Bug 26270

Summary: sends session-accept before the Tp user has responded
Product: Telepathy Reporter: Simon McVittie <smcv>
Component: fargoAssignee: Simon McVittie <smcv>
Status: RESOLVED FIXED QA Contact: David Laban <david.laban>
Severity: normal    
Priority: high Keywords: patch
Version: git master   
Hardware: Other   
OS: All   
URL: http://git.collabora.co.uk/?p=user/smcv/telepathy-fargo.git;a=shortlog;h=refs/heads/acceptance
Whiteboard: milestone4.7 est=7h
i915 platform: i915 features:
Bug Depends on: 26413    
Bug Blocks:    

Description Simon McVittie 2010-01-27 06:11:20 UTC
# FIXME: should Fargo really accept the call before the XMPP client
            # has accepted the call?
            # In this case, it's the only way to get the codecs out of
            # telepathy so we can send session-initiate, so yes.
            EventPattern('dbus-method-call', handled=False,
                interface=cs.CHANNEL_IFACE_GROUP, method='AddMembers',
                path=chan.object_path, args=[[conn.self_handle], '']),

14:00 < smcv> alsuren: er, is it really necessary to call AddMembers before 
              doing all the codec wibbling?
14:00 < smcv> alsuren: I would expect that that ought to work even while the 
              local user is still in local-pending state

14:07 < alsuren> smcv: I was going by the description on: 
file:///usr/share/gtk-doc/html/telepathy-spec/org.freedesktop.Telepathy.Channel.Type.StreamedMedia.html. I guess you could set the stream direction yourself 
                 before you accept the call
14:08 < alsuren> not tried it though
14:08 < smcv> mm
14:08 < smcv> I should try that
14:09 < smcv> it doesn't look very good for a user of this gateway if Fargo 
              claims the XMPP user has answered the call, before they actually 
14:09 < smcv> the SIP user will potentially be going "hello? hello? you there?" 
              for several seconds
Comment 1 Simon McVittie 2010-01-27 09:44:38 UTC
7h? It's just a matter of re-ordering events, but they're quite subtle events and will need careful testing.
Comment 2 Simon McVittie 2010-02-18 05:26:49 UTC
smcv/acceptance has patches.
Comment 3 Simon McVittie 2010-02-18 10:32:31 UTC
Fixed in git
Comment 4 Simon McVittie 2010-02-18 14:07:36 UTC
The reverse of the originally-reported problem has been reinstated as a hackish way to get calls working again.
Comment 5 Simon McVittie 2010-02-18 14:21:08 UTC
... or, in fact, not, because Sjoerd is ninja.

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.