<?xml version="1.0" encoding="utf-8"?>
<feed xml:lang="en" xmlns="http://www.w3.org/2005/Atom"><title>Recent changes to patches</title><link href="https://sourceforge.net/p/peers/patches/" rel="alternate"/><link href="https://sourceforge.net/p/peers/patches/feed.atom" rel="self"/><id>https://sourceforge.net/p/peers/patches/</id><updated>2011-06-25T00:45:38Z</updated><subtitle>Recent changes to patches</subtitle><entry><title>Making Peers an Applet and an Application</title><link href="https://sourceforge.net/p/peers/patches/20/" rel="alternate"/><published>2011-06-25T00:45:38Z</published><updated>2011-06-25T00:45:38Z</updated><author><name>anirudh r</name><uri>https://sourceforge.net/u/anirudhrx/</uri></author><id>https://sourceforge.netb1130e98a73ad690882da86635af47bdf0e70324</id><summary type="html">Just figured out the patch system\!
Please review and apply.
If the patch applied is incomplete, please use http://www.mediafire.com/?3z8enp6uxu68un8</summary></entry><entry><title>send 200 OK on notify requests</title><link href="https://sourceforge.net/p/peers/patches/19/" rel="alternate"/><published>2011-02-12T12:34:26Z</published><updated>2011-02-12T12:34:26Z</updated><author><name>yohannmartineau</name><uri>https://sourceforge.net/u/yohannmartineau/</uri></author><id>https://sourceforge.net09cfadd164780f0ab8d8be3cf10496309ec24526</id><summary type="html">some asterisk instances seem to consider user-agents offline if they don't answer notify requests.

To bypass this issue, a patch has been developped by cristian david cepeda piscal. This patch adds a NotifyHandler that simply answer 200 OK to notify requests.

After this modification, asterisk apparently considers peers online and can route calls towards peers.</summary></entry><entry><title>Wrong order in dialog state and SipListener handling.</title><link href="https://sourceforge.net/p/peers/patches/18/" rel="alternate"/><published>2011-01-11T13:39:52Z</published><updated>2011-01-11T13:39:52Z</updated><author><name>uiltje</name><uri>https://sourceforge.net/u/tuijldert/</uri></author><id>https://sourceforge.netc48f5f18f56fcb84c45a70d6719bfe3cafb0360d</id><summary type="html">Within the InviteHandler.handleInitialInvite\(\), order of execution is:
1\. send 180 Ringing
2\. tell SipListener about incoming call
3\. update dialog state/handling.

3\. should be done before 2. because the dialog has the wrong state when calling the SipListener.
If the SipListener decides to immediately accept the call, a 200 Ok will be sent, the dialog will be set from init to confirmed, followed by an attempt to set it from confirmed to early, leading to an IllegalStateException.

Attached patch reverses the order.
Please review and apply.

Hope this helps,
Tom.</summary></entry><entry><title>No serialization in client transaction.</title><link href="https://sourceforge.net/p/peers/patches/17/" rel="alternate"/><published>2010-12-15T11:08:54Z</published><updated>2010-12-15T11:08:54Z</updated><author><name>uiltje</name><uri>https://sourceforge.net/u/tuijldert/</uri></author><id>https://sourceforge.net4835a725b1bb0e7f49159a0aa69cf6136753628b</id><summary type="html">Multiple SIP responses can be executed in parallel within 1 ClientTransaction \(probably interfering with transaction state handling as well\).

This can lead to 200 OK responses overtaking 180 Ringing responses.

Attached patch serializes this by synchronizing the "receivedResponse\(\)" method.

