchive/forum.php?forum_id=8549 Homepage: http://www.gnugk.org/