Hope this helps,
Tom.</summary></entry><entry><title>start cseq at 1 instead of 0</title><link href="https://sourceforge.net/p/peers/patches/16/" rel="alternate"/><published>2010-10-29T13:43:50Z</published><updated>2010-10-29T13:43:50Z</updated><author><name>yohannmartineau</name><uri>https://sourceforge.net/u/yohannmartineau/</uri></author><id>https://sourceforge.net68698f5cc985311c1b1f3c6e826f2004da957c48</id><summary type="html">Some buggy registrars seem to refuse REGISTER with a cseq of 0. This value is valid in RFC3261, but most sip softphones use 1 as their first cseq value.
So it makes sense to use 1 as the first cseq value in peers.
To use 1 as the first cseq value in peers, modify net.sourceforge.peers.sip.core.useragent.UserAgent.UserAgent. In constructor, use cseqCounter = 1;</summary></entry><entry><title>modified timerE handling</title><link href="https://sourceforge.net/p/peers/patches/15/" rel="alternate"/><published>2010-10-22T15:55:20Z</published><updated>2010-10-22T15:55:20Z</updated><author><name>uiltje</name><uri>https://sourceforge.net/u/tuijldert/</uri></author><id>https://sourceforge.net287600126d8521f6494a2654399fc9e4247574ce</id><summary type="html">On receiving provisional responses \(1xx\) within a non-INVITE client transaction, it is logical to reset the re-transmission timer, as the server side has indicated that it is working on it.

This is currently not happening. The timer is not reset and a retransmission is done when timer E fires regardless.

Attached patch will reset the timer on receiving a provisional response.

Hope this helps,
Tom.</summary></entry><entry><title>stop sending rtp when not running.</title><link href="https://sourceforge.net/p/peers/patches/14/" rel="alternate"/><published>2010-10-22T15:47:12Z</published><updated>2010-10-22T15:47:12Z</updated><author><name>uiltje</name><uri>https://sourceforge.net/u/tuijldert/</uri></author><id>https://sourceforge.netf90de7b04453052211f51a8abceb88a73402b9e3</id><summary type="html">When an rtp-stream is  stopped, any packets, waiting to be sent, are still sent.

This patch prevents that.

Hope this helps,
Tom.</summary></entry><entry><title>Timing off on closing rtp streams</title><link href="https://sourceforge.net/p/peers/patches/13/" rel="alternate"/><published>2010-10-21T15:05:06Z</published><updated>2010-10-21T15:05:06Z</updated><author><name>uiltje</name><uri>https://sourceforge.net/u/tuijldert/</uri></author><id>https://sourceforge.net1d37fd7c35516744cb25a4489179d5eea133c53d</id><summary type="html">When refreshing \(dropping and creating\) rtp-streams, time to wait for closure should depend on RtpSession.TIMER.

Attached patch does just that.

Hope this helps,
Tom.</summary></entry><entry><title>Remove superfluous JPanel</title><link href="https://sourceforge.net/p/peers/patches/12/" rel="alternate"/><published>2010-10-21T09:48:52Z</published><updated>2010-10-21T09:48:52Z</updated><author><name>uiltje</name><uri>https://sourceforge.net/u/tuijldert/</uri></author><id>https://sourceforge.net7776b6168a8c5fda7a75d5d2bcd5bac1aeaf0ede</id><summary type="html">Creation of MainFrame.mainPanel is not needed.
The JFrame.getContentPane\(\) will get you the Container needed.

Attached patch gets rid of the JPanel creation.

Hope this helps,
Tom.</summary></entry><entry><title>Enhancement of register handler</title><link href="https://sourceforge.net/p/peers/patches/11/" rel="alternate"/><published>2010-10-15T13:09:52Z</published><updated>2010-10-15T13:09:52Z</updated><author><name>uiltje</name><uri>https://sourceforge.net/u/tuijldert/</uri></author><id>https://sourceforge.net9617b2589daa7bebec98548fe3ebe44e7b1ee0fa</id><summary type="html">Attached, a patch that enhances the handling of registrations, including:
\- introduction of "Expires" header \(including default\).
\- addition of unregister all \(wildcard\)
\- improved separation of functionality - RegisterHandler  handles REGISTERs under control of InitialRequestManager

Hope this helps,
Tom.</summary></entry></feed>